|
|
|
|||||||
|
|||||||||
(Received for publication, May 8, 1995) From the
Unstable hemoglobins and oxidative conditions tend to produce
hemichromes which demonstrably release their heme to the erythrocyte
membrane, with consequent lipid peroxidation and cell lysis. High
levels of non-heme iron are also found in such circumstances, but the
origin of this iron is uncertain. In the present work, we show that
reduced glutathione (GSH) is able to degrade heme in solution with a pH
optimum of 7. Degradation depended on the presence of oxygen and on
heme and GSH concentrations. It was inhibited by catalase and
superoxide dismutase, implicating the involvement of perferryl reactive
species in the process of heme degradation. Heme degradation at pH 7
and 37 °C is rapid (t
The most sizable and concentrated store of heme in the body
resides in erythrocytes hemoglobin (Hb). As long as heme is bound
tightly to its hydrophobic pocket in globin, it mediates its normal
physiological role in oxygen transport. However, in some pathological
conditions or under oxidative stress, heme may be released and exert
various noxious actions. The pathobiology of heme has been reviewed
extensively(1, 2, 3, 4) , and the
main points relevant to the present work will be mentioned briefly. The
physiological oxidation of deoxyhemoglobin involves the sharing of an
electron of hemoglobin's Fe In the present work we have investigated the
interactions between heme and GSH. We have shown that co-incubation of
these compounds lead to the destruction of heme and the generation of
oxidative radicals. The heme iron and oxygen are essential for these
reactions. Addition of heme to intact red blood cells increases their
hexose monophosphate shunt activity, and, in the presence of GSH, heme
in ghost membrane is decomposed releasing its iron. These results
indicate that intracellular GSH interacts with membrane heme to produce
an oxidative stress and increase the membrane iron content. Materials were obtained from the following sources: GSH,
diethylenetriaminepentaacetic acid (DTPA), (
To measure the
time-dependent activation of HMS by heme, 5 The effect of iron on HMS activity of RBC was tested by
preincubation of cells (5
Figure 1:
Heme degradation by GSH. Heme (10
µM) was mixed with 2 mM GSH + 1 mM DTPA in Hepes buffer, pH 7.0, and incubated at 37 °C.
Absorption spectra (300-500 nm) were taken at 100-s intervals
starting immediately after mixing (5 upper continuous traces)
and then after 2 h (dotted trace). Inset, the
absorbance at 365 nm was regressed against time to fit the kinetics of
single exponential decay + offset, yielding a rate constant of
decay of 9.91
Hematoporphyrin that does
not contain iron was not degraded by GSH, even when exogenous
Fe As shown in Fig. 2A, the rate of heme degradation was
[GSH]-dependent displaying Michaelis-Menten-like kinetics
with K
Figure 2:
The
dependence of heme degradation on GSH and heme concentration. Heme and
GSH were mixed + 1 mM DTPA at the desired concentrations
in Hepes buffer, pH 7.0, at 37 °C, and heme degradation was
monitored for 200 s at 365 nm. The initial rates were calculated and
converted to nmol of heme/min. A, dependence on GSH
concentration, [heme] = 10 µM. The continuous trace is the best fit to Michaelis-Menten kinetics,
yielding k
Figure 3:
The dependence of heme degradation on pH.
Heme (10 µM) was mixed with 2 mM GSH and 1 mM DTPA in buffers preset to the indicated pH. The rate of heme
degradation was determined as described in the legend to Fig. 2.
In
absence of molecular oxygen from the reaction, or in the presence of
catalase (128 units/ml), GSH-dependent degradation did not occur.
Superoxide dismutase (90 units/ml) inhibited heme degradation by 91%.
H
Figure 4:
Iron release from heme in the presence of
GSH in vitro. The reaction conditions were the same as
described in Fig. 1. The concentration iron was determined by
the Ferrozine method. Data were fitted by nonlinear least square
regression to first order + offset kinetics (continuous
trace), yielding a rate constant of 0.527 ± 0.1
min
The destruction of heme
was also studied fluorometrically. A fluorescent product is formed
during GSH-dependent heme degradation having excitation and emission
maxima at 473 and 550 nm, respectively. Since GSH can form a
fluorescent adduct with Cu
Figure 5:
The evolution of the final degradation
product of GSH-degraded heme. Heme (10 µM) was mixed with
2 mM GSH ± 1 mM DTPA (D), in either
Hepes buffer (H) or in PBS (P), and the resulting
fluorescence was monitored with time.
Figure 6:
Degradation of membrane-associated heme by
GSH. White RBC ghosts were loaded with 30 µM heme. After
the removal of nonassociated heme, the membranes were incubated in the
presence of 10 mM GSH, and membrane heme (
To test if heme is also degraded in intact RBC, cells
were loaded with heme and subsequently incubated in PBS + sucrose
(sucrose was found to minimize spontaneous lysis of heme-loaded RBC).
The cell-associated heme was determined after extraction in the cold
with GSH in alkaline pH(10) . The amount of heme thus extracted
correlated with the concentration of heme during the loading, hence
validating this assay (data not shown). Incubation of heme-loaded RBC
resulted in a time-dependent decrease of cell-associated heme (Fig. 7). The presence of glucose in the incubation medium had
no effect on the rate of heme depletion, suggesting that intracellular
GSH, even without replenishment, was sufficient to explain the
disappearance of heme.
Figure 7:
Time-dependent depletion of heme from
heme-loaded intact RBC. RBC were loaded with heme, thoroughly washed,
and incubated at 37 °C in PBS, with (
Figure 8:
Effect of heme load on intraerythrocytic
GSH level. RBC were loaded for 30 min with 20 µM heme, and
non-cell-associated heme was washed away. Unloaded and heme-loaded RBC
were then incubated in PBS buffer (pH 7.4, 37 °C) ± 5 mM glucose, and the intracellular GSH concentration was determined as
a function of incubation time as described under ``Materials and
Methods.''
In order to test whether the
activation of HMS is due to H
Abnormal hemoglobins readily autooxidize to give
methemoglobin and hemichromes which lead to the formation of Heinz
bodies. Those were shown to bind preferentially to the membranes of
RBC(1, 26) . Structural changes in the globin molecule
may lead to serious modifications of the hydrophobic heme binding
pocket(1, 4) , ensuing in the loss of heme to other
cell components, mainly to the membrane compartment of the RBC. High
heme levels were detected in the membrane of abnormal RBC, such as
Knowledge about the mechanism which leads to the release of free
iron detected in abnormal RBC is scarce. Iron can be released from
hemoglobin or heme molecules by hydroperoxides and cause lipid
peroxidation(29) . The parallel increase of membrane heme and
non-heme iron concentrations suggests that iron derives from
heme(13) , but the mechanism responsible for this phenomenon,
or the elimination of excess intracellular heme altogether, remains an
enigma. The interaction between heme and GSH has been known for some
time(10) , but here we show directly, and for the first time,
that this interaction leads to a destruction of the tetrapyrrole ring
of heme. This is evidenced by the unique absorbance and fluorescent
spectra of the final product of the reaction, the [GSH] and
[heme] dependence of this process (Fig. 2), and the
release of iron from the destroyed tetrapyrrole ring (Fig. 4).
Heme degradation was not observed previously because experiments were
done in 0.14 M phosphate buffer at pH 8.0. Indeed, we have
observed that the degradation of heme is maximal at pH 7 and almost
undetectable in PBS at pH 8 (Fig. 3). The conclusion of these
experiments is that GSH-dependent heme degradation occurs at
physiological conditions, even when heme is bound nonspecifically to
protein. The decomposition of heme by GSH was not affected by traces
of free iron which usually contaminate various chemicals, since Chelex
100 treatment of buffers did not alter the kinetics of heme breakdown.
No effect was seen in the presence of the iron chelators DTPA and EDTA,
suggesting that iron released due to heme decomposition does not
participate in the destruction of the tetrapyrrole ring. However, DFO
completely abolished the destruction of heme, probably due to its
ability to bind to heme iron(22) . ( At the
present time, one can only speculate about the mechanism of
GSH-dependent heme degradation. The fact that it does not occur in
anaerobic conditions implies that O
displays the formation of
O
The efficient inhibition of heme degradation in presence of
either superoxide dismutase, cytochrome c, or catalase
suggests that the process depends on the oxidative effect of the
perferryl radical which is an intermediate of reaction 1. A suitable
proportion of Fe The pH dependence of GSH-mediated heme
degradation suggests that heme binds preferentially to the
nonprotonated GSH. Indeed, no spectral changes upon mixing of heme with
GSH could be observed at pH 4.4 but were conspicuous at pH > 6.5
increasing at higher pH values (not shown). The decreased ability of
GSH to bind heme at low pH will prevent the reduction of heme
Fe In
discussing the GSH-dependent mechanism of heme destruction, one should
also consider the possible involvement of thiyl radicals
(GS
Hence, in presence of O During the
destruction of heme, a fluorescent product is formed showing complex
kinetics (Fig. 5). Since the fluorescence measurement monitors
the final product, the simplest explanation is that the lag time is
required for the generation of the final product from the various
intermediates that are insinuated from the lack of distinct isosbestic
points in the time-dependent changes in the absorption spectra.
Accordingly, the t The GSH-dependent decomposition of heme also
occurs when it is dissolved in RBC membranes. White RBC ghosts retain
loaded heme for long times (18 h). When GSH is added to this system,
heme disappears with the concomitant emergence of free iron, most of
which (75%) remains associated with the membrane (Fig. 6). The
rate of production of iron, however, was slower (t The present results reveal for the first time
a mechanism that could account for the observed high levels of non-heme
iron in sickle cell membranes that initially display high levels of
heme(12, 13, 34) . Furthermore, the increased
consumption of GSH in this process is compatible with higher levels of
oxidized sulfhydryl groups found in abnormal red blood cell
membranes(3) . This and the prooxidant activity of iron in the
membrane(2, 26, 30, 34, 35, 36, 37) connect
the thiol status, the endogenous generation of oxidative stress, and
the sensitivity to prooxidants in cells with abnormal hemoglobins and
cell injury. The oxidation of GSH and the production of oxidative
radicals during the decomposition of membrane-associated heme is
expected to increase the HMS activity in intact RBC. We have clearly
demonstrated that higher levels of H In
conclusion, the present investigation indicates that heme is decomposed
by GSH, either when heme is free, bound to proteins, or dissolved in
membranes. During this process, GSH is oxidized, oxidative radicals are
produced, and iron is released, and most of it remains associated with
the membrane. It seems, however, that it is the redox cycling of the
released iron which is the dominant factor in the activation of HMS. It
is quite plausible that in sickle or thalassemic RBC, the
transformation of the unstable hemoglobin into hemichromes that release
their heme to the cell membrane, the oxidant damage due to heme
degradation, surpasses the antioxidant defense mechanisms in these
cells. This would explain their lower GSH and increased HMS activity.
Further research is needed in order to verify the nature of the
degradation products of the tetrapyrrole ring, and the types of the
free radicals generated during GSH-dependent heme degradation. However,
the elucidation of this mechanism and the identification of the
reaction products, are not pertinent to the present work which seeks to
investigate the biochemical source of membrane iron and some of the
consequences of heme degradation by GSH to RBC physiology.
Volume 270,
Number 42,
Issue of October 20, 1995 pp. 24876-24883
©1995 by The American Society for Biochemistry and Molecular Biology, Inc.
A PROPOSED MECHANISM TO ACCOUNT FOR THE HIGH LEVELS OF NON-HEME
IRON FOUND IN THE MEMBRANES OF HEMOGLOBINOPATHIC RED BLOOD CELLS (*)
ABSTRACT
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
FOOTNOTES
ACKNOWLEDGEMENTS
REFERENCES
= 70 s) and
results in the release of iron from heme. Heme that was dissolved in
red blood cell ghosts is also degraded by GSH with a concomitant
increase in non-heme iron, most of which (75%) remains associated with
the cell membrane. Loading of intact erythrocytes with heme was
followed by time-dependent decrease of membrane-associated heme and
caused an acceleration of the hexose monophosphate shunt due to the
production of H
O and the oxidation of
intracellular GSH. Most of the activation of the hexose monophosphate
pathway was due to redox cycling of iron, since iron chelators
inhibited it considerably. These results explain the origin of non-heme
iron found in the membrane of sickle cells and the oxidative stress
that is observed in these and other abnormal erythrocytes.
with oxygen,
leading oxyHb to form superoxiferriHb. Whereas this reaction is usually
reversible upon deoxygenation, occasionally the oxygen dissociates as
superoxide that creates oxidative stress and metHb. The normal red cell
is endowed with efficient mechanisms to remedy these physiological
deviations, but variant erythrocytes such as sickle and thalassemic
cells are unable some times to do so, either because their Hb has a
higher tendency to autooxidize or is less stable. Unstable metHb
readily forms hemichromes which have a tendency to bind to the cell
membrane and sometimes may release their heme. Normal HbA and HbS were
also shown to release their heme to cell membranes and to liposomes
made of aminophospholipids, although at a reduced extent compared with
metHb or hemichromes. It seems that the red cell's membrane is
the major ``sponge'' for free heme: some of it interacts
nonspecifically with the membrane proteins, and some dissolves in the
membrane's lipid bilayer. Heme has been shown to rapidly
destabilize the bilayer structure, thus increasing its permeability to
ions and leading to hemolysis(5) , and to induce the
peroxidation of the membrane lipids(6) . Its binding to
membrane proteins diminishes their reduced thiol content and leads to
cross-linking. Both latter processes are apparently due to the chronic
effect of increased membrane heme. Hence, efficient mechanisms must
exist to prevent the build-up of heme in membranes. Since heme is able
to translocate across membranes(7, 8) , it's
fate and membrane concentration are determined by the presence of
various ligands, such as serum's albumin and hemopexin, and their
relative affinities to heme vis à vis that of
the membrane(9) . Reduced glutathione (GSH) has also been shown
to remove heme from membranes (10) , but it is not known if
this scavenging is effective in vivo in protecting membranes
from the deleterious effects of heme. Altogether, it is presumed that
the concentration of free heme in the membrane is physiologically kept
at low (micromolar) levels(11) . In conditions that favor
higher membrane heme association, the concentrations of non-heme iron
also increase in the membrane(12, 13) . The origin of
this iron is not well established, but it has been suggested that it
could result from the destruction of heme by organic and lipid
peroxides and by H
O; peroxidation of sickle
cell ghosts causes an increase in free iron and a parallel decrease in
membrane-bound heme.
)catalase,
superoxide dismutase, fatty acid-free bovine serum albumin (BSA), EDTA,
deferoxamine (DFO), 3-amino-1,2,4-triazole (3-AT), Ferrozine
(3-(2-pyridyl)-5,6-bis(4-phenylsulfonic acid)-1,2,4-triazine),
neocuproine (2,9-dimethyl-1,10-phenanthroline), and ascorbic acid from
Sigma. [1-
C]Glucose was from Amersham, United
Kingdom, Chelex 100 from Bio-Rad, FeCl
and thiourea from
B.D.H. Heme, Zn-protoporphyrin and hematoporphyrin IX were from
Porphyrin Products Inc. (Logan, UT), 5,5`-dithiobis(2-nitrobenzoic
acid) from Pierce, and Hepes from Research Organics Inc. (Cleveland,
OH). Pyridoxal isonicotinoyl hydrazone (PIH) and salicylaldehyde
isonicotinoyl hydrazone (SIH) were kindly donated by Dr. P. Ponka.
Outdated human blood was a generous gift from the Blood Bank of the
Shaarei Zedek Medical Center, Jerusalem. All other chemicals were of
the best available grade. All solutions were prepared in deionized,
charcoal-filtered water.Investigation of the Destruction of the Heme Molecule by GSH:
Spectral Changes in the Heme Molecule
Heme was prepared fresh at
the beginning of each experiment as a stock solution of 1 mM in 0.2 N NaOH and was kept in the dark on ice. DTPA was
prepared in Hepes buffer as a stock solution of 20 mM. Fresh
GSH solution was prepared as a stock of 0.1 M in 0.2 M Hepes buffer, pH 7.0, or as detailed under ``Results.''
First, DTPA was added (1 mM final concentration) to the Hepes
buffer prewarmed to 37 °C. DTPA is needed in order to prevent
nonspecific oxidation of GSH(14) . Then GSH was added to a
final concentration of 2 mM, and finally heme was added to the
desired final concentration. The spectral changes between 300 and 800
nm were measured immediately after gentle mixing in a Milton Roy
Spectronic 3000 Array spectrophotometer and thereafter at 100-s
intervals. The rate of heme degradation was measured by calculating the
decrease at 365 nm, a wavelength characteristic of the heme
GSH
complex(10) . The pH dependence of heme degradation by GSH was
measured as described above, in Hepes buffer or PBS adjusted to the
desired pH. The pH was determined at the beginning and at the end of
the reaction and was found to be constant. It was ascertained that heme
did not precipitate out of solution during the measurement. For
calculation of the amounts of heme that were degraded, we used a
millimolar extinction coefficient which was determined in our
laboratory to be 64.1 ± 1.3 at 365 nm. The effect of the hypoxic
conditions were measured by bubbling nitrogen through the buffer and
the stock solutions, and the degradation of heme was followed as
described above under continuous exposure to nitrogen.Iron Release from the Heme Molecule during the
Degradation Reaction
A reaction mixture of the heme decomposing
system was prepared in a final volume of 5 ml at 37 °C. The heme
concentration was 10 µM, and the GSH concentration was 2
mM. DTPA was not added to this reaction because it interferes
with iron determination. Free iron was measured by the Ferrozine
method(15) . Briefly, a 0.45-ml sample taken from the reaction
mixture was mixed with 50 µl of 100% (w/v) trichloroacetic acid.
Then, 0.5 ml of 0.02% ascorbic acid in 0.1 N HCl was added,
the system was incubated for 5 min at room temperature, and 0.4 ml of
ammonium acetate (10%) and 0.1 ml of Ferrozine solution (75 mg of
Ferrozine and 75 mg of neocuproine in 25 ml water) were added. After an
additional incubation for 5 min at room temperature, the color
developed was measured at 562 nm. The iron concentration was calculated
using a millimolar extinction coefficient of 27.9.Fluorometric Determination of GSH-dependent Heme
Degradation
GSH (2 mM) and heme (10 µM)
were mixed in Hepes buffer and incubated at 37 °C for 2 h to allow
for complete degradation of heme. The excitation and emission spectra
of the final product were determined in a SPEX Fluorolog II (all slits
were set at 5 mm), and the maxima were found to be 473 and 550 nm,
respectively. The evolution of the final product with time was followed
in various incubation conditions as described under
``Results.''GSH Oxidation during Heme Degradation
The
oxidation of GSH in the presence of heme was measured by the method of
Beutler (16) with a minor modification: DTPA was added at 1
mM to the assay medium to chelate heavy metals which may
participate in redox reactions and lead to a further GSH oxidation.
Hence, the measured decrease in GSH concentration was due to the
interaction of heme with GSH.Degradation of RBC Ghosts-associated Heme by GSH,
Detection and Compartmentalization of the Iron Released
Ghosts
from RBC were prepared by 1:10 dilution of a known number of RBC into
an ice-cold solution of 5 mM NaHPO
, pH 8.0 (5P8).
Ghosts were washed with cold 5P8 until they became white (27,000
g, 10 min at 4 °C; usually 5 wash cycles were
needed with a 5-min incubation on ice between washes). White ghosts
were suspended in 0.2 M Hepes, pH 7, to a concentration of
50-80 10
ghosts/ml. A known amount of the
ghosts (usually 7.4-16 10) were suspended in
3 ml of Hepes buffer containing 30 µM heme and incubated
for 60 min at 37 °C. Ghosts were then washed twice in 10 ml of
ice-cold 5P8 and finally resuspended in Hepes buffer. In order to
determine the heme inside the membranes, those were dissolved by adding
SDS (1% w/v final concentration) and the absorption spectrum was
measured between 300 and 800 nm. For the calculations of heme
concentration, we used a millimolar extinction coefficient (in the
presence of 1% SDS) of 83.5 ± 1.8 at 399 nm which was determined
in our laboratory. The same amount of ghosts was taken for the
measurement of total non-heme iron (free and membrane-bound). Iron was
assayed as described before (13) with a minor modification.
Briefly, 200 µl of the ghost suspension was dissolved with 500
µl of 0.6% SDS in 0.2 M sodium acetate, pH 4.5. A
400-µl aliquot of the SDS-dissolved ghosts were taken into 500
µl of reductants solution (0.2% ascorbic acid and 0.2% sodium
dithionite dissolved in 0.2 M sodium acetate, pH 4.5) and
incubated for 5 min at room temperature. Then, 100 µl of the color
developing solution were added (200 mg of Ferrozine and 1.25 g of
thiourea or 200 mg of neocuproine dissolved in 50 ml of water). After 5
min of incubation at room temperature, the absorbance was measured at
562 nm. A millimolar extinction coefficient of 27.9 was used for
calculations of iron concentration. To measure the ghost-associated
iron, the ghost suspension was spun down and the supernatant was
discarded. The ghosts were incubated for 10 min at 37 °C with 1 ml
of 0.5 mM DTPA in order to chelate all the free iron from the
solution, were washed three times in 1 ml of Hepes buffer in order to
remove the residual DTPA, and the iron content was measured as
described above. Controls (ghosts not loaded with heme) were treated in
the same way to measure the basal levels of iron and heme.Degradation of Heme in Heme-loaded Intact RBC
GSH
at alkaline pH has been shown previously to extract heme from RBC ghost
membranes and to form a hemeGSH complex absorbing at 368
nm(10) . This technique has been used to determine the
degradation of heme in heme-loaded intact RBC. Washed RBC in PBS (10 ml
at 5% hematocrit) were incubated with 28 µM heme for 10
min at 37 °C. Cells were spun down and washed twice with 10 ml of
PBS supplemented with 50 mM sucrose to reduce lysis. The cell
pellet was resuspended in the same buffer ± 10 mM glucose and incubated at 37 °C. Aliquots of 1 ml were taken
immediately and at different time intervals and centrifuged, and the
cells were resuspended in ice-cold alkaline (pH 8.7) PBS supplemented
with 50 mM sucrose and 2 mM GSH. After 7 min on ice,
cells were spun down, and the absorption spectrum of the supernatant
was determined. The contribution of hemoglobin (present due to
spontaneous lysis) was subtracted from the absorbance at 368 nm, and
the amount of heme was calculated using an extinction coefficient of
60.85 mMcm.
Intracellular Hydrogen Peroxide Production in Heme-loaded
RBC
Catalase bound to HO is irreversibly
inhibited by 3-amino-1,2,4-triazole (3-AT; (17) ). Hence, the
extent of inhibition is proportional to the concentration of
HO. RBC were suspended in PBS containing 40
µM heme and 40 mM 3-AT, and various additives as
described under ``Results,'' at 5 10 cells/ml, and incubated at 37 °C. Samples of 1 ml were washed
three times with 1 ml of PBS with 2-min intervals at 37 °C and then
lysed in 1 ml of ice-cold 5P8. After centrifugation (10 min at 16,000
g), catalase activity was measured as described
before(18) . Briefly, to 10 µl of 4-fold PBS-diluted lysate
(equivalent to 1.25 10
cells) were added 990 µl
of 6 mM HO, and the decomposition of
the substrate was followed by monitoring the absorbance at 236 nm. A
millimolar extinction coefficient of 0.071 was used for the calculation
of catalase activity.Effect of Heme and Iron on the Hexose Monophosphate Shunt
of Human Red Blood Cells
Red blood cells (RBC) were separated
from other cells and plasma by 3-4 washes in RPMI 1640 medium
supplemented with 8 mM NaHCO and 25 mM Hepes. The activity of the RBC hexose monophosphate shunt (HMS)
was assayed by measuring the evolution of CO from [1-C]glucose as described
before(19) . Normal RBC were suspended at 5 10 cells/ml in RPMI 1640 and preincubated with different
concentrations of heme for 30 min at 37 °C. RBC were then washed 5
times in 5 ml of glucose-free RPMI 1640 supplemented as above and 5
mM glucose. To measure the effect of heme on HMS activity, RBC
were incubated with the desired concentration of heme for 30 min at 37
°C, washed of nonassociated heme, and the HMS activity was assayed
without or with 2 mM GSH and/or 1 mM DTPA. The
combination between DTPA and GSH without heme or DTPA alone did not
affect the HMS activity. On the other hand, GSH alone caused a
significant activation of the RBC HMS. This was probably due to
nonspecific oxidation of the GSH by the heavy metals which usually
contaminates the solutions. This was abolished completely when DTPA was
used at a 1 mM final concentration. 10 RBC/ml were incubated for different times with a 40 µM heme. At the end of each incubation period, the cells were treated
by the same way as described above, and the HMS activity was measured. 10 cells/ml) for 30 min
at 37 °C in RPMI 1640 supplemented with 10-100 µM of the following additives:
Fe(NH)(SO),
FeCl:PIH, and FeCl:SIH both at 1:1 molar
ratios, and FeCl:Na-citrate (1:20).
Fe
and the various forms of chelated Fe
are known to penetrate into cells(20, 21) .
After the preincubation, cells were washed and resuspended in
glucose-free RPMI 1640 supplemented with 5 mM glucose and the
same concentration of iron additive, and HMS was measured as described
above. Controls containing 100 µM PIH or SIH were run in
parallel. To test the effect of iron released during the degradation of
heme in heme-loaded RBC on HMS activity, RBC (5% hematocrit) were
incubated for 30 min at 37 °C in RPMI 1640 containing 50 µM freshly prepared heme, in the absence or presence of 100
µM DFO, DFO + 100 µM PIH, or DFO +
100 µM SIH. Heme was washed off, and HMS activity was
measured in the presence of the chelators. Control RBC were run in
parallel.
Degradation of Heme by GSH
Mixing of heme with
GSH results in a blue shift in the maximal absorbance of heme from
387.5 to 365 nm as observed previously(10) , probably due to
the formation GSHheme complexes, and a rapid decline in peak
absorbance indicating the degradation of the heme molecule (Fig. 1). The absence of well-defined isosbestic points suggests
that degradation succeeds through various intermediate products. The
rate of heme degradation was fastest in 0.2 M Hepes and
considerably slower in either PBS, 0.1 M
NaHPO, or 0.05 M Tris base titrated
with HSO, all at pH 7.0. As shown in the inset of Fig. 1, heme degradation follows a
monoexponential decay with t of 70 ± 0.45
s. The final product of this reaction (lowest trace in Fig. 1) is obtained after substantially longer incubations
(
2 h). The chemical nature of the end product(s) has not been
investigated although it has been ascertained that it is neither
biliverdin nor bilirubin (data not shown). That heme destruction is not
due to contaminating traces of iron was evidenced either by using
buffer solutions that were passed through a Chelex 100 resin, or by
adding DTPA to the reaction mixture. In both cases, there was no change
in the rate of heme degradation. In spite of this fact, we included
DTPA regularly in order to discard any nonspecific heavy
metal-catalyzed GSH oxidation reactions.
10 ± 6.00
10 s
. The derived t
was 70 ± 0.45
s.
was added as FeCl
(data not shown). No
degradation could be observed with Zn-protoporphyrin. Deferoxamine
(DFO) that interacts with the heme iron as evidenced by a blue shift in
the maximal absorbance (22) almost completely inhibited the
degradation of heme by GSH (data not shown). = 1.59 ± 0.08
nmol
min and an apparent K
= 0.49 ± 0.07 mM. Heme degradation is also
[heme]-dependent with a K =
8.68 ± 0.92 nmol
min and an apparent K
= 59.04 ± 9.28 µM (Fig. 2B). These kinetics are reminiscent of those
described for heme degradation by
HO(23, 24) , but no further
attempts were made to determine the precise molecular mechanism. Heme
degradation is pH-dependent in either 0.2 M Hepes or PBS,
peaking at pH 7 (Fig. 3). The technique used for the assay of
heme degradation precludes the possibility that the low rates observed
at acid pH are due to precipitation of heme. Degradation also occurs
when heme is bound nonspecifically to protein. Heme was complexed with
defatted BSA in Hepes buffer by mixing the two compounds for 30 min at
room temperature at a molar ratio of 70:1 (heme:BSA) and then subjected
to the same spectrophotometric assay in the presence of GSH as done
with free heme. The rate of degradation of BSA-bound heme was somewhat
lower than that of free heme, e.g. 1.005 and 1.69
nmolmin, respectively (Fig. 3).
= 1.586 ± 0.079
nmol/min and K
= 0.492 ±
0.107 mM. B, dependence on heme concentration,
[GSH] = 2 mM. K = 8.68 ± 0.92 nmol/min and K
= 59.04 ± 9.28
µM.
, 0.2 M Hepes; , PBS;
, defatted BSA at 70:1
(heme:BSA) molar ratio.
O is known to destroy
heme(23, 24) , and the absorption spectrum of heme
degraded in presence of high (0.5 mM) HO concentration is similar to, but not identical with, the final
product of GSH-dependent heme degradation (not shown).The Liberation of Iron from Degraded Heme
The
decomposition of the heme molecule leads to the liberation of the heme
iron as determined by the Ferrozine method (Fig. 4). At the same
time, GSH undergoes oxidation to GSSG: during the 8.3 min that are
required for the destruction of most heme by GSH, 200 µM GSH (10% of the total amount of the GSH present) was oxidized.
This amount exceeds the quantity of heme present (10 µM),
indicating that oxidation of GSH did occur.
, t
= 1.3 ±
0.25 min.
(25) , we have
verified that the final product is not due to GSH:iron adduct
formation. It is due neither to the mere release of iron nor to the
production of bilirubin or biliverdin, as neither hematoporphyrin nor
the latter compounds ± GSH, ± iron, yielded any specific
fluorescence. The kinetics of end product formation shown in Fig. 5are rather complex, characterized by a lag phase, then a
rapid rate of production followed by a slower one. The rapid rate is
faster in Hepes buffer compared to PBS. Addition of DTPA reduces the
lag time and increases the rapid rate of end product formation, but has
no effect on the slower rate.
= 473
nm;
= 550 nm.
Effect of GSH on RBC Membrane-associated
Heme
Since in RBC heme is released from denatured or oxidized
hemoglobin into the cell membrane, and normal red cells contain
millimolar GSH concentrations, it was interesting to study whether
membrane-associated heme can also be degraded by GSH. Our results
demonstrate that GSH interacts with heme associated with RBC ghost
membranes and decompose the heme molecule thereby releasing iron. White
ghosts loaded with heme do not lose the heme. However, when GSH was
added, heme was degraded as assessed by dissolving the membranes in 1%
SDS and measuring the absorption spectrum, from which the accurate
amount of the heme in the membranes was calculated. The final spectrum
after long incubations (6 h) resembles that obtained at the end of
the direct reaction between GSH and heme in solution (not shown).
During the degradation of heme, iron is liberated. As indicated in Fig. 6: 1) heme degradation and iron release were found to be
linear with time up to 3 h; 2) the free iron detected in the whole
system is stoichiometrically related to the amount of heme that
decomposed; and 3) about 75% of liberated iron remains in the membrane
compartment, while the rest is released into the aqueous phase of the
assay system.
), total
non-heme iron (
), and membrane-associated non-heme iron ()
were determined in aliquots taken at various time intervals as
described under ``Materials and
Methods.''
) or without ()
glucose. The heme content in the RBC was determined with incubation
time, as described under ``Materials and Methods.'' The continuous line depicts the best fit of the data to a single
monoexponential decay process. The derived t is
73.4 ± 3.8 min.
Depletion of Intracellular GSH in Heme-loaded
RBC
Control RBC and RBC loaded with 20 µM heme were
suspended in PBS with or without 5 mM glucose and were
incubated at 37 °C. At different time intervals, samples were taken
and assayed for their GSH content in the presence of 1 mM DTPA. As shown in Fig. 8, in the presence of glucose,
control cells maintained a steady GSH level. The omission of glucose
resulted in an expected decrease in GSH reflecting its normal
endogenous utilization. In heme-loaded RBC, GSH decreased substantially
even in the presence of glucose, but much more so in its absence,
indicating that intracellular GSH is consumed by the degradation of
heme.
, control RBC + glucose;
, control
RBC without glucose; , heme-loaded RBC + glucose; ,
heme-loaded RBC without glucose.
Heme and GSH Generate Hydrogen Peroxide
Intracellularly
Heme can translocate across membranes (7, 8, 9) and thus can interact with both
intracellular and exogenous GSH. During the GSH-mediated decomposition
of membrane-associated heme, HO can be
generated (as evidenced by the inhibitory effect of catalase, see
above). To test this presumption, RBC were incubated in PBS ±
heme and ± 3-AT, an irreversible inhibitor of
HO-associated catalase(17) . Hence, the
extent of inhibition of catalase is used as a measure of intracellular
HO. Results shown in Table 1indicate
that incubation with heme alone resulted in a time-dependent
inactivation of catalase, indicating that HO was generated. In the presence of 2 mM GSH (+ 1
mM DTPA), catalase was inhibited to a much greater extent,
indicating that HO was generated, and at least
part of it was released to the cellular compartment during the
degradation of heme by GSH. Extracellular GSH (+1 mM DTPA) in absence of heme did not affect catalase activity.
Heme and GSH Activate the Hexose Monophosphate Shunt in
RBC
The effect of heme on the hexose monophosphate shunt (HMS)
activity was utilized as an indicator for the interaction of the
endogenous GSH and heme in intact RBC. Since part of the GSH is
oxidized and HO was shown to be produced during
this process, both effects should increase the activity of the HMS due
to the consumption of NADPH during the reduction of oxidized
glutathione by glutathione reductase. To test the effect of heme on HMS
activity, RBC were loaded with 40 µM heme for various
lengths of time, and HMS activity was measured. The relation between
HMS activity and time of incubation with heme was analyzed by nonlinear
regression of the data to fit first order kinetics + offset (the
basal HMS activity). The basal HMS activity thus calculated was 0.153
± 0.015 µmol of glucose consumed/10 cells/h,
and it was maximally increased by 2.3-fold. The heme-dependent activity
was 0.199 ± 0.024 µmol of glucose consumed/10
cells/h. The half-time of HMS activation was 24.8 ± 7.8
min. To test the effect of heme concentration on HMS, RBC were
preincubated with increasing heme concentrations for 30 min, and HMS
activity was measured. Data were analyzed by nonlinear regression to
fit the Michaelis-Menten equation. The activation of HMS is saturable
with [heme], the derived K
and k increases in HMS activity are 27.05 ±
9.48 µM and 0.193 ± 0.028 µmol of glucose
consumed/10
cells/h. In the presence of externally added
GSH, the respective values were 64.22 ± 24.02 µM and 0.685 ± 0.144 µmol/10
cells/h. Since
GSH does not enter into RBC, these results suggest that extracellular
GSH can degrade membrane-associated heme with consequent activation of
HMS. Catalase (128 units/ml) and superoxide dismutase (90 units/ml)
added to the bathing medium did not affect the HMS activation by GSH
and heme (data not shown).
Activation of HMS by Iron
Since iron derived from
degraded heme can enter into redox cycling and generate
HO, it could in itself activate HMS. To test
this possibility, RBC were preincubated in the presence of increasing
concentrations of Fe or membrane-permeable forms of
Fe
, and HMS was measured.
Fe
-citrate had no effect, and Fe
or
Fe
:PIH at 100 µM increased HMS activity
only slightly (0.067 over a basal activity of 0.165 µmol of glucose
consumed/10
cells/h). In presence of
Fe
:SIH, HMS increased by 0.155 at 10 µM and 0.333 at 100 µM, suggesting that this chelator
was the best vehicle for the introduction of iron into RBC. Indeed, the
1-octanol:water partition coefficient of SIH is 1 order of magnitude
higher than that of PIH(25) .
O generated during
the degradation of heme or due to the release of iron which may enter
into redox cycling and the consequent generation of the peroxide, RBC
were loaded with heme in the absence or presence of DFO alone, DFO
+ PIH, or DFO + SIH. Thereafter, cells were washed and HMS
was measured in the presence of chelators alone. Results shown in Table 2indicate that none of the chelators had any effect on the
HMS activity of control RBC. DFO alone or DFO + PIH had no effect
on the activity of heme-loaded cells, but in presence of DFO +
SIH, HMS activity was considerably reduced, although the remaining
activity was higher than control levels. These results suggest that SIH
can chelate intracellular iron that is released during the degradation
of heme and translocate it to the extracellular medium where it is
chelated by DFO (the stability constant of Fe:DFO is
substantially higher than that of Fe
:SIH). (
)Consequently, iron freed from heme is the major culprit
responsible for HMS activation.
-thalassemic (27) and sickle
cells(11, 12) . In parallel, a significant
phospholipid-bound fraction of non-heme iron was detected in the
membranes of these cells(13, 28) . Iron
decompartmentalization was recently suggested to be an important
feature of abnormal RBC and an important cause for cell
lysis(29) . This membrane-bound iron may play an important role
in RBC disorders and may be the causative factor for oxidative membrane
damage. Lipid peroxidation and oxidation of thiol groups were shown to
be characteristic features of the abnormal RBC(3) . ) is involved. Since heme
degradation is not observed with hematoporphyrin or Zn-protoporphyrin,
heme iron is essential for the complexation of heme and GSH, as
previously suggested(10) , and this iron is essential for the
destruction of heme by GSH. The heme iron is most probably in the
Fe state and can be reduced to Fe
by GSH while the latter is oxidized to GSSG. The following
reactions are then bound to occur (even if iron is still bound to the
porphyrin ring, but not to DTPA):
, whereas reactions 2 and 3 show the
nonenzymatic dismutation of O
which does
occur at physiological pH, although 4 orders of magnitude slower than
the superoxide dismutase-mediated reaction(30) . Once
O
and H
O are
present in the system, the following reactions could take place:
O
and
OFeO
in this radical
may be necessary for the destruction of heme, as has been suggested for
the mechanisms of iron-mediated lipid peroxidation(31) . and could alter this proportion, thus
explaining the inhibitory effect of superoxide dismutase and catalase
which could prevent this alteration.
OH radicals which
can be formed from the oxidation of Fe
and
H
O are probably not involved in heme
destruction, as the rate of destruction in the presence of OH radicals scavenger Hepes is faster than in any
other buffer tested.
. At the high pH range, binding does occur but the
spontaneous dismutation of O
to
H
O is considerably decreased, thus potentially
altering the composition of the perferryl intermediate.) which could be formed by the reaction of GSH with
O
, HO
and/or
OH radicals. As previously reviewed(32) , the
following reactions can take place:
, GSH can form a
GS radical and H
O, and the
radical can then undergo a series of reactions that lead to the
formation of GSSG and O. These reactions
may explain the greater than stoichiometric oxidation of GSH observed
in the present and other studies (14, 32, 33) and insinuates the involvement
of the GS
, GSOO
, and
GSS
]G
radicals in the
destruction of the tetrapyrrole ring of heme. Further precise
investigations on the mechanism of heme degradation by GSH are needed
to clarify the details of this important phenomenon.
for the generation of the
final product is considerably longer (approximately 180 s or longer in
the absence of DTPA or in PBS, Fig. 5) than that of heme
destruction monitored by the decline of absorbance (70 s, Fig. 1, inset). Apparently, the precise mechanism of
heme destruction and iron release from it is rather complex, as well as
the effect of Hepes which accelerates both heme destruction and the
formation of the final fluorescent product, and the expediting effect
of DTPA on the formation of the final product. The precise sequence of
events obviously requires further investigation, but it does not
preclude us from pondering the biological implications of GSH-dependent
heme destruction.
180 min) than that observed in aqueous solutions (t = 1.3 min, Fig. 4), probably
due to limited accessibility of GSH to membrane-associated heme. Heme
has also been shown to be depleted from heme-loaded intact RBC with a t
of 73.4 ± 3.8 min (Fig. 7),
indicating that heme is also degraded in intact RBC. In parallel, GSH
is oxidized (in heme-loaded RBC and in the absence of glucose, t
= 67.9 ± 11.4 min), implicating
GSH in heme depletion.
O are
produced in RBC incubated in the presence of heme and more so when GSH
was added to the medium (Table 1). That neither extracellular
catalase nor superoxide dismutase influenced this process suggests that
oxidative radicals are delivered to the intracellular compartment. In
the absence of extracellular GSH, intracellular GSH fulfills the task,
as evidenced by the decline of GSH levels in heme-loaded RBC (Fig. 8). This decrease can be accounted for by direct oxidation
of GSH to GSSG through thiyl radical formation (see above) or due to
the detoxification of HO by GSH peroxidase.
That this effect is exacerbated in the absence of glucose suggests that
the amplification of HMS could provide much of the reducing equivalents
(in the form of NADPH) to reduce GSSG back to GSH by glutathione
reductase. Indeed, considerably higher HMS activity was detected in
heme-loaded RBC and more so when GSH was added to the bathing medium.
The activation of HMS was rapid (t of 25 min),
in agreement with the fast translocation of heme across
membranes(7, 8) , and depended on the heme
concentration. It could not be ascertained whether the rate-limiting
step in HMS activation is the translocation of heme through the
membrane or the interaction of heme with intracellular GSH. Most of the
activation of HMS is due to iron liberated from heme and dissociating
from the membrane into the intracellular compartment, which enters into
redox cycling to generate H
O. Indeed,
introduction of Fe into the cells increased HMS
activity. In the presence of chelators that can trap the iron released
during the degradation of heme in heme-loaded RBC, the heme-dependent
HMS activity is reduced by 77% (Table 2). Thus,
H
O produced by the redox cycling of released
iron seems to play a major role in the activation of HMS.
)))
We thank Dr. P. Ponka for the generous gift of PIH and
SIH and Professor Z. I. Cabantchik for helpful discussions.
©1995 by The American Society for Biochemistry and Molecular Biology, Inc.
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  D. J. Creek, W. N. Charman, F. C. K. Chiu, R. J. Prankerd, Y. Dong, J. L. Vennerstrom, and S. A. Charman Relationship between Antimalarial Activity and Heme Alkylation for Spiro- and Dispiro-1,2,4-Trioxolane Antimalarials Antimicrob. Agents Chemother., April 1, 2008; 52(4): 1291 - 1296. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. T. Huy, K. Mizunuma, K. Kaur, N. T. T. Nhien, M. Jain, D. T. Uyen, S. Harada, R. Jain, and K. Kamei 2-tert-Butyl-8-Quinolinamines Exhibit Potent Blood Schizontocidal Antimalarial Activity via Inhibition of Heme Crystallization Antimicrob. Agents Chemother., August 1, 2007; 51(8): 2842 - 2847. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. T. Huy, D. T. Uyen, A. Maeda, D. T. X. Trang, T. Oida, S. Harada, and K. Kamei Simple Colorimetric Inhibition Assay of Heme Crystallization for High-Throughput Screening of Antimalarial Compounds Antimicrob. Agents Chemother., January 1, 2007; 51(1): 350 - 353. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. de Vogel, D. S.M.L. Jonker-Termont, E. M.M. van Lieshout, M. B. Katan, and R. van der Meer Green vegetables, red meat and colon cancer: chlorophyll prevents the cytotoxic and hyperproliferative effects of haem in rat colon Carcinogenesis, February 1, 2005; 26(2): 387 - 393. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. T. Huy, S. Serada, D. T. X. Trang, R. Takano, Y. Kondo, K. Kanaori, K. Tajima, S. Hara, and K. Kamei Neutralization of Toxic Heme by Plasmodium Falciparum Histidine-Rich Protein 2 J. Biochem., May 1, 2003; 133(5): 693 - 698. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. C. P. Steele, R. J. Phelps, M. S. J. Simmonds, D. C. Warhurst, and D. J. Meyer Two novel assays for the detection of haemin-binding properties of antimalarials evaluated with compounds isolated from medicinal plants J. Antimicrob. Chemother., July 1, 2002; 50(1): 25 - 31. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. T. Huy, K. Kamei, T. Yamamoto, Y. Kondo, K. Kanaori, R. Takano, K. Tajima, and S. Hara Clotrimazole Binds to Heme and Enhances Heme-dependent Hemolysis. PROPOSED ANTIMALARIAL MECHANISM OF CLOTRIMAZOLE J. Biol. Chem., February 1, 2002; 277(6): 4152 - 4158. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. Atamna, J. Liu, and B. N. Ames Heme Deficiency Selectively Interrupts Assembly of Mitochondrial Complex IV in Human Fibroblasts. RELEVANCE TO AGING J. Biol. Chem., December 14, 2001; 276(51): 48410 - 48416. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| All ASBMB Journals | Molecular and Cellular Proteomics |
| Journal of Lipid Research | ASBMB Today |