Citation
Release and uptake of ferritin absorbed at tin doped indium oxide electrode

Material Information

Title:
Release and uptake of ferritin absorbed at tin doped indium oxide electrode
Creator:
Pyon, Moon Son
Publication Date:
Language:
English
Physical Description:
x, 87 leaves : illustrations ; 28 cm

Thesis/Dissertation Information

Degree Divisions:
Department of Chemistry, CU Denver
Degree Disciplines:
Chemistry

Subjects

Subjects / Keywords:
Ferritin ( lcsh )
Biological transport ( lcsh )
Iron proteins ( lcsh )
Biological transport ( fast )
Ferritin ( fast )
Iron proteins ( fast )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Bibliography:
Includes bibliographical references (leaves 83-87).
Thesis:
Department of Chemistry
Statement of Responsibility:
by Moon Son Pyon.

Record Information

Source Institution:
University of Colorado Denver
Holding Location:
Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
42613755 ( OCLC )
ocm42613755
Classification:
LD1190.L46 1999m .P96 ( lcc )

Full Text
RELEASE AND UPTAKE OF FERRITIN
ADSORBED AT TIN DOPED INDIUM OXIDE ELECTRODE
by
Moon Son Pyon
B.S., University of Colorado at Denver, 1997
A thesis submitted to the
University of Colorado at Denver
in partial fulfillment
of the requirements for the degree of
Master of Science
Chemistry


This thesis for the Master of Science
degree by
Moon Son Pyon
has been approved
by
Donald C. Zapien
tf -xi m
Date


Moon Son Pyon (M.S., Chemistry)
Release and Uptake of Ferritin Adsorbed at Tin doped Indium Oxide
Electrode
Thesis directed by Assistant Professor Donald C. Zapien
ABSTRACT
In this experiment, the uptake and release of iron by ferritin using
direct electrochemical techniques has been investigated. Cyclic
voltammetry, the principal method used in this study, reveals that the
ferritin in pH 7.0, p. = 1.0 M phosphate buffer adsorbed onto an ITO
electrode at open circuit potential. The ferritin adsorbed at the ITO
electrode is electroactive and results in about one monolayer of ferritin.
The pH dependence of the reduction of the ITO electrode with an adsorbed
layer of ferritin results the slope of-125 mV/pH, which suggests that 2H+
transferred into iron core upon reduction reaction. This result suggests that
adsorbed ferritin behaves similar to ferritin in solution. Iron is released
from ferritin in presence of EDTA after the iron is first reduced from Fe (III)
to Fe (II), however, EDTA does not directly act on ferric iron without the
corporation of reduction. Adsorbed ferritin, emptied by electrochemical
reduction in presence of EDTA, was exposed to ferrous ion at 0.20 V. The
m


current-potential curves show the original peaks pattern, suggesting that
iron had been re-incorporated into the apoferritin shell. An authentic
sample of adsorbed apoferritin exposed to ferrous ion exhibited the same
current-potential response as electrochemically emptied ferritin,
supporting that apoferritin results from the ITO/ferritin electrode in the
presence of EDTA. Applying molecular oxygen for the oxidation of Fe(ll)
results the same current-potential response as applying the controlling
potential at 0.20 V, suggesting that the oxidizing potential has the same
oxidation effect as molecular oxygen. These studies show that ferritin
adsorbed at the ITO electrode is a promising venue to study not only
ferritins electrochemistry but also its function.
This abstract accurately represents the content of the candidates thesis.
I recommend its publication.
Signed
IV


ACKNOWLEDGMENTS
I would like to take this opportunity to thank my advisor, assistant professor
Dr. Donald C. Zapien, for his support, guidance, patient, and
encouragement through this work. 1 also wish to express my sincere
thanks to my committee members, Dr. Lanning and Dr.Levy for their
technical advice, assistance, and encouragement. Special thanks to
Peggy Lore, Asian American student services' advisor, for her support and
understanding.
v


CONTENTS
Chapter
1. Introduction....................................................1
1.1 The structure of horse spleen ferritin..........................1
1.2 The function of horse spleen ferritin...........................7
1.3 Iron rlease of horse spleen ferritin............................7
1.3.1 Iron release by chelators.......................................8
1.3.2 Iron release by reductants......................................8
1.3.3 Iron release by redox reactions.................................9
1.3.4 The pH dependence of iron release..............................10
1.4 Iron uptake of horse spleen ferritin.....................'.....11
1.4.1 The general mechanism of iron uptake...........................11
1 4.2 The inhibitory effects on iron uptake..........................13
1.5 Recombinant H/L chain in horse spleen ferritin.................14
1.5.1 Recombinant H chain in ferritin (rHF)..........................15
1.5.2 Recombinant L chain in ferritin (rLF)......................... 16
1.5.3 The pH dependence of recombinant H/L chain in ferritin.........16
1.6 The role of phosphate in ferritin..............................17
1.7 Electrochemistry...............................................19
VI


1.7.1 Electrochemical systems..........................................19
1.7.2 Nernst equation..................................................21
1.7.3 Cyclic voltammetry...............................................22
1.7.4 Electrochemical measurements.....................................25
1.8 Research goal....................................................31
2. Experimental section.............................................32
2.1 Chemicals and equipments used....................................32
2.2 Prevention of contamination......................................33
2.3 Ferritin (or apoferritin) purification...........................34
2.4 Concenetration determination of ferritin (or apoferritin)........35
2.5 Elelctrochemical equipments......................................36
2.5.1 Electrochemical cell.............................................36
2.5.2 Preparation of tin doped indium oxide eledctrodes................39
2.5.3 Preparation of reference and auxiliary electrodes................39
2.6 Experimental procedure...........................................40
2.6.1 The preparation of ITO/ferritin electrodes
with a layer of adsorbed ferritin................................40
2.6.2 CV of an ITO/ferritin electrode in phosphate buffer..............42
2.6.3 The removal of iron from horse spleen ferritin...................43
2.6.4 Iron uptake by horse spleen "apoferritin"........................44
2.7 Caculation of surface concetration...............................46
VII


3. Result and discussion...............................................47
3.1 Characterization of an ITO/ferritin electrode.......................47
3.2 CV of an ITO/ferritin electrode in phosphate buffer.................48
3.2.1 The scan rate dependence of the ITO/ferritin electrode..............51
3.2.2 The effects on packing density of the ITO/ferritin electrode........54
3.2.3 The pH dependence of the ITO/ferritin electrode.....................57
3.2.4 CV of an ITO/ferritin electrode in HEPES buffer.....................59
3.3 The removal of iron from horse spleen ferritin......................61
3.3.1 Iron release by an electrochemical reduction
in the presence of EDTA.............................................62
3.3.2 Iron release without the reduction in the presence of EDTA..........63
3.3.3 Iron release in HEPES buffer ..................................... .66
3.4 Iron uptake by horse spleen "apoferritin"...........................68
3.4.1 Iron uptake by the immersion potential of electrode at 0.20 V.......68
3.4.2 The immersion potential of electrode dependence.....................72
3.4.3 Iron uptake by molecular oxygen.....................................74
3.4.4 Iron uptake by commercially availale apoferritin....................78
4. Conclusion..........................................................81
5. Bibliography........................................................83
VIII


FIGURES
Figures
1.1 Schematic representation of ferritin................................ 2
1.2 A cross section of ferritin..........................................3
1.3 Ribbon diagram of the alpha backbone
of a ferritin subunit................................................5
1.4 A three-electrode system............................................20
1.5 Cyclic voltammetry..................................................23
1.6 Indirect electron transfer method by a mediator.....................26
1.7 Indirect electron transfer method by a promotor.....................28
1.8 Direct electron transfer method.....................................30
2.1 Electrochemical cell configuration..................................37
2.2 Glass-sheathed ITO electrode........................................38
2.3 Overal configuration of electrochemical equipments..................41
3.1 Current-potential curve of an ITO/ferritin electrode
in phosphate buffer.................................................49
3.2 The scan rate dependence of an ITO/ferritin electrode
(Peak current vs. scan rate)........................................52
3.3 The scan rate dependence of an ITO/ferritin electrode
(AEP of the reconstruced layer vs. scan rate).......................54
3.4 The adsorption time dependence of an ITO/ferritin electrode
(Rel. initial cathodic peak area vs. adsorption time)...............55
IX


3.5 The concentration of ferritin dependence of an ITO/ferritin
electrode (Rel. initial cathodic peak area vs. concentration)....56
3.6 The pH dependence of an ITO/ferritin electrode
(pH vs. initial cathodic peak potential).........................58
3.7 Current-potential curve of an ITO/ferritin electrode
in HEPES buffer..................................................60
3.8 Current-potential curve of an ITO/ferritin electrode
in phosphate/EDTA buffer.........................................64
3.9 Schematic representation of iron release
of an ITO/ferritin electrode ....................................65
3.10 Current-potential curve of an ITO/ferritin electrode
in HEPES/EDTA buffer.............................................67
3.11 Current-potential curve of an ITO/"apoferritin" electrode
in uptake of iron................................................71
3.12 The immersion potential of electrode dependence
in uptake of iron................................................73
3.13 Oxygen control study in uptake of iron..............................77
3.14 Apoferritin control study in uptake of iron.........................80
x


1.
Introduction
1.1 The structure of horse spleen ferritin
Ferritin is a large protein, which can be found in most cell types for
human, and other vertebrates, and in higher plants, fungi, and bacteria.
Ferritin is found in high concentrations in the liver, spleen, and the bone
marrow in vertebrates [1]. Ferritin is roughly spherical in shape having a
diameter of 12.5 nm and 1.0 nm in thickness. It is composed of a three
dimensional organization of protein shell that has a molecular weight of
450,000g mol'1. The protein shell is composed of individual 24 subunits
(figure 1.1). The subunits surround a core of hydrous ferric oxide
tFe(lll)00H] that is capable of storing as many as 4,500 iron Fe(lll) atoms
with variable amounts of phosphate, as known as the iron core. The average
composition of the iron core is ((FeOOH)a FeOP3H2) (figure 1.2). When the
protein shell doesnt contain any iron, it is referred to as apoferritin [2, 3].
The arrangement of the 24 subunits that are associated in 4:3:2
symmetry results in eight hydrophilic and six hydrophobic channels which link
outside to interior of the shell [4].
1


Figure 1.1
8 Hydrophilic Channels
6 Hydrophobic Channels
125 A --------------1
Apoferritin consists of twenty-four subunits with 4-3-2 symmetry
composed of varying ratio of L and H chain [3],
2


Protein subunit
The mi tral structure of
[(FeOOHJsFeOPaHJ
Protein channel
Fe (III) core
~12.5 nm
pi = 4.5
Figure 1.2 A cross-section of the ferritin.
3


Interactions of three subunits near the N terminus of each subunits, which is
lined with glutamic and aspartic acid residues, results in eight hydrophilic
channels, and four subunits near the E terminus of each subunits which is
lined with leucine residues which result in six hydrophobic channels. Since
the hydrophilic channels have been proposed as metal binding sites, they are
probable passages where iron passes into/out of the interior of the protein for
iron formation/removal processes [5]. The postulated role of these six
hydrophobic channels has been known as a passageway for oxygen and
phosphate [6].
Each subunit is represented by a sausage-shaped building block with
N and E terminus. Each of 24 subunits, contained 174 amino acids, is
composed of four closely packed major helices (A, B, C, and D) with a shorter
helix E located to the adjacent to the primary axis and opposite to the N-
terminus (figure 1.3). The four major helices are held closely together due to
hydrophobic interactions. The C terminus of helix B with the N terminus of
helix C is joined by a long loop L [3]. The study of horse spleen ferritin at a
2.8 A resolution indicates that 129 of the 174 amino acids residues are
located in these five helices [7j. Helices A and C locate on the outside and
helices B and D are on the inside of the protein.
4


Figure 1.3 Ribbon diagram of the alpha-carbon backbone
of a ferritin subunit [3].
5


There are many hydrogen and salt bridges contained within each subunit and
between subunits to give a high level of strength to the tertiary structure and
thermostability [6].
The subunits are composed of varying ratio of two distinct types of
subunits. The L-type or light subunits have a molecular mass of 19,766 g
mol'1 and the H type or heavy subunits have a molecular mass of 21,099 g
mol'1. Previous experiment suggests that both L and H subunits have
virtually identical tertiary conformation and share 55% of amino acid
sequence identity [8]. The major difference of two types of subunits are that
L subunit contains a metal binding site for iron storage, and H subunit
contains a ferroxidase center for catalyzing iron oxidation. Moreover, the
difference of H and L subunits occurs in the amino acid sequence, surface
charge, and iron content in ferritin [2]. The heavier H subunits are
associated with a lower pi and lower iron content, and no major iron storage
function, and predominates in horse heart ferritin, whereas the lighter L
subunit is associated with a higher pi and higher iron content, and
predominates in horse spleen ferritin. In horse spleen ferritin 22 subunits
and 2 subunits out of 24 subunits are L and H respectively thus,
approximately more than 90% of the subunits is L type. In contrast, horse
heart ferritin contains approximately 90% of H type subunit. The molecular
6


weights of the relative proportion of H and L subunits increases from
460,000g mol'1 for spleen and to 515,000g mol'1 for heart apoferritin [9].
1.2 The function of horse spleen ferritin
The main function of ferritin is to sequester excess iron in the cell.
Ferritin also stores excess iron in soluble form in the mineral core within
protein shell, until it is needed in the synthesis of other iron containing
protein such as hemoglobin, myoglobin, and cytochrome c [2]. Although
ferritin sequesters principally iron, it also binds several nonferrous metal
ionssuch as Cu2+, Zn2+, Cd2+, Tb3+, U022\ Cr3+, or V02+, though in lesser
amounts. Horse spleen ferritin is useful standard against which other species
are compared because the primary and three-dimensional structures are
already known and because it is nearly entirely composed of identical
subunits [2].
1.3 Iron release of horse spleen ferritin
An understanding of the mechanisms by which iron is released by
ferritin is of importance not only to biochemists, but also physicians who are
interested in managing diseases of iron overload [2]. Since the available
evidence suggests that iron release occurs through three fold eight
7


hydrophilic channels, determining the mechanism by which iron exits the
protein shell is one of the most pressing issues in ferritin research.
1.3.1 Iron release by chelators
In the earliest in vitro studies of iron release, it has been suggested
that Fe(lll) can be released from ferritin by chelators with high affinity for
Fe(lll) attesting to the stability of the ferritin complex, but the action of Fe(lll)-
chelators alone is extremely slow. It was thought that chelators penetrate
the protein shell and act directly on the iron core [11]. Then, Fe(lll)-chleate
complexes may exit through the hydrophilic channels [12]. The result of
previous study shows that about 2 ~ 10 % of iron was released from horse
spleen ferritin by chelators thus, chelation of Fe(lll) is not likely to be a
physiologically significant mechanism. Agents such as 1,10-phenanthroline,
bipyridyl, desferrioxamine B, rhodotorulic acid, and EDTA have all been
used as in vitro chelators for Fe(lll) [2, 10].
1.3.2 Iron release by reductants
Using in vitro experiments, it was found that iron can be induced to exit
ferritin faster and more completely if the iron is first reduced from Fe3+ to Fe2+
by reducing agents like thioglycollate and dithionite, at pH 4.5-5.0 and
8


reduced flavins at physiological pH values. Reductants that are effective in
vitro also include reduced flavins, sulfhydryl compounds (thioglycolic acid,
dithionite, and dihydrolipoic acid), Cu(ll) plus ascorbate, free radicals of
oxygen (superoxide) and methylviologen [2, 13]. When ferritin is treated with
reducing agents, the ferritin readily releases iron at approximately one million
times faster than the rate when ferritin was not treated with reductants [14].
This result was supported by another study that showed that ferritin must be
reduced before the release of the inner core ferritin iron [15, 57]. Reduction
of Fe3+ to Fe2+ appears to be the most likely mechanism of iron release in
vivo. Jones and co-workers study suggests that the most effective iron
release is effected when the Fe(lll) is first converted to Fe(ll), which in turn is
readily lost in the presence of Fe chelators [11].
1.3.3 Iron release by redox reactions
Several reductants and chelators known to be capable of releasing
iron are apparently too large to pass through the channel. It was once
believed those small reductants, such as dithionite and thoglycolic acid might
enter the protein shell and effect the iron release. However, large reductants
such as FMNH2 [16], xanthine [17], and NADH [13], that have been used to
reduce ferritin, are too large to pass through the channels. Therefore, it was
9


suggested that redox reactions involving the Fe(ll)/Fe(lll) components of the
iron core in the ferritin can occur without direct interaction of the redox
reagent at the mineral core surface. Electron tunneling has been suggested
as a possible mechanism for a long-range electron transfer in ferritin.
Alternatively, the electrons might pass through the shell mediated by iron
atoms residing in the channels [18].
1.3.4 The pH dependence of iron release
In micro coulometric experiments in which electron transfer mediators
were used, it was observed that 2H+ are transferred to the core for every iron
reduced. The following reaction occurring in the core was proposed [19, 20]:
The reduction of Fe(lll) ions in ferritin requires to take up two protons from the
surrounding medium per electron transfer to the protein shell. Two protons
are used for the protonation of hydroxide ions coordinated to surface Fe(lll).
Rates of reduction and iron release are also affected by a variety of
conditions, including phosphate content in the core, iron core size, the
modification of protein shell and buffer ions [21].
10


1.4 Iron uptake of horse spleen ferritin
1.4.1 The general mechanisms of iron uptake
In 1972, it was shown that the uptake of iron could be followed
spectrophotometically using apoferritin or ferritin containing low amounts of
free Fe2+ and an oxidizing agent. Iron is oxidized at a greater rate in the
presence of ferritin than free iron alone; thus ferritin was considered an
enzyme [22]. In another early study, it was suggested that iron, as Fe(ll) or
Fe(lll), most probably entered into the protein shell through three fold
hydrophilic channels, as a neutral chelate complex, but the entering as bare
Fe2+ or Fe3+ ions was unlikely to have occurred [23]. Later, it was proved that
the iron core could not be formed by adding Fe(lll) to apoferritin. Many
experiments have shown that the formation of iron core in apoferritin can be
achieved by starting with Fe(ll) and apoferritin in the presence of suitable
oxidants. In other words, the formation of the iron core in ferritin involves the
oxidation of Fe(ll) to Fe(lll), catalyzed by apoferritin in the presence of either
02 or KI03, yielding peroxide as a by-product. This was followed by the
hydrolysis of the hydrated Fe(lll) and the formation of core crystallites [3, 24,
56], Then, the crystallites of hydrous ferric oxide begin to grow within the
protein shell and once the iron core is formed, further additions of Fe2+ can be
deposited and oxidized directly on the surface of these crystals. At that
11


stage, the rate of core growth is dependent on the surface area of hydrous
ferric oxide crystallite [25]. In wild-type ferritin, the formation of ferritin from
apoferritin is self-catalyzing and probably involves the formation of a dimer,
Fe(ll)-0-Fe(lll) or a cluster of 3-4 Fe(lll)-oxo atoms; such dimers have been
detected in several iron-protein by EPR spectroscopy [2]. Studies using
Mossbauer and EXAFS spectroscopy and other methods have revealed that
larger clusters form, probably at nucleation sites on the inner surface of the
protein, followed by continued iron oxidation on the developing mineral core
[27],
The details of the iron core formation mechanism have been described
in several ways. One study suggests that while Fe(ll) passes through the
three fold hydrophilic channels, the Fe(ll) becomes oxidized. Then, Fe(lll)
binds directly to the nucleation site on the inner surface of the protein shell as
soon as Fe(lll) arrives at the central cavity [3]. Another study proposes that
Fe(ll) enters the central cavity through three-fold hydrophilic channels in the
protein shell, without the oxidation of Fe(ll). It was previously suggested that
oxidation sites and nucleation sites are located in distinctly different regions
of the shells inner surface of the protein shell. The inter-subunit channels
provide a passageway for the entry of Fe(ll), and Fe(ll) is not oxidized until it
is inside the shell. Once Fe(ll) arrives inside of the protein shell, Fe(ll) is
12


bound at the oxidation site. Following oxidation of Fe(ll), Fe(lll) migrates and
consequently binds to a specific nucleation site [19]. Alternatively, the site of
oxidation also serves as a nucleation site for iron core formation [11, 22].
Another description proposes that after Fe(lll) is generated, autonucleation
occurs spontaneously forming a mineral core without specifically binding to
the protein [3],
1.4.2 The inhibitory effects on iron uptake
Apoferritin has the biggest effect on iron core formation at the
beginning stages of core formation, presumably involving the complexation of
iron by carboxyl groups. It has been shown that chemical modification
(esterification) of the carboxyl groups in the channel prevents the iron core
formation. It has been suggested that one or more carboxyl groups are
essential for the catalytic activity for uptake and oxidation of iron by
apoferritin. The modification of more than seven carboxyl groups per subunit
eliminates the capability of apoferritin to uptake iron, and its catalytic
properties. [28].
Reconstitution of ferritin is inhibited by a variety of metal ions including
Zn2+, Tb3+, Cr3*, and Vo2+. Zinc, as Zn2+ has been found to inhibit both the
formation of ferritin from Fe2+ and the formation of FeOOH. This suggests
13


that Zn2+ competes both for common binding sites on the protein at which the
oxidation the Fe2+ is catalyzed, and the sites on the hydrous ferric oxide
particles, hence inhibiting the iron core growth. However, although Tb3+ may
be competing with Fe2+ and Fe3+ for channel sites, effectively displacing them,
Tb3+ may also promote cluster formation by binding at or near the nucleation
site. Thus the iron core nucleation site may be at or near one at which Tb3+
binds but not on at which Zn2+ binds. The stoichiometry of the displacement
of Mn2+ or Vo2+ by Fe2+ suggests that these ions bind either at the same sites
or more than one site presenting in ferritin [29, 30, 58].
Cross-linking in ferritin, that is intramolecular dimers between subunits,
affects iron uptake and release by protein shell in vitro. Since the presence
of subunit dimers can explain the iron content of ferritin in vivo\ molecules
with subunit dimers have a low iron content which leads to exhibiting slower
rates of iron uptake and faster rates of iron release in vitro [2, 38, 39]
1.5 Recombinant H/L chain in horse spleen ferritin
Recent studies involving recombinant H and L chain ferritin have
shown not only marked difference in the catalytic (for Fe(ll) oxidation)
properties between the subunit types, but they also have s cooperative role in
the iron uptake mechanism and the ability to form a stable iron [31, 32, 59,
14


60], The determining cooperative role of H and L chain ferritin in iron uptake
may help to understand the biological function of an isoferritin [31].
1.5.1 Recombinant H chain in ferritin (rHF)
It is known that recombinant H chain in ferritin (rHF) contains a
ferroxidase center where catalyzes the oxidation of Fe(ll) that produces the
first product, a Fe(lll) p-oxo-bridged dimer and peroxide [24]. The oxidation
can be proceeded by the fixation of two Fe2+ ions by catalytic sites on pairs of
adjacent polypeptide chains, followed by the binding of a dioxygen molecule
between the iron atoms that is located in close proximity to enable fast two
electrons transfer to the dioxygen. Then, oxidation of two irons is
consequently followed by splitting of p-oxo-bridged Fe(lll) dimer. Hydrolysis
of this peroxo-complex, which is favored at low pH, generates that the rHF
which may be responsible for being released from the ferroxidase center for a
ferrihydrite nucleation in the cavity [28]. This is further supported by that the
Fe(ll) oxidation that occurs only at the ferroxidase center in the rHF and not in
the threefold channels even though the channels are used as the pathway for
entering Fe(ll) [32]. The presence of two H subunits in horse spleen ferritin is
sufficient to catalyze the iron oxidation.
15


1.5.2 Recombinant L chain in ferritin (rLH)
The role of recombinant L chain in ferritin (rLF) is the incorporation
and storage of iron in solution at physiological pH values. The rLF, which
lacks the ferroxidase center, has a higher capability than the rHF to induce
iron core formation while the rHF is promoting Fe(ll) oxidation. Moreover, the
presence of carboxylate ligands such as a glutamic acid residue in the inner
surface of the cavity may serve an important role either as metal binding sites
or as negatively charged environment to lower their activation energy of
cluster formation [31, 33-35].
In conclusion, each H chain ferritin contains the ferroxidase center
were Fe(ll) oxidation is catalyzed so Fe(lll) is provided rapidly for the
ferrihydrite nucleation [54]. The L chain ferritin lacks the ferroxidase center,
but it favors the nucleation process because the rLF is relatively rich in
carboxylic amino acids [8]. Once again, the L chain rich ferritin is associated
with long-term iron storage and the H chain rich with iron detoxification [55].
1.5.3 The pH dependence of
recombinant H/L chain in ferritin
Since the iron uptake is pH dependent, the oxidation and the
nucleation processes, that are segregated by the H and L chain respectively,
16


are affected by pH. It was previously shown that the ferroxidase center in
ferritin is only active to catalyze the Fe(ll) oxidation at pH below 6.0 [36]. At
pH 5.5-6.0, the oxidation reaction is favorable because an iron hydrolysis
does not occur. Thus ferroxidase center in the rHF provides a sufficient
concentration of Fe(lll) to allow the nucleation to proceed. On the contrary,
the rLF has higher efficiency in its role when it has a higher concentration of
negative charges in the cavity. The formation of a ferrihydrite is favored by
pH values above 7.0 because of the release of protons associated with
hydrolytic polymerization. Higher pH values promote in the iron hydrolysis
and induce the iron core formation in the apoferritin [31].
1.6 The role of phosphate in the iron core
Phosphate, which is a constituent of the ferritin core, alters the core
structure, crystallinity, redox properties, and local environment of iron in the
core [24], No matter where the phosphate locates, the environment of iron
near phosphate is different from iron atom in three-dimensional order and the
iron coordination to phosphate results in the heterogeneity of the rate of iron
release from ferritin in vivo [4], It has been suggested that phosphate on the
surface of the iron core interacts with amino groups on the inner surface of
the protein shell, acting as a bridging ligand for the binding of the iron core to
17


apoferritin. Such binding of phosphate throughout the core is considered
more stable than surface adsorption alone [37]. The effect of the stabilization
is to shift the reduction potential negative relative to the iron core that does
not contain phosphate [21].
Phosphate has significant influence on an iron deposition. The
pathway may involve the protein ferroxidase site where phosphate may bind
iron (II), shifting its redox potential to a more negative value and thus
facilitating its oxidation [24].
The study of the role of phosphate in the initial phase of the iron core
formation in apoferritin suggests that phosphate accelerates the rate of Fe(ll)
oxidation by decreasing the concentration of the intermediate the
mononuclear Fe(ll)-apoferritin complex and becomes bound to the iron
clusters so formed. However, it has been shown that the degree of the
disorder of the core increases as the amount of phosphate present increases
[37]. Much of phosphate is released during the early stage of iron release
[19] and can be reconstituted from apoferritin and Fe(ll) in vitro without
phosphate [24].
18


1.7 Electrochemistry
1.7.1 Electrochemical systems
/
The measurement of electrode currents as a function of the voltage
applied to the electrolysis cell can provide remarkably detailed information
about the mechanism of the reaction of interest. These measurements are
accomplished by the use of a three-electrode cell (figure 1.4). In this cell
configuration, surface of the working electrode is where the reaction of
interest occurs. An auxiliary electrode, which is a conductor, ensures that the
current only flows between the working and auxiliary electrodes, not a
reference electrode where a stable and fixed potential is provided. It is
because the passage of a current through the reference electrode will change
its chemical composition, thus changing its potential. When a sufficiently
large potential is applied between the working and auxiliary electrodes, the
oxidation or reduction reaction occurs at the working electrode. The current
that is generated by the reaction passes through the circuit and is measured
by an ammeter, and the potential of the working electrode is monitored with
respected to the reference electrode. Thus with the three-electrode
configuration no current passes through the reference electrode.
19


Figure 1.4 A three-electrode system.
20


1.7.2 Nernst equation
An equilibrium exists between the oxidized and reduced forms of a
system. The Nernst equation gives the relationship between the electrode
potential and the activities of the oxidized and reduced forms of the electrode
system.
E = E + _RT In jqx]
nF [red]
Where (E) is the electrode potential, (E) is the standard electrode potential,
R is the molar gas constant (8.313 JIC1mor1), T is the absolute temperature
measured in K, n is the number of electrons being transferred during the
redox reaction, F is the Faraday constant (96,487 C mol'1), [ox] is the
concentration of the oxidized species, and [red] is the concentration of the
reduced species.
Equation can be summarized as following;
E = E + 0.0591 log _[ox] at 25C
n [red]
The Nernst equation indicates that the equilibrium potential (E) of the
electrode results from the standard electrode potential (E) of the reaction
and the concentrations of [ox] and [red] at the electrode surface, which,
under equilibrium concentrations, are the same as their values in bulk
solution.
21


1.7.3 Cyclic voltammetry
Linear-sweep voltammetry is often used as the first technique
employed in an electrochemical investigation since it is easy to perform and
provides quick and useful information about the system under the
investigation. The potential of the working electrode is swept from a value of
E, at which species cannot undergo oxidation or reduction, to a potential E2
where the electron transfer is driven rapidly. The applied potential E is a
function of the speed at which the potential is swept (v) and the point in time
of the sweep (t).
Cyclic voltammetry involves the continuous variation of externally
applied potential while measuring the current response. Cyclic voltammetry
is an extension of linear sweep voltammagram in that when the potential of
the working electrode reaches the values of E2, the direction of sweep is
reversed and the electrode is scanned back toward the original value Ev
When the potential of the working electrode reaches the designated limits of
potential value, the potential will reach a level which will induce dissolved
species to liberate electrons (oxidation) or take up electrons (reduction).
Figure 1.5 (a) shows the term of the potential of the working electrode is
ramped negatively and positively, uniformly with time (triangular potential
cycle) and the corresponding current responses, which is shown in figure
22


(a)
Figure 1.5 (a) Variation of the applied potential as a function of time in CV.
(b) CV for a reversible electron transfer reaction.
23


1.5 (b). The scan indicates that the potential of the working electrode is first
swept from a value of E1P at which cannot undergo reduction, to a potential
E2i which generates the current, indicating the reduction reaction. The
maximum current is known as peak current, /p. On reaching E2 the potential
is then swept back, oxidizing species formed at the electrode during the
forward scan to E,. The currents acquired from the scans, are observed due
to the reduction and oxidation reaction. The relative shape of the forward
peak and the reverse peak depend on that reversibility of redox couple.
Electrons passed per unit time from a given reaction constitutes a current.
Therefore, the higher the rate of the redox reaction the higher the current.
amount of charge passed (coulombs) = current (Amps)
unit time (sec)
For any system, reversible or not in an adsorbed state, /p is directly
proportional to the surface concentration of oxidized species, the area of
electrode, and increases with the scan rate.
ip= _rFF vAr0X*
4RT
Where ip is the peak current, n is the number of electrons being transferred
during the redox reaction, F is the Faraday constant (96,487 C mol"1), A is the
area of electrode (cm2), rox* is the surface concentration of the oxidized
24


species, v is the scan rate, R is the gas constant (8.313 JK'1mol'1), and T is
the absolute temperature (measured in K).
/
i
1.7.4 Electrochemical measurements
' Electrochemical methods are well-suited to probing the
electrochemical properties of proteins, and the environment surrounding the
electroactive center. These techniques can entail either the direct electron
transfer between the redox centers of the proteins and the electrode, or
indirect electron transfer where mediators are involved.
1.7.4.1 Indirect electron transfer method using mediator
Historically, the electron transfer of proteins has been studied using
mediators such as methyl viologen. Figure 1.6 illustrates the reduction of
ferritin mediated by methyl viologen. By applying reducing potentials, the
mediator is reduced at the electrode surface and in turn, the mediator is
oxidized at the same time when ferritin is reduced. Then the mediator returns
to the electrode surface again to be reduced. This method can give the value
of n and emptying a titration can give an estimate for the formal potential.
However, no information is given for the electron transfer kinetics.
25


Figure 1.6
Ferritin (Oxidized) Ferritin (Reduced)
Reduction and oxidation of ferritin using methyl viologen (MV)
as a mediator.
26


n = (Charge) x (96,485 C/mol)
number moles of ferritin
j
1.7.4.2 Indirect electron transfer method using promoters
Modified electrodes are found in a wide range of applications, since
they offer novel properties such as the ability to hold charge, induce novel
specific reactions, act as organic semiconductors and catalyze reactions.
Functional groups on electrode surfaces, which catalyze redox reactions, are
called promoters. Promoters do not act as electron acceptors but will usually
contribute to establishing an analyte-electrode interaction favorable to
effective arrangement of the adsorbed layer of the surface electron transfer.
Promoters can be molecules adsorbed or chemically bonded to the surface.
Or it also can be groups that are inherently present on the surface of the
electrode material. The promoter, 4,4-dithiodipyrine is used to induce
cytochrome c to approach the electrode in favorable orientation for electron
transfer to occur (figure 1.7).
1.7.4.3 Direct electron transfer method
The indirect electron transfer method using a mediator has been used
extensively in early work, but it is not possible to observe the direct electron
27


v e- }
Figure 1.7 The orientation of cytochrome c to a gold electrode surface
modified by promoter, 4-thiopyridine [39].
28


transfer between the redox centers of a protein and the electrode. Direct
electrochemical methods have been used extensively to explore the redox
properties of proteins. The direct electron transfer of the protein with the
electrode allows one to observe how the protein responds directly to an
applied potential and the measurement of electron transfer kinetics.
However, it poses some difficulties. The electroactive center is sometimes
located within an insulating globular matrix. Thus, when the protein is in
contact with an electrode, electron transfer between the electrode and redox
center of the protein can not occur at appreciable rates. Also, proteins
generally absorb strongly and irreversibly on metal electrodes which cause
forming an insulating layer of denatured protein that blocks the electrode
surface form further electron transfer [39].
Many studies have shown that proteins are electroactive at tin-doped
indium oxide electrodes, and exhibit the direct electron transfer at the tin
doped ln203 electrode (figure 1.8). The class of electrodes made from
semiconducting materials, such as tin-doped indium oxide (ITO) has found
widespread utility in characterizing the electrochemical properties of proteins
other than cytochrome c [40, 41], such as myglobins [43, 44], ferredoxins [45,
46], and hemoglobin [47]. In aqueous solution, the surface of the ITO
electrode is negatively charged.
29


Ferritin (oxidized form)
Ferritin (reduced form)
Figure 1.8 The direct electron transfer of ferritin at an ln203 electrode.
30


1.8 Research goal
Since it is clear that the uptake and release of iron by ferritin involve
the oxidation and reduction reaction, describing the role of electron transfer is
key to understanding the mechanism by which iron is loaded and unloaded
by the iron storage protein, ferritin. In this work, that the electrode surface will
be exploited as the venue for observing these processes since they can be
induced electrochemically. A layer of ferritin will be formed at the ln203
electrode surface, and then the loading and unloading of iron will be induced
by controlling the electrode potential. In this study, all ferritin molecules of the
layer experience the same applied potential. In addition, the ferritin layer will
be characterized using cyclic voltammetry.
31


2.
Experimental section
2.1 Chemicals and equipments used
The commercial supplier of sodium chloride (NaCI) (99.2%), sodium
hydroxide (NaOH) (Analytical Reagent), HEPES (C8H18N204S), hydrochloric
acid, nitric acid (analytical reagent grade), and sodium phosphate, monobasic
(Analytical Reagent) was Mallinkrodt Specialty Chemicals Co. (Paris, KY).
Sulfuric acid was purchased from VWR scientific company (West Chester,
PA). Ethylene diaminetetraacetic acid (EDTA), disodium salt, dihydrate
(Analyzed Reagent), and Ferrous Ammonium Sulfate (Fe(NH4)2(S04)26H20)
were obtained from J.T. Baker Chemical Company (Philipsburg, NJ). Bio-
Rad Protein Assay G-250 Dye was purchased from Bio-Rad Laboratories
(Hercules, CA). Sigma Chemical Company (St. Louis, MO) was the source of
the following compounds: 3-mercaptopropionic acid (MPA) (99.3%), sodium
dichromate (Na2Cr204H20) (98%), bovine serum albumin (BSA) (fraction 5
powder), horse spleen ferritin (108 mg/ml), horse spleen apoferritin (50
mg/ml), sodium azide (NaN3) (> 99.0%), and phenyl methyl-sulfonyl fluoride
(PMSF) (> 99.0%). All of the chemicals except for ferritin and apoferritin were
used as obtained without further treatment.
32


The purification of commercially available ferritin utilized a Flex-
Column gravity liquid chromatography column purchased from Kontes Glass
Co.(Vineland, NJ). A Cypress Model Omni 90 potentiostat (Lawrence, KS)
and a BioAnalytical Systems Model R XY recorder (West Lafayette, IN) were
used to perform voltammetric scans. A Perkin-Elmer (Nonwalk, CT) Model
552 UV-Visible Spectrophotometer was used to determine the concentration
of ferritin following size-exclusion chromatography.
2.2 Prevention of contamination
The contamination resulting from non-measurable surface active
species coming from water, electrochemical components, all grasswares, etc.
can make a big difference on the result of the electrochemical experiments.
Prevention of the contamination was proceeded as following: Water was
purified using a Milli-Q water treatment system manufactured by Millipore
Corporation (Bedford, MA). The Mili-Q system treated water was purified
again by distilling de-ionized water vapor through a heated platinum catalyst
in the presence of oxygen, subsequently distilled again. The water is called
pyrolytically distilled water (PDW). PDW was used in cleaning procedures, in
the preparation of all solution and in all experimental procedures. All
glassware and electrochemical cells were cleaned by soaking in chromic acid
33


for at least 20 min. prior to use in all experiments followed by rinsing with
PDW. The chromic acid solution was prepared by dissolving 92 grams of
sodium dichromate (Na2Cr204*H20) into 458 ml of water followed by the
addition of 800 ml of sulfuric acid while stirring.
2.3 Ferritin (or apoferritin) purification
A gravity column (25 cm X 2.5 cm) packed with a gel filtration medium
of sephadex G-200 (protein fraction range, 5,000-600,000 Da) and G-200
buffer (20 mM pH 7.0 phosphate buffer, 0.9% NaCI, 0.2 mM phenylmethyl-
sulfonyl fluoride (PMSF), and 0.05% sodium azide (NaN3)) was used to purify
the ferritin protein. The sephadex G-200 solution was prepared by hydrating
4g of G-200 sephadex gel in 300 ml of G-200 buffer for 24 hrs. One third of
the gravity column was packed with G-200 sephadex gel and the G-200
buffer flushed the packed G-200 sephadex gel through the column under
pressure of N2. A 3 ml sample of horse spleen ferritin (or apoferritin) was
applied to the column and eluted with G-200 buffer. The eluted fractions were
collected in 1ml increments and the ferritin fractions were combined.
34


2.4 Concentration determination of ferritin
(or apoferritin)
The concentration of the ferritin sample was determined by the
(
Bradford method [42] as follows; A series of standard solutions were
prepared by adding 1 ml of different concentrations (1ppm ~ 5ppm) of Bovine
Serum Albumin (BSA) into 5 ml of protein assay dye G-250. 1ml of the
ferritin sample was also treated with 5ml of the protein assay dye G-250, and
the absorbance of each standard of blue complexes (standards and sample)
was measured at 595 nm using a Perkin-Elmer (Nonwalk, CN) Model 552 UV-
visible spectrometer. The absorbance of ferritin sample was projected onto
the calibration curve to find the ferritin sample concentration.
The average number of iron atoms per ferritin molecule was
determined by the UV-visible spectrometer to determine the concentration of
ferritin as described above and a Perkin-Elmer (Norwalk, CN) model 5000
atomic absorption spectrophotometer to determine the number of iron atoms
per liter. In order to determine the number of Fe atoms per liter, the iron
standards were prepared in a range of 5.0 to 10.0 ppm using ferric
ammonium sulfate in 1M nitric acid. The measurements were conducted at
248.3 nm operating at a hollow cathodic lamp current of 30 mA.
The number of iron atoms = # moles Fe/L_______
per ferritin molecule # moles ferritin/L
35


2.5 Electrochemical equipments
2.5.1 Electrochemical cell (H-cell)
Electrochemical cells having the H-type configuration were used in all
cyclic voltammetry experiments (figure 2.1). The H-cell contains two
chambers separated by a 1 cm sintered glass fritted disc that allowed ions to
flow between the compartments while keeping contaminants from the side
compartment from entering the main chamber. The side chamber contained
the reference (Ag / AgCI) and auxiliary (platinum wire) electrodes, while the
main chamber contained the working electrode (tin doped indium oxide (ITO))
immersed in the electrochemical solution under study. The nitrogen inlet
allowed for the deaeration of the solutions. Nitrogen flowed through the
bottom of the main chamber, and exited through the outlet located at a
position above the solution surface. All potentials are given with respect to
the Ag/AgCI reference electrode. To provide the oxygen free environment
when an electrode was transferred between different cells, a tube was used
to suspend the working electrode. A Teflon stopcock in the nitrogen line
controlled the flow of pressurized nitrogen (figure 2.2).
36


N2 inlet
Figure 2.1 Electrochemical cell configuration (H-cell).
Auxiliary electrode: Platinum wire
Reference electrode: Ag/AgCI (in 1M KCI) = 0.222V
Working electrode: Tin doped ln203 on a glass
37


A. Teflon stopcock
B. 10 mm glass tubing
C. stopper
D. electrode lead
E. alligator clip
F. 6x10 mm ITO electrode
Figure 2.2 Glass-sheathed ITO electrode.
38


2.5.2 Preparation of tin doped indium oxide electrodes
Tin-doped indium oxide (ln203) (100 pm) on glass was donated by
Applied Films Corporation (Boulder, CO). Tin-doped indium oxide on glass
was cut to 6 mm X 10 mm pieces, cleaned by sonication in saturated Alconox
(in 95% ethanol) for ten minutes, rinsed with Milli-Q water, then sonicated
twice for ten minutes in Milli-Q water. Finally, the electrodes were allowed to
hydrate for 24 hours. In the experiments, the electrode was clipped to an
alligator type connector and suspended in solution by an electrical lead wire
so that only a 6 mm X 6 mm area was immersed in solution (figure 2.2).
2.5.3 Preparation of reference and auxiliary electrodes
The reference electrode was Ag/AgCI in 1 M KCI. A silver wire was
coated with silver chloride by oxidizing a silver wire in 6M HCI. The reaction
is shown below:
AgCI (s) + e' (metal) - Ag (s) + Cl' (aq)
The silver chloride coated silver wire was placed in a 6 mm diameter glass
cylinder capped at the bottom with a ground glass cap. The tube was then
filled with 1 M KCI solution. The auxiliary electrode was a platinum electrode
coiled around the reference electrode. All electrodes (working, reference,
39


and auxiliary electrodes) were connected to the potentiostat. The data were
recorded on the X/Y recorder (figure 2.3).
2.6 Experimental procedure
The pH 7.0 phosphate buffer (p=1.0 M) was prepared by mixing 50 ml
of 0.1 M sodium dihydrogen phosphate with 29.1ml of 0.1 M sodium hydroxide
then diluted to 1L with PDW. The pH 7 HEPES buffer was prepared by
mixing 15.9 g HEPES, 0.49 g of NaOH, and 5.84 g NaCI, then diluted to 100
ml with PDW. The 1x1 O'4 M ferrous ion solution was prepared by dissolving
0.001 g of ferrous ammonium sulfate into 25ml of pH 7.0 M HEPES buffer
which was previously deaerated with N2 for 20 minutes .
2.6.1 The preparation of ITO electrodes with
a layer of adsorbed ferritin (ITO/Ferritin electrodes)
In this study, the concentration of ferritin that used for adsorption onto
the ITO electrode surface was 0.1 mg/ml. A ferritin sample (C=0.1 mg/ml)
was prepared by diluting purified ferritin with pH 7.0 phosphate buffer (p=1.0
M). The hydrated ITO electrodes were immersed into the ferritin sample at
open circuit potential for at least 60 hours. After 60 hours, an ITO electrode
40


Figure 2.3 Overall configuration of electrochemical equipments.
41


was removed, rinsed by PDW, and then immersed into an electrochemical
cell containing pure pH 7.0 phosphate buffer (ji=1.0 M) for scanning.
2.6.2 CV of the an ITO/ferritin electrode in phosphate buffer
A general cyclic voltammagram of an ITO electrode with the adsorbed
layer of ferritin was scanned negatively between 0.10 V and -0.80 V at scan
rate of 100 mV/s.
The cyclic voltammetry was also performed at various scan rates while
other condition were held constant in order to observe the scan rate
dependence on ITO/ferritin electrode. The dependence of ferritin packing
density on exposure time was done by varying the immersion times of the
ITO electrode into the ferritin solution at open circuit potential, then the scans
were performed as described above.
In order to study the dependence of packing density on dissolved
ferritin concentration, ITO electrodes were exposed to various concentrations
of the ferritin solution at open circuit potential for least 60 hours.
Pure phosphate buffer at different pH was used to dilute the purified
ferritin in order to examine the effect of different pH on adsorbed ferritins
packing density. The phosphate buffers of different pH were prepared by
42


varying the concentration of sodium hydroxide while maintaining a constant
ionic strength of 1.0 M.
2.6.3 The removal of iron from horse spleen ferritin
(the formation of apoferritin)
Modification of ITO electrodes with an adsorbed layer of ferritin is
described above. Followed the rinsing of dissolved ferritin, the ITO/ferritin
electrode was immersed into a H-cell that contained phosphate buffer
containing 10 mM EDTA. The potential was scanned negatively from 0.10 V
to -0.80 V at a scan rate of 100mV/s in order to reduce adsorbed ferritin.
In order to determine whether iron release was induced by the
reduction of ferritin iron or simply caused by the direct chelation of Fe(lll), the
following procedure was used: An ITO/ferritin electrode was immersed in 50
mM EDTA/phosphate solution for 6 hours. The ITO/ferritin electrode was
removed, rinsed with PDW, and immersed into a cell containing pure
phosphate buffer. A cyclic voltammagram of the electrode was cycled
between 0.10 V and -0.80 V. The iron release dependence on buffer anion
was studied by cycling the potential of the electrode in 10 mm EDTA/pH 7.0
HEPES buffer between 0.20 V and -1.00 V.
43


2.6.4 Iron uptake by horse spleen apoferritin
The iron uptake experiment proceeded in four steps as following; 1)
the preparation of ITO electrodes with an adsorbed layer of ferritin, 2) the
removal of iron, 3) exposure to Fe2+ at 0.20 V, and 4) scanning of ITO/ferritin
electrode. The exposure of an ITOfapoferritin electrode to ferrous ion was
done in pH 7.0 HEPES buffer instead of using pH 7.0 phosphate buffer (p =
1.0 M) because HEPES is a non-iron binding buffer while phosphate readily
precipitates Fe2+. When the ITO/ferritin electrode was transferred between H-
cell, the electrode was blanketed by pressurized N2 to provide an oxygen free
environment.
After removing the iron from adsorbed ferritin, an
ITO/electrochemically emptied apoferritin electrode was rinsed with pure
electrolyte and then re-immersed in pure HEPES buffer containing 1x10^ M
ferrous ammonium sulfate for 20 minutes. Immersion was carried out at 0.20
V, the potential which can sustain the adsorbed ferritin in the oxidized from.
The electrode was removed, rinsed, and re-immersed under N2 into a cell
containing pure HEPES buffer. The potential cycled from 0.20 V and -0.70 V
at a scan rate of 100 mV/s.
The ITOfapoferritin electrode was exposed to buffer at immersion
potentials of-0.40, -0.20, 0.00, 0.10, 0.20, 0.30, and 0.50 V. The effect of
44


the oxidation by molecular oxygen on the reconstitution of adsorbed ferritin
was investigated. The ITOf apoferritin electrode was immersed into the
Fe2+/pure HEPES buffer at open circuit potential for 20 minutes, followed by
rinsing and transferring to a cell containing pure HEPES buffer that contained
saturated molecular oxygen for 20 min. Then the electrode was transferred
to another cell containing pure HEPES buffer under nitrogen. The current-
potential curves were cycled between 0.20 V and -0.70 V at 100 mV/s.
In order to compare the apoferritin that was made by removing iron
from ferritin electrochemically with commercially available apoferritin, the
following controlled experiment proceeded: The commercially obtained
apoferritin was purified and diluted with pH 7.0 phosphate buffer (ji = 1.0 M)
to give a 0.1mg/ml solution of apoferritin. The ITO electrodes were then
immersed into the apoferritin sample at open circuit for at least 60 hours. An
ITO/apoferritin electrode was removed, rinsed, and immersed into Fe2+/pure
HEPES buffer at potential 0.20 V for 20 minutes. This was followed by rinsing
and transferring under N2 to a cell containing pure HEPES buffer, and the
potential cycled between 0.20 V and -0.70 V.
45


2.7 Calculation of surface concentration
The area underneath the cathodic peak is proportional to the amount
of charge passed when adsorbed ferritin is electrolyzed. The baseline of the
anodic peak was estimated by drawing a straight line from the leading and
tailing limits of the peak. The area of the anodic peak was integrated by the
method of cutting and weighing. Experimental packing densities r were
determined by integrating the charge of the initial cathodic peak of adsorbed
ferritins current-potential curve, followed by substitution in the Faraday law
shown below;
r = q
nFA
T is the packing density (mole/cm2), Q is the integrated charge of the anodic
peak (coulombs), A is the geometric electrode area in cm2, n is the number of
electrons being transferred during the redox reaction, and F is the Farday
constant (96,487 C mol"1). The electrode surface area occupied by each
ferritin molecule was calculated from the reported radius of the protein
(12.0 nm). The theoretical packing density of ferritin was estimated by from
the geometric electrode area and the calculated footprint area of ferritin.
46


3.
Results and discussions
3.1 Characterization of the ITO/ferritin electrode
Using the reported diameter of tissue ferritin of 12.0 nm, and given that
the quaternary structure of the ferritin is nearly spherical, the theoretical area
occupied per ferritin molecule is calculated to be 1.1 x 10'12 cm2. Using this
estimate, a theoretical packing density for the ferritin is calculated to be 1.5
pmol/cm2. An experimental packing density of 1.3 pmol/cm2 is determined
from the charge represented by initial cathodic peak, the Faraday law, an
electrode area of 0.36 cm2, and the number of iron atoms per ferritin
molecule, which is approximately 1,500. Therefore, the close agreement
between the experimental and theoretical packing densities indicates that
roughly a monolayer of ferritin is formed onto ITO electrode.
The pi of horse spleen ferritin is 4.5 [2, 4], while that of indium oxide is
about 6 [48]. Hence, at neutral pH the ionic interaction between the ferritin
and the electrode surface is not expected to be as strong as that between a
protein such as cytochrome c and the ITO electrode. However, it has been
reported that the packing density of the ferritin increases with ionic strength,
suggesting that the adsorption of the ferritin to the ITO electrode probably
involves a hydrophobic interaction [49].
47


3.2 CV of an ITO/ferritin electrode in phosphate buffer
A clean tin-doped indium oxide electrode (ITO) were immersed in
ferritin solution for 60 hours, followed by rinsing away dissolved ferritin and
immersing into a H-cell containing pure pH 7.0 (p=1.0 M) phosphate buffer.
When the potential were cycled between 0.10 V and -0.80 V, the
current-potential curve (figure 3.1) shows three major features; an initial
cathodic peak at -0.62 V, an anodic peak at -0.19 V in the return scan, and a
new cathodic peak at -0.36 V with the absence of initial cathodic peak at
-0.62 V in the second cycle. The initial cathodic peak is the reduction of
Fe(lll) in the adsorbed ferritin layer. The anodic peak of the return scan and
a new cathodic peak of second cycle indicate the oxidation of Fe(ll) and re-
reduction of Fe(lll) of the layer, respectively.
The area of the anodic peak and new cathodic peak are smaller than
the initial cathodic peak. Since the area under a peak represents a charge
passed due to a reaction, and the charge is proportional to the number of
moles of ferritin react, based on the Faraday law (Q = nFAT), this result
suggests that some desorption of ferritin might have resulted from the
reduction of the initial layer.
48


0--
Figure 3.1
-0.8 -0.6 -0.4 -0.2 0.0 0.2
POTENTIAL (VOLTS vs. Ag/AgCI)
Current-potential curve of ferritin adsorbed at an ITO electrode
in pH 7.0, ji=1.0 M phosphate buffer,
ferritin concentration: 0.1 mg/ml, electrode area: 0.36 cm2,
scan rate: 100 mV/s.
49


In addition, in the second cycle, a new cathodic peak appears at more
positive potential, while the initial cathodic peak disappears. These results
suggest that major changes have occurred in the layer upon reduction. Upon
reduction, ferritin in the initial layer became kinetically and thermodynamically
different ferritin. On the return scan of second cycle, the anodic peak
potential has not changed. Repeated cycles reveal the presence of only the
anodic and new cathodic peaks, however decreasing in size. These results
suggest that the initial layer appears to reconstruct into a new layer which is
electro-inactive. The anodic and new cathodic peaks appear to be
associated with each other, belonging to the same redox couple of the new
layer.
The thermodynamics of the electron transfer reaction are different
since the mid-point potential of the redox couple is a more positive value.
The initial layer gives a cathodic peak with no return anodic peak, indicative
of a kinetically slow redox system. The AEP between new constructed
cathodic and anodic peaks is -0.17 V. The couple exhibits quasi-reversible
behavior indicating a system with faster kinetics than the original layer. It is
not clear what the reconstruction of entails. Some possibilities are 1)
molecular re-orientation, 2) changes in the composition of the iron core, or
3) a change in the rigidity of the protein coat and/or the core.
50


Though it is not clear what might be responsible for the reconstruction
of the initial layer upon its reduction, conformational changes in the global
protein occurring with changes to the metallic centers are known for other
proteins. It has been shown that upon reduction, the heme of cytochrome c
moves into the interior of the protein causing the heme crevice to close. The
conformational change of cytochrome c accompanying its reduction results in
a change in the interaction of the protein with the electrode surface [50].
3.2.1 The scan rate dependence of the ITO/ferritin electrode
In order to confirm whether the current-potential behavior is due to ferritin in
an adsorbed state, the current-potential curve of ITO /ferritin electrodes was
obtained between 0.20 V and -0.70 V at scan rates of 50, 100, 200, and 400
mV/s. Each current-potential curve shows basically the same features, a
cathodic peak, anodic, peak, and new constructed cathodic peak. According
to the equation (ip = (n2F2/4RT)vA rox*), for the adsorbed state, the cathodic
peak currents increase correspondingly with scan rate. As expected for the
adsorbed species, the currents of all initial cathodic peaks increase linearly
with scan rate (correlation coefficient = 0.999) shown in figure 3.2. Therefore,
ferritin is adsorbed onto the ITO electrodes. The AEp of the reconstructed
layer increases gradually over the same scan rate range, indicating
51


25
0 100 200 300 400 500
SCAN RATE (mV/sec)
Figure 3.2 The scan rate dependence of ferritin adsorbed
at an ITO electrode, (peak current vs. scan rate)
ferritin concentration: 0.1 mg/ml & electrode area: 0.36 cm2.
52


an electron transfer couple exhibiting fairly slow kinetics (figure 3.3).
3.2.2 The effects on packing density of
the ITO/ferritin electrode
In order to determine which factors affect packing density, clean ITO
electrodes were immersed in the ferritin solution at various adsorption time,
concentration, and ionic strength of the ferritin solution.
As the time of exposure to the ferritin increases from 0.25 to 96 hours,
the packing density of ferritin onto the ITO electrode increases slowly, and
requires 60 hours to reach a limiting value (figure 3.4). This time dependent
adsorption behavior suggests that the adsorption kinetics of ferritin is such
that the adsorption-desorption equilibrium takes place over longer periods of
time compared to ferritin adsorbed onto bare gold or modified gold surfaces.
In general, when long periods of time are required, re-orientation or
conformational changes accompanies surface attachment [42]. Perhaps
since both ferritin and the ITO electrode possess charged surfaces, it
probably takes considerable time for the adsorbed molecules to arrange
themselves in the most efficient packing. Figure 3.5 shows the dependence
of packing density of dissolved adsorbate concentration of ferritin. As
dissolved ferritin concentration increases from 0.0125 to 0.1 mg/ml,
53


POTENTIAL (VOLTS vs. Ag/AgCI)
SCAN RATE (mV/sec)
0 100 200 300 400 500
Figure 3.3 The scan rate dependence of ferritin adsorbed at an ITO
electrode (AEp of the reconstructed layer vs. scan rate)
ferritin concentration: 0.1 mg/ml & electrode area: 0.36 cm2.
54


REL. INITIAL CATHODIC PEAK AREA
ADSORPTION TIME (hours)
Figure 3.4 The adsorption time dependence of ferritin adsorbed
at an ITO electrode.
(Rel. initial cathodic peak area vs. adsorption time)
ferritin concentration: 0.1 mg/ml, scan rate: 100 mV/s
electrode area: 0.36 cm2.
55


REL. CATHODIC PEAK AREA
CONCENTRATION (mg/ml)
Figure 3.5 The dissolved ferritin concentration dependence of ferritin
adsorbed at an ITO electrode.
(Rel. initial cathodic peak area vs. concentration)
electrode area: 0.36 cm2 pH: 7.0, scan rate: 100 mV/s.
56


the packing density of ferritin adsorbed at open circuit potential, also
increases. The packing density depends on the equilibrium between species
adsorbed and species left in solution. It also appears that, the interaction
with the electrode surface has not adversely affected the electroactivity of
ferritin over time.
The packing density of ferritin increases with ionic strength, suggesting
that the adsorption of ferritin to ITO probably involve a hydrophobic
interaction. It has been reported previously that at lower ionic strengths, the
reduction process is represented by two peaks, rather than one. It indicates
that ferritin appears to adsorb in two different states, suggesting that while
both states involve a hydrophobic interaction, one state perhaps has greater
component of charge repulsion than the other [49].
3.2.3 The pH dependence of the ITO/ferritin electrode
When the potential was scanned in different pH (ji = 1.0 M) phosphate
buffers, all three peaks occurred in the current-potential curve. However, the
pH has an effect on the reduction potential, but not in the current peak area.
As the pH increases the peak potential becomes more negative, shifting by
-125 mV/pH (figure 3.6), suggesting that the reduction is thermodynamically
favored at lower pH. This result suggest that 2H+ are transferred to
57


POTENTIAL (VOLTS vs. Ag/AgCI)
4.5
5.5
pH
6.5
7.5
8.5
Figure 3.6 The pH dependence of ferritin adsorbed at an ITO electrode.
(pH vs. initial cathodic peak potential)
ferritin concentration: 0.1 mg/ml, electrode area: 0.36 cm2,
scan rate: 100 mV/s.
58


the core for every Fe(lll) reduced. The result is also supported by the
Nemest equation that the potential is shifted by -120 mV per pH changes if
the reduction involves the uptake of two protons and one electron.
If [ox] + 2H+ + e' [red]
E = E 0.0592 log [red] [H]~2 => E = E -120 mV log [red] [H]'1
1 [ox] [ox]
In addition the result suggests that adsorbed ferritin behaves similarly to
ferritin in solution in that two protons are absorbed for each electron
transferred [21]. The reduction of the iron in ferritin is pH dependent,
becoming thermodynamically more difficult with increasing pH. Interestingly,
the mid-potential between anodic and new cathodic peaks are shifted only by
-15 mV/s, indicative of a pH independent redox reaction. This result suggests
that there is no longer proton transfer involved in redox reaction on the new
constructed layer, suggesting that a composition changes in the core.
3.2.4 CV of an ITO/ferritin electrode in HEPES buffer
Figure 3.7 shows the current-potential curve of an ITO/ferritin
electrode in pH 7.0 (ji=1 .5 M) HEPES buffer. The general features of the
scan in the phosphate buffer can be found in this curve. However, the initial
cathodic peak appears broader in HEPES, and its potential is shifted from
59


POTENTIAL (VOLTS vs. Ag/AgCI)
Figure 3.7 Current-potential curve of ferritin adsorbed at an ITO electrode
in pH 7.0, p = 1.5 M HEPES buffer,
ferritin concentration: 0.1 mg/ml, electrode area: 0.36 cm2,
scan rate: 100 mV/s.
60


-0.62 V (in phosphate) to -0.67 V (in HEPES). The quasi-reversible couple
has undergone changes in potential as well; the midpoint potential shifting
from -0.275 V to -0.235 V. In addition, the peak potential difference has
increased from 170 mV to 220 mV. It is known that phosphate is binds to the
lysine residues in the vicinity of cytochrome cs heme edge, stabilizing
ferricytochrome c. The effect of the stabilization is to shift the reduction
potential negative relative to cytochrome c in non-binding electrolytes [41].
The binding ability of HEPES is considerably less significant than that of
phosphate. Fe3(P04)2 -> 3Fe2+ + 2PO43* = 1035
As for cytochrome c, phosphate can also bind to the positively charged
residues on ferritins surfaces. There is much evidence, which suggests that
the mineral core-apoferritin interface is important in ferritins function [2].
Agents affecting the surface of ferritin may also affect the energetics of the
electroactive mineral core.
3.3 The removal of iron from horse spleen ferritin
It is known that iron can be induced to exit the iron core following the
reduction of ferritin iron. Since the adsorbed ferritin is electroactive at the ITO
electrode, it is of interest to see whether electrochemical reduction can
induce iron release from ferritin in the adsorbed state.
61


3.3.1 Iron release by an electrochemical reduction
in the presence of EDTA
The removal of iron was performed by an ITO electrode with a layer of
adsorbed ferritin immersed in pH 7.0 (p=1.0 M) phosphate buffer containing
10 mM EDTA. The potential was scanned between -0.70 V and 0.20 V at a
scan rate of 100 mV/s. The current-potential curve in figure 3.8 (a) shows the
initial cathodic peak at -0.58 V and the absence of anodic peak and cathodic
peak in the return wave and in the second cycle respectively. This suggests
that when the potential was negatively scanned, the Fe(lll) in ferritin was
electrochemically reduced to Fe(ll), followed by the complexation and thus
the removal of ferrous ion by the surrounding EDTA, a strong complexing
agent for many metal ions. Switching the scan direction to more positive
potentials reveals the absence of the anodic peak. An explanation of this
result is that the anodic peak is absent because iron has been captured by
EDTA and is no longer present in the ferritin core. Moreover, the missing
cathodic peak in the second cycle strongly supports the absence of iron in
the ferritin core. In this experiment, reducing agents were not used, though
reducing agents are effective in removing iron from ferritin. Instead, electrons
are directly transferred from the electrode material to the protein core by
using the electrochemical method. These data suggest that either electrons
62


pass through the protein coat by a tunneling mechanism, or that the electron
transfer is mediated by iron atoms residing in the channels.
3.3.2 Iron release without the reduction
in the presence of EDTA
It is important to determine whether iron release is induced by the
electrochemical reduction of iron in ferritin or simply caused by a direct
chelation of Fe(lll). An ITO/ferritin electrode was immersed in pH 7.0 (ji=1.0
M) phosphate buffer containing 50mM EDTA for 6 hours, providing ample
time for EDTA to remove iron from ferritin. The electrode was then rinsed
free of residual EDTA, and immersed in pH7.0 (p=1.0 M) phosphate buffer.
The potential was scanned between -0.80 V and 0.20 V at a scan rate of 100
mV/s. The current-potential curve in figure 3.8 (b) reveals three major
features again; an initial cathodic peak at -0.62 V, an anodic peak at -0.19 V
in the return scan, and a new cathodic peak at -0.36 V in the second cycle.
These three peaks occur at the same peak potentials that in the scan of
ITO/ferritin electrode in pH 7.0 phosphate buffer (p=1.0 M). The presence of
the familiar features indicates that iron (III) in the core is not complexed
directly by EDTA over 6 hours. Rather, iron can be only mobilized after it is
first reduced (figure 3.9).
63


Figure 3.8 Current-potential curves of ferritin adsorbed at an ITO electrode
(a) with the applied reducing potentials in the presence of 10 mM
EDTA. (b) after immersing in 50 mM EDTA solution at OCP.
ferritin concentration: 0.1 mg/ml, electrode area: 0.36 cm2,
scan rate: 100 mV/s.
64


\\\w\\
EDTA
\ EDTA
Fe(III) )
^ y EDTA
EDTA
"Apoferritin"
Figure 3.9 Schematic representation of iron release of ferritin adsorbed
at an ITO electrode.
65


3.3.3 Iron release in HEPES buffer
It was fond that ferritin does not adsorb onto the ITO electrodes from
HEPES buffer, however ferritin does adsorb from phosphate buffer. The
scan in HEPES buffer shows a smaller cathodic current, and peak potentials
occur at different values as compared to the scan in phosphate buffer (figure
3.10). The current-potential curve of the ITO/ferritin electrode in EDTA/pH
7.0 HEPES buffer shows that the anodic and cathodic peaks of the
reconstructed layer are still present, though small, whereas they are
completely absent in the scan in phosphate. Though in each buffer the curve
shows that the reduction of ferritin iron has been effected, there remains
some unreacted Fe(ll) in the core when the experiment is done in HEPES
buffer. This result suggests that the driving force for the formation of the
Fe2+-EDTA complex is smaller in HEPES buffer than in phosphate buffer.
The attenuating effect on the driving force may be due to the limited access
of EDTA to the ferritin layer due to the presence of the large HEPES
molecule, coupled with the decreased ability of HEPES to bind to iron, as
compared with phosphate.
66


POTENTIAL (VOLTS vs. Ag/AgCI)
Figure 3.10 Current-potential curve of ferritin adsorbed at an ITO electrode
in pH 7.0, n = 1.5M HEPES buffer containing 10 mM EDTA.
ferritin concentration: 0.1 mg/ml, electrode area: 0.36 cm2,
scan rate: 100 mV/s.
67


3.4 Iron uptake by horse spleen apoferritin
In vitro studies have shown that ferritin can be reconstituted from
apoferritin in the presence of Fe(ll), and not Fe(IH) [23].
3.4.1 Iron uptake by the immersion potential
of electrode at 0.20 V
An ITO electrochemically emptied apoferritin electrode was placed in
HEPES buffer containing 1x1 O'4 M ferrous ammonium sulfate for 20 min. at
0.20 V. The uptake of iron is increased with increasing exposure time into
Fe2+/pure HEPES buffer, until a limited integrated charge is reached at 20
minutes. The controlled potential of 0.20 V is negative enough to maintain
dissolved iron in its reduced form, while sustaining ferritin iron in its oxidized
from. For the Fe3+/Fe2+ couple which in solution:
Fe3+ + e- - Fe2+ E = 0.55 V (vs. Ag/AgCI)
The electrochemical system is kept in an oxygen free environment by
deaerating the HEPES buffer with nitrogen though the bottom of the cell.
Anaerobic conditions are essential in order to eliminate the possibility of
oxidation of ferritin by dissolved oxygen.
An ITOfapoferritin electrode was immersed in pure HEPES buffer,
and the potential cycled between 0.20 V and -0.70 V. The current potential
68


curve exhibits the initial cathodic wave of the initial layer and the quasi-
reversible anodic/cathodic couple of the re-constructed layer (figure 3.11).
The appearance of the familiar peak pattern peaks suggests that iron has
entered the protein shell. In the absence of molecular oxygen, Fe(ll) is still
converted to Fe(lll) by an applied electrode potential (0.20 V), suggesting
that during, or after entry into the protein sphere, the iron is oxidized.
Moreover, given that the peak potentials are similar to those of adsorbed
ferritin, it is probable that the iron oxidation leads to the ferrihydrite structure
that has been found with reconstituted ferritin in other studies. However, the
peak potentials are not exactly the same in figure 3.1; the initial cathodic
peak has shifted from -0.67 V to -0.46 V; the midpoint potential of the
quasi-reversible couple from -0.235 V to -0.230 V. Also, the peak potential
difference has changed from 220 mV to 280 mV. The more negative peak
potentials in natural ferritin may be due to the presence of phosphate in its
core. When ferritin is emptied of its iron, phosphate is also released [19].
Since phosphate is largely adventitious, ferritin can be reconstituted without
phosphate [24]. In addition, it was found through Mossbauer spectroscopy,
that the iron core of reconstituted ferritin was indeed that of the ferrihydrite
structure, and that the structure was more ordered than that of native ferritin
[52]. It should be mentioned that phosphate is no longer in the iron core
69


when ferritin is reconstituted electrochemically, thus the absence of
phosphate may be responsible for more order in structure and the lower
redox potentials. These results are further supported by the observation that
the reduction potential of bacterioferritin from Azobacter vinlandii is lower
than that of mammalian ferritin, presumably due to the higher concentration
of phosphate in the latter [51]. It can be noticed that the cathodic peak area
of the i-E curve in Figure 3.11 is three times larger in size than those shown
in figure 3.1 (i-E curve of adsorbed, natural ferritin). In the original layer of
ferritin, it was found that 1,500 Fe/molecule presented. Since it was
increased by three times larger than the original amount of irons, 4,500
Fe/molecule would be present, which agrees with the maximum number of
iron atoms contained by ferritin. This increase is probably due to a greater
number of iron atoms in the reconstituted ferritin than was present originally.
This result is also supported by vitro studies which have reported that
reconstituted ferritin contained three times the iron as native ferritin [52]. In
general, the current potential curves of re-constituted ferritin and natural
ferritin are more similar than they are different.
Iron uptake results in all of the adsorbed ferritin becoming electroactive
again as indicated by the reappearance of the initial cathodic peak. It has
been previously reported that the reduction induced reconstruction of natural
70


-0.8 -0.6 -0.4 -0.2 0.0 0.2
POTENTIAL (VOLTS vs. Ag/AgCI)
Figure 3.11 Current-potential curve of an ITO electrode with an adsorbed
layer of reconstituted ferritin electrode following
by the uptake of iron at 0.20 V.
Fe2+/HEPES buffer concentration: 1 x 10-4 M, pH: 7.0,
electrode area: 0.36 cm2, scan rate: 100 mV/s.
71


ferritin may be accompanied by desorption [49]. However, these new data on
iron uptake indicate that ferritin which was thought to have desorbed have
probably undergone a re-orientation, conformational change, or some other
reconstruction rendering it electroinactive.
3.4.2 The immersion potential of electrode dependence
In order to better define the role of electrode potential on the uptake of
iron, the immersion of ITOfapoferritin electrodes in Fe2+/pure HEPES buffer
is made for the same duration, but immersing the ITOfapoferritin electrodes
at different oxidizing potentials. Figure 3.12 shows the initial cathodic peak
area plotted against electrode potential under which ferritin was reconstituted.
The immersing potential above 0.10 V do not increase the level of iron
loading, indicating that this potential was sufficiently positive to oxidize iron
entering the protein shell to Fe (III). At -0.20 V ferritin is maintained in its
reduced form according to cyclic voltammetry. The experiment is repeated at
-0.20 V and -0.40 V, however, no peaks were observed in the ensuing scan.
This suggests two possibilities that at these negative potentials iron had
either not entered the protein, or entered but has not been oxidized. In order
to show whether iron enters the apoferritin shell at -0.20 V, the
ITOfapoferritin electrode was exposed to Fe2+/pure HEPES buffer
72


REL. CATHODIC PEAK AREA
POTENTIAL (VOLTS vs. Ag/AgCl)
Figure 3.12 The immersion potential of electrode dependence of iron uptake.
(Rel. initial cathodic peak area vs. potential)
Fe2+/HEPES buffer concentration: 1 x 10^ M, pH: 7.0,
electrode area: 0.36 cm2, scan rate: 100 mV/s.
73


at -0.20 V for 20 minutes. Following a rinse under N2i the ITO/"apoferritin
electrode was immersed in pure HEPES buffer for 20 minutes at 0.20 V to
oxidize any Fe(ll) which may have entered apoferritin shell. However, no
appreciable peaks were displayed on the current-potential curve indicating
that iron does not enter the apoferritin
sphere at -0.20 V. This result suggests that iron does not enter the
apoferritin shell at potentials more negative than the potential -0.20 V.
3.4.3 Iron uptake by molecular oxygen
Most in vitro iron uptake studies have subjected apoferritin to ferrous
ions in the presence of dissolved oxygen. In order to show whether ferritin
controlled by the electrochemical potential behaves similarly to that in other in
vitro studies, the applied oxidizing potential was replaced by an exposure to
dissolved oxygen. An ITO/apoferritin" electrode was prepared as described
above, then immersed into deaerated ferrous ammonium sulfate in HEPES
buffer at open circuit potential. The electrode was rinsed under nitrogen to
provide an oxygen free environment, and immersed in pure HEPES buffer
saturated with oxygen for 20 min. Ideally, the exposure to oxygen should be
one simultaneously with ferrous ion exposure, however, the oxidation of a
74


ferrous ion is observed when this condition was attempted as following
reaction;
Fe2+ + 02 Fe3+ + H20 -> Fe(OH)3 (s)
Therefore, the exposure to oxygen is made after the exposure to ferrous ions.
After rinsing under N2 and immersion in pure HEPES buffer, the potential is
cycled between 0.20 V and -0.70 V. The current-potential curve, shown in
figure 3.13 (a) displays similar electroactivity as shown in Figure 3.11,
suggesting that oxygen has a same effect as the oxidizing potential on the
oxidation of iron taken up by ferritin. However, the peak area are ten times
smaller than when the oxidizing potential was applied during the exposure to
iron. Since the immersion of ITOrapoferritin electrode in Fe27pure HEPES
buffer was made at open circuit, which was approximately -0.17 V, the
potential did not allow the limiting amount of iron to enter the shell (see the
potential dependent study in section 3.4.2 & figure 3.12). These
current-potential data also show that oxidation can take place after iron
uptake, suggesting that some Fe (II) can be loaded into the protein. These
results are consistent with EXAFS experiments, which reveal that Fe (II) can
load into ferritin without oxidizing [53].
Though it appears that oxygen is responsible for the oxidation of Fe(ll),
it was necessary to determine whether the application of 0.20 V at the start of
75


the scan had an effect on the oxidation of Fe(ll). The ITOfapoferritin
electrode was immersed in deaerated Fe2+/HPEES buffer at open circuit,
/
rinsed under N2, and immersed in pure HEPES buffer. The electrode was not
exposure to molecular oxygen at this time. The resulting current-potential
curve is shown in Figure 3.13 (b). The i-E curve shows essentially no current
indicating that oxidation was not effected by the applied potential (0.20 V)
during the scan. In addition, iron remains in reduced form by omitting the
exposure to oxygen, supporting the conclusion that oxygen effects the
oxidation of iron (II) already present in ferritin.
It was of interest to determine whether iron can be oxidized by an
applied oxidizing potential after reconstitution at open circuit potential.
Following immersion of an ITOfapoferritin electrode at open circuit potential
in Fe2+/HEPES buffer for 20 minutes, the electrode was placed into pure
HEPES buffer at 0.20 V for 20 minutes. The current-potential curve shows
identical currents as when molecular oxygen was used as the oxidizer. This
result suggests that oxidation by oxygen or applied oxidizing potential is
equally effective following the loading of Fe (II).
76


Figure 3.13 (a) Current-potential curve of an ITO with an adsorbed layer of
reconstituted ferritin electrode following iron uptake and
oxidation by dissolved 02 at open circuit potential.
(b) Same as in figure 3.13 (a), but omitting immersion into 02
saturated HEPES buffer.
Fe2+/HEPES buffer concentration: 1 x 10-4 M, pH: 7.0,
electrode area: 0.36 cm2, scan rate: 100 mV/s.
77


3.4.4 Iron uptake by commercially available apoferritin
Presumably, it is apoferritin, which remains on the ITO surface after
the reduction of adsorbed ferritin in the presence of EDTA. However, it is
necessary to ensure whether electrochemically formed "apoferritin" behaves
similar to commercially available apoferritin under same iron uptake condition.
A control experiment was proceeded by the purification and formation of a
layer, from a commercially available authentic apoferritin sample, on an ITO
electrode, as was done for ferritin. The ITO/apoferritin electrode was
immersed in pure HEPES buffer, following by scanning between -0.70 V and
0.20 V. The current-potential curve for the ITO/apoferritin electrode is shown
in Figure 3.14 (a). The current potential curve reveals that no peaks are
exhibited due to the absence of iron in adsorbed apoferritin. The absence of
currents in apoferritin strongly suggests that the voltammetric waves present
in ferritin originate from the oxidation and reduction of iron in the inorganic
ferritin core.
In a related controlled experiment, an ITO/apoferritin electrode was
immersed into Fe27 pure HEPES for 20 minutes at 0.20 V in order to induce
the loading of iron. The electrode was rinsed under N2 again, immersed in
the pure HEPES buffer, and the potential cycled between 0.20 V and -0.70 V,
generating the current-potential curve shown in Figure 3.14 (b). The same
78


peak potentials are displayed as those in Figure 3.11 indicating that the
ferritin, emptied electrochemically, behaves similarly to apoferritin under the
iron uptake process. This result strongly supports the conclusion that ferritin
is converted to apoferritin following the electrochemical reduction in the
presence of a complexing agent. However, the cathodic peak area in figure
3.14 (b) is slightly larger than the cathodic peak area in figure 3.11. The
difference in charge possed might be due to the difference in the adsorption
strength between ferritin and apoferritin.
79


Figure 3.14 (a) Current-potential curve of an ITO with an adsorbed layer of
apoferritin electrode in pH 7.0, p = 1.5 M HEPES buffer.
(b) Current-potential curve of an ITO with an adsorbed layer of
apoferritin electrode following uptake of iron at 0.20 V.
Fe2+/HEPES buffer concentration: 1 x 10^ M,
electrode area: 0.36 cm2, scan rate: 100 mV/s.
80


4.
Conclusion
Horse spleen ferritin irreversibly adsorbs at a tin doped indium oxide
electrode at open circuit potential, and is electroactive in the adsorbed state.
The current-potential curve of an ITO/ferritin electrode in phosphate buffer
indicates an anodic peak and a new cathodic peak with the disappearance of
the initial cathodic peak. These results suggest that reconstruction of the
ferritin layer accompanies reduction of the initially formed. Repeated cycles
reveal the presence of only the anodic and new cathodic peaks, however
decreasing in size. These results suggest that the new layer becomes
electroinacitve with potential cycling, and that the anodic and new cathodic
peaks appear to be associated with each other, belonging to the same redox
couple. The result of the pH dependence study suggests that two protons
are transferred into the iron core upon the one electron reduction of the core
iron (III); however, the redox reactions of the reconstructed layer is pH
independent, suggesting that proton transfer is not involved in the redox
reaction on the new layer, perhaps due to composition change in the core.
Cyclic voltammetry in the presence of EDTA results in the
disappearance of the anodic and new cathodic peaks. This result indicates
that the reduction of ferritin iron in the presence of chelating agents induces
81


the transport of iron from the protein shell of adsorbed ferritin. However, the
iron in adsorbed ferritin in its oxidized form is not readily complexed by
/
chelating agents in solution.
Electrochemically emptied ferritin, when exposed to ferrous ion, takes
up iron at a potential sustaining ferritin in the oxidized form. However, the
iron does not enter the emptied shell at potential more negative than
-0.20 V. Adsorbed apoferritin takes up iron similarly to electrochemically
emptied ferritin, suggesting that apoferritin is formed by the electrochemical
reduction of ferritin in the presence of EDTA. In addition, the oxidizing
potential has the same effect as molecular oxygen for oxidizing iron.
Ferritin adsorbed at an ITO surface exhibits iron uptake and release in
fashions similar to ferritin in solution. Clearly, the results presented in this
report show the ITO/ferritin electrode to be a system by which the primary
functions of ferritin, the loading and unloading of iron, can be examined by
controlling the potential of the electrode. This system thus provides the
means to probe the role of electron transfer in the mechanisms of ferritin's
functions. The detail of ferritin's behavior how it takes up and release the iron
may be studied by surface analytical techniques in future experiments.
82


5.
Bibliography
[1] Gutteridge, J. M. C., Halliwell, B., Treffry, A., Harrison, P. M., & Blake, D.
(1983) Biochem. J. 209: 557-560.
[2] Thiel, E. C. (1987) Ann. Rev. Biochem. 56: 289-315.
[3] Ford, G. C., Harrison, P. M., Rice, D. W., Smith, J. M. A., Treffry, A.,
White, J. L., and Yariv, J. (1984) J. Phil. Trans. R. Soc. London B. 304:
551-565.
[4] Thiel, E. C. in Advances in Inorganic Biochemistry; Thiel, E. C.,
Eichhorn, G. L., and Marzilli, L. G., Eds., (1983) Elsevier, New York. 5:
1-38.
[5] Wardeska, J. G., Vigliorek, B., and Chasteen, N. D; (1986) J. of Biol.
Chem. 1261: 6677-6683.
[6] Harrison, P. M., White, J. L., Smith, A., Farrants, G. W., Ford, G. C.,
Rice, D. W., Addison, J. M., and Treffry, A. (1985) Proteins of Iron
Storage and Transport] Spik, G., Montreuil, J., Crichton, R. R. &
Mazurier, J., Eds., Elsevier Science Publishers, B.V., 67-79.
[7] Heusterspreute, M. and Crichton, R. R. (1981) Fedn. Eur. Biochem.
Soc. Lett. 129: 156-162.
[8] Levi, S., Salfeld, J., Franceschinelli, F., Cozzi, A., Domer, M. H., and
Arosio, P. (1989) Biochem. 28: 5179-5184.
[9] Stefanini, S., Chiancone, E., Arosio, P., Finazzi-Agro, A., and Antonini,
E. (1982) Biochem. 21: 2293-2299.
[10] Pape, L., Multani, J. S., Stitt, C. and Saltman, P. (1968) Biochem. 7:
606.
[11] Jones, T., Spencer, R., and Walsh, C. (1978) Biochem. 17: 4011-4016.
83


[12] Harrison, P. M., Ford, G. C., Rice, D. W., Smith, J. M., Treffry, A. and
White, J. L. (1987) Biochem. Soc. Trans. 15; 744-748.
[13] Sirivech. S., Frieden, E., and Osaki, S. (1974) Biochem. J. 143: 311-
315.
[14] Harrison, P. M., Clegg, G. A., and May, K. Ferritin Structure and
Function. In Iron in Biochemistry and Medicine II, Jacobs, A. and
Worwood, M. eds., (1980) academic press, London. 131-172.
[15] Funk, F., Lenders, J. P., Crichton, R. R., and Schneider, W. (1985) Eur.
J. Biochem. 152: 167-172.
[16] Ulvik, R. J., Romslo, I., Roland, F., and Crichton, R. R. (1981) Biophys.
Acta. 351: 224-229.
[17] Topham, R., Goger, M., Pearce, K., and Schultz, P. (1989) Biochem J.
261: 137-143.
[18] Watt, G. D., Jacobs, D., and Frankel, R. B. (1988) Proc. Natl. Acad. Sci.
85: 7457-7461.
[19] Clegg, G. A., Fitton, J. E., Harrison, P. M., and Treffry, A. (1980) Aprog.
Biophys. Molec. Biol. 36: 56-86.
[20] Watt, G. D., Frankel, R. B., Papaefthymiou, G. C., Spartalian, K., and
Stiefel, E. I. (1986) Biochem. 25: 4330-4336.
[21] Watt, G. D., Frankel, R. B., and Papaefthymiou, G. C. (1985) Biochem.
82: 3640-3643.
[22] Macara, I. G., Hoy, T. G., and Harrison, P. M. (1972) Biochem. J. 135:
151-162.
[23] Harrison, P. M., Treffry A., and Lilley, T. H. (1986) J. of Inorganic
Biochem. 27: 287-293.
[24] Chasteen, N.D. and Xu, B. (1991) J. Biol. Chem. 266: 19965-19979.
84


[25] Hoy, T. G., Harrison, P. M., and Shabbir, M. (1974) Biochem. J. 39:
603-607.
[26] Paques, E. P., Paques, A., and Crichton, R. R. (1979) J.of Molecular
Catalysis 5: 363-375.
[27] Yang, C. Y., Meagher, A., Huynh, B. H., Sayers, D. E., and Thiel, E. C.
(1987) Biochem. 26: 497-503.
[28] Wetz, K. and Crichton, R. R. (1976) Eur. J. Biochem. 61: 545-550.
[29] Macara, I. G., Hoy, T. G., and Harrison, P. M. (1973) Biochem J. 135:
785-789.
[30] Treffry, A. and Harrison, P. M. (1984) J. Inorganic Biochem. 21: 9-20.
[31] Levi, S., Yewdall, S. J., Harrison, P. M., Santambrogio, P., Cozzi, A.,
Rovida, E., Albertini, A., and Arosio, P. (1992) Biochem. J. 288: 591-
596.
[32] Treffery, A., Bauminger, E. R., Hechel, D., Modson, N. W., Nowik, I.,
Yewdall, S. J., and Harrison, P. M. (1993) Biochem. J. 296: 721-728.
[33] Wade, V. J., Levi, S., Arosio, P., Treffry, A., Harrison, P. M., and Mann,
S. J. (1991) Mol. Bio. 221: 1443-1452.
[34] Bauminger, E. R., Harrison, P. M., Mechel, D., Nowik, I., and Treffry, A.
(1991) Biochem. Biophys. Acta. 118: 48-58.
[35] Santambrogio, P., Levi, S., Arosio, P., Palagi, L., Vechio, G., Lawson, D.
M., Yewdall, C. J., Artymiuk, P. J., Harrison, P. M., Jappelli, R., and
Cesaveni, G. J. (1992) Biol. Chem. 267: 14077-14083.
[36] Mann S., Bannister, J. V., and Williams, R. J. P. (1986) J. Mol. Bio.
188: 225-232.
[37] McKenzie, R. A., Yablonski, M. J., Gillespie, G., and Theil, E. C. (1989)
Archives of Biochemistry and Biopysics 272: 88-96.
85


[38] Mertz, J. and Theil, E. C. (1983) J. Biol. Chem. 258: 1719-11726.
[39] Hitchens, G. D. (1989) TIBS 14: 152-155.
[40] Yeh, P. and Kuwana, T. (1977) Chem. Lett. 1145-1148.
[41] Koller, K. B., Hawkridge, F. M., Faugue, G., Legall (1987) J. Boichem.
Biophys. Res. Comman. 145: 619-624.
[42] Norde, W. (1986) Adv. Coll. Interface. Sci. 25: 267-340.
[43] King, B. C. and Hawkridge, F. M. (1987) J. Electroanal. Chem.
Interfacial Electrochem. 237: 81-92.
[44] Tominaga, M., Kumagai, T., Takita, S., and Taniguchi, I. (1993) Chem.
Lett. 1771-774.
[45] Taniguich, I., Hayashi, K., Tominaga, M., Muraguchi, R., and Hirose, A.,
(1993) Denki Kagaku Oyobi Kogyo Butsuri Kagaku 61: 774-775.
[46] Nishiyama, K., Ishida, H., and Taniguchi, I. (1994) J. Electroanal. Chem.
373: 255-58
[47] Grubbs, W. T. and Richard, L. H. (1989) Charge Field eff. Biosyst. -2.
[Proc. Int. Symp] 129-136. Editor(s): Allen, M. J., Cleary, S. F., and
Hawkridge, F. M., Plenum, New York, N. Y.
[48] Daido, T. and Akaike (1993) J. Electroanal. Chem. 344: 91-106.
[49] Cherry, R. J., Bjornsen, A. J., and Zapien D. C. (1998) Langmuir 14:
1971-1973.
[50] Koller, K. B. and Hawkridge, F. M. (1985) J. Am. Chem. Soc. 107:
7412- 7417.
[51] Watt, G. D., Frankel, R. B., Jacobs, D., Huang, H., and Papaefthymiou,
G. C. (1992) Biochem. 31: 5672-5679.
86


[52] Mann, S., Williams, J. M., Treffry, A., and Harrison, P. M. (1987) J. Mol.
Bio. 198: 405-416.
[53] Rohrer, J. S., Joo, M.-S., Dartyge, E., Sayers, D. E., Fontaine, A., and
Thiel, E. C. (1987) J. Biol. Chem. 262: 13385-13387.
[54] Lawson, D. M., Artymiuck, P. J., Harrison, P. M., Yewdall, S. J.,
Luzzago, A., Cesarent, G., Levi, S., and Arosi, P. (1989) FEBS Lett.
254: 207-210.
[55] Bauminger, E. R., Harrison, P. M., Mechel, D., Hodson, N. W., Nowik, I.,
Treffry, A., and Yewdall, J. J. (1993) Biochem. J. 296: 709-719.
[56] Crichton, R. R. and Roman, F. J. (1978) J. Mol. Ctalysis 4: 75-82.
[57] Crichton, R. R., (1973) Struct. Binding 17: 67-134.
[58] Wardeska, J. G., Vigliorek, B., and Chasteen, N.D. (1986) J of Biol.
Chem. 261: 6677-6683.
[59] Treffry, A., Sowerby, J. M., and Harrison, P. M. (1979) FEBS Lett. 100:
33-35.
[60] Yablonski, M. J. and Theil, E. C. (1992) Biochem. 31: 9680-9684.
87