From www.bloodjournal.org by guest on June 18, 2017. For personal use only.
Catalysis of Soluble Hemoglobin Oxidation by Free Iron on
Sickle Red Cell Membranes
By Oded Shalev and Robert P. Hebbel
Abnormal deposition of hemichrome on the inner aspect of
the sickle red cell membrane promotes premature cell demise. The steps proximate to hemichrome formation in
these cells are poorly understood. To test the hypothesis
that the pathologic deposits of free ferric iron located on the
inner aspect of sickle red cell membranes would be redox
active and promote oxidation of soluble oxyhemoglobin, we
incubated native versus iron-stripped sickle or normal ghost
membranes with oxyhemoglobin S. We found that sickle
membranes exerted an exaggerated effect on methemoglobin formation in solution, an effect completely accounted
for by their abnormal content of free iron. This ability of
sickle membranes to promote hemoglobin oxidation was
not diminished by catalase or by presence of a high-affinity,
iron-inactivating chelator that is unable to remove membrane iron. Examination of those membranes likewise revealed that their free iron content promoted deposition of
additional heme-protein. These resuks establish that the potential redox couple formed by membrane-associated ferric
iron and cytoplasmic oxyhemoglobin is promotive of hemoglobin oxidation and deposition of hemichrome on the membrane. This predicts that removal of pathologic membrane
iron might help prevent the detrimentalformation of methemoglobin and hemichrome in vivo, insofar as this is accelerated by transition metal.
0 1996 by The American Society of Hematology.
T
dressable therapeutically, by which the membrane itself
might directly promote hemoglobin oxidation. Under solution conditions, ferric iron can oxidize oxyhemoglobin to
methemoglobin, the free iron and the hemoglobin molecule
thus comprising a redox couple affecting valence cycling of
both iron m01ecules.l~~’~
The free iron on the inner aspect of
sickle erythrocyte
is largely, if not exclusively, ferric and is bioavailable for valence
We
thus postulated that this pathologic membrane iron would be
able to catalyze soluble methemoglobin formation, leading
ultimately to deposition of denatured hemoglobin on the red
cell membrane.
To test this hypothesis, it was necessary to compare the
effect of sickle red cell membranes that were in native state
with those that had been completely stripped of their free
iron deposits. To accomplish this we used the iron chelator,
deferiprone (“Ll”; 1,2-dimethyl-3-hydroxypyrid-4-one),
which has already been documented to completely remove
pathologic free iron deposits from the cytoplasmic surface
of the membranes even of intact sickle red cells.” The effective use of L1 in the current studies punctuates its potential
efficacy in vivo for a new indication, prevention of methemoglobin formation accelerated by the presence of transition
metal.
HE ABNORMAL ACCUMULATION of catalytic iron
on the cytoplasmic surface of sickle erythrocyte mem-
branes has been implicated as being pathophysiologically
relevant to their dysfunction and premature demise.’ The
diverse functional and structural defects that have been ascribed, at least in part, to iron-mediated oxidative damage
in these cells include: enhanced KC1 co-transport activity,’
increased leak susceptibility to def~rmation,~
abnormal tendency to vesiculation: deficient microrheological properties: and hemichromehand 3 co-~lustering.~~~
The latter is
particularly important because it leads to accumulation of
immunoglobulin and complement, and ultimately to destruction of red cells by macrophages.6.” Hence, understanding
the events leading to hemichrome formation on the sickle
cell membrane would enhance our understanding of sickle
disease pathophysiology.
Since methemoglobin formation is a necessary step proximate to formation of hemichromes, we have focused on
potential mechanisms that might underlie this conversion.
One possible proximate event might be the tendency of soluble hemoglobin S to undergo abnormally rapid auto-oxidation of its heme.” In theory, methemoglobin or hemichrome
thus formed in the cytoplasm could migrate to the membrane
and there accumulate. Alternatively, the oxidizing and denaturing effect on cytosolic hemoglobin of aminophospholipids
at the inner leaflet of the lipid bilayer’’ could lead to local
formation of methemoglobin and hemichrome directly at the
cytoplasdmembrane interface.
However, there is another mechanism, one potentially ad~~
From the Department of Medicine-Division of Hematology, University of Minnesota Medical School, Minneapolis, MN.
Submitted July 31, 1995; accepted December 26, 1995.
Supported by the National Institutes of Health (Grant No.
HL37528).
Address correspondence to Robert P. Hebbel, MD, Department of
Medicine-Division of Hematology, University of Minnesota Medical
School, 420 Delaware Sr, SE, Box 480, Minneapolis, MN 55455.
The publication costs of this article were defrayed in part by page
charge payment. This article must therefore be hereby marked
“advertisement” in accordance with 18 U.S.C. section 1734 solely to
indicate this fact.
0 1996 by The American Society of Hematology.
0006-4971/96/8709-0022$3.00/0
3948
MATERIALS AND METHODS
RBC membrane preparation. Heparinized or citrated blood was
obtained from volunteer normal individuals and patients homozygous for HbS. All samples were processed within 24 hours of bloodletting by washing them three times with 10 volumes of ice-cold
phosphate buffered saline (PBS, pH 7.4),removing the buffy coat
between washes. Washed red cells (0.5 mL) were lysed in 45 mL
ice-cold, iron-free 5 mmol/L sodium phosphate (5P) buffer, pH 8
containing 0.5 mmoVL EDTA, and centrifuged for 15 minutes at
14,000rpm, 4°C.The membrane pellet was washed four more times
with the same buffer with multiple passages of the pellet through a
22 gauge needle between each wash, a method that preserves all
membrane component^.'^ The membranes were then washed two
more times in iron-free 5P buffer, pH 7.4,without EDTA, at which
point amounts of membrane free-iron, heme and protein were determined. All solutions used for preparing RBC membranes were rendered iron-free by pretreatment with the chelating resin iminodiacetic
acid (Sigma Chemical Co, St Louis, MO).
Measurement of RBC membrane free-iron. The nonheme nonfemtin (“free”) iron content of the membranes was determined as
Blood, Vol 87, No 9 (May I), 1996: pp 3948-3952
From www.bloodjournal.org by guest on June 18, 2017. For personal use only.
MetHb FORMATION CATALYZED BY MEMBRANE IRON
3949
as previously described." Results are expressed as nanomoles per
milligram membrane protein.
Stnristicd andysis. For statistical analysis we used the Student's
t-test. paired or unpaired as appropriate.
20 1
i
e
._
RESULTS
W
E
c
-
0
0.025
0.05
T
T
0.125
0.25
0.5
L1 concentration, mM
Fig 1. Sickle RBC membrane free iron (mean 2
SD)before (16.1
? 0.7 nmollmg) and following incubation of the ghost membranes
with various concentrations of L1. L1, 0.5 mmol/L, is sufficient to
eliminate all detectable free iron on the membranes (UD = undetectable). The results, at each concentration of L1, represent data from
three separate experiments performed in duplicate.
previously described".'"'" by its rapid reactivity with ferrozine
(which does not detect heme iron) in the presence of the denaturant
sodium dodecyl sulfate and the reducing agents ascorbate and sodium metabisulfite. All reagents used for measurement of the membrane iron content were rendered iron-free by pretreatment with the
chelating resin iminodiacetic acid. Amounts of iron are expressed
in nanomoles of iron/milligram of membrane protein. The protein
content of the membranes was determined by the BioRad (Richmond. CA) microprotein assay.
R e m o l d of memhmne free-iron. To determine the conditions
required for deferiprone to remove all detectable membrane freeiron. the ghost membranes were incubated (at 1 mg membrane protein/mL in SP buffer, pH 7.4) with various concentrations (0 to 0.5
mmol/L) of deferiprone at 37°C for 30 minutes. At the end of the
incubation, the membranes were washed extensively with ice-cold
5P buffer (pH 7.4) and immediately assayed for free-iron content as
above.
Conversiori of HhS to MetHbS. To determine whether normal
and sickle RBC membranes can affect the conversion of oxyHbS to
metHbS, and whether this is catalyzed by the pathologic free iron
on sickle membranes. native and deferiprone-treated normal and
sickle membranes ( I mg/mL) were incubated with HbS ( I mg/mL)
in 5P buffer (pH 7.4). at 37°C for various periods of time. Parallel
incubations of HbS ( I mg/mL) alone were performed to determine
the rate of spontaneous (ie. membrane-free) metHb formation under
identical conditions. At the end of the incubation the samples were
centrifuged (15 minutes, 14,000 rpm. 4°C) and their supematants
analyzed spectrophotometrically for metHbS content, calculated as
described" from absorbance at 560,577,630. and 700 nm. Of note,
even at the maximal time point used in these studies (8 hours),
membranes incubated in this manner preserved the same ratio of
lipid phosphorus to protein that they had at zero-time (data not
shown).
Deposition of heme on the memhmnes. The membrane pellet
obtained from the previous step was assayed for total heme (ie,
hemoglobin, hemichrome, free heme) content by dissolving the
memhranes in formic acid and recording the absorbance at 398 nm.
Dose response studies showed that incubation of sickle
ghost membranes with increasing concentrations of L1 resulted in a gradual decrease in the membrane free iron such
that at 0.5 mmol/L none could be detected by the ferrozine
assay (Fig 1). Therefore, native sickle ghost membranes and
sickle ghost membranes pretreated with 0.5 mmol/L L1 were
used for the following experimental purposes.
Oxidation of HbS to metHbS (Fig 2). The primary objective of the current study was to determine whether the pathologic deposits of free iron on sickle membranes promote
oxidation of oxyHbS to metHbS. To address this issue we
incubated oxyHbS with red cell membranes for up to 8 hours
and found that native sickle membranes provoked oxidative
conversion of oxyHbS to metHbS, and they did so to a
significantly greater extent than did native normal membranes. Since the membranes were incubated with -15.5
pmol/L Hb (62 pmol/L heme), the data depicted in Fig 2
represent spontaneous and membrane-mediated conversion
of approximately 15% and 20%, respectively, of Hb to
metHb at the 4-hour time point, for example. Significantly,
this exaggerated Hb-oxidizing effect of sickle membranes
was completely abrogated by prior removal of their free iron
deposits. Indeed, metHbS formation by sickle membranes
stripped of free iron was identical to that generated by normal
membranes (which do not have detectable free iron deposits'.').
HbS without membranes
N
f
7.5
-
N+ L1
vi
S
.-c
sB
S+Ll
5-
f
5
2
2.5
-
04
8
Hours of incubation
Fig 2. Methemoglobin S content (mean ? SD for 3 to 6 experiments) in supernatants from normal and sickle membranes (N and
S,respectively)and from iron-strippednormal and sickle membranes
(N L1 and S L1) incubated with 1 mg/mL (-15.5 pmollL) HbS
for 4 to 8 hours. Parallel incubation of HbS in the absence of membranes also is indicated.For each timepoint, and +* indicatestatistically significant differences between normal versus sickle membranes and between sickle versus iron-stripped sickle membranes,
respectively.
+
+
From www.bloodjournal.org by guest on June 18, 2017. For personal use only.
SHALEV AND HEBBEL
3950
4 1
f 3 N
T
N+LI
3 1
o s
0
m
L-T
0
4
a
Hours of incubation
Fig 3. Heme content (mean 5 SD for 3 to 6 experiments)of native
normal and sickle RBC membranes (N and S, respectively1 and of
iron-stripped normal and sickle RBC membranes (N + L1 and S + L1,
respectively) following incubation with HbS for up to 8 hours. For
each timepoint, * indicates statistically significant difference between normal versus sickle membranes, ** indicates statistical significance between native sickle membranes versus iron-stripped
sickle membranes, and # indicates statisticallysignificant differences
between iron-strippednormal membranesversus iron-strippedsickle
membranes.
Inclusion of catalase ( I UlmL) during incubation of HbS
with sickle membranes did not affect metHbS formation
(3 experiments; data not shown). Likewise. inclusion of a
high-affinity. iron-inactivating chelator (DTPA. diethylenetriaminepentaacetic acid) that is incapable of removing membrane-associated free iron’‘ did not prevent the metHb-inducing effect of native sickle membranes (data not shown).
Deposition of lierne on RBC membmnes (Fig 3). As previously observed,’’ the heme content of the starting native
sickle membranes was substantially higher than that of native
normal membranes. During incubation with HbS, the native
sickle membranes promoted nearly twice as much additional
heme deposition as did native normal membranes. This new
heme accumulation was accompanied by additional globin
deposition, as documented by polyacrylamide gel electrophoresis (data not shown). Signiticantly, as was the case for
formation of metHb in solution (above), the degree of heme
accumulation by sickle membranes was diminished by prior
removal of membrane iron. It is interesting to note, though,
that whereas stripping free iron from sickle membranes
“normalized” their methemoglobinogenic effect (so that it
was not significantly different from that of normal membranes), it only partially corrected (by about 30%) the excess
membrane heme accumulation promoted by sickle ghosts.
DISCUSSION
The abnormal deposition of hemichrome on the inner aspect of the sickle cell membrane is believed to be a proximate
“trigger” leading to premature cell demise. As noted earlier.
several mechanisms have been proposed to underlie this oxidative denaturation event, including hemoglobin auto-oxidation in solution” and the denaturing effect of membrane
lipid on cytosolic hemoglobin,” In the present studies we
provide evidence supporting the existence of an additional
mechanism: that free iron associated with pathologic red cell
membranes can catalyze conversion of soluble oxyhemoglobin to methemoglobin. Although only sickle membranes
were used in the present study, these results would be applicable to the iron-laden membranes of thalassemic cells as
well.
The exaggerated influence of the sickle membranes in
promoting methemoglobin conversion in these studies is
completely accounted for by their content of free iron. Although oxidizing oxygen species resulting from subsequent
interaction between reduced iron and oxygen also could. in
theory. contribute to methemoglobin formation. this appears
not to be the case since inclusion of catalase had no discemible effect on degree of methemoglobin formation. The enhanced promotion of metHb formation also is not explained
by release of iron from membrane into solution, since free
iron is associated with the sickle red cell membrane with
extremely high affinityz3and since presence of DTPA (which
chelates iron in solution with very high affinity but is unable
to remove iron from the membranez3)was found to have no
effect on methemoglobin formation catalyzed by membranes. On the other hand. it is entirely possible that free
iron liberated by the heme destruction that can accompany
Hb denaturation in a membrane environment” might readd some free iron to the iron-stripped membranes, thus
diminishing somewhat the observable difference between
native and stripped membranes.
Two aspects of our experimental design, selected for technical and practical reasons. theoretically might raise concem
whether our results are strictly applicable to red cell physiology. First, we used hypotonic buffer (having 5 mmolL KCI)
for co-incubation of Hb and membranes. conditions that
might accelerate Hbhembrane interaction. for example.’’
To address this, we performed limited experiments using a
higher-salt buffer (having 100 mmol/L KCI). This condition,
in fact, decreased the rate of metHb formation by about
25%. but it neither abrogated apparent membrane-dependent
metHb formation nor altered the protective effect of ironremoval in this regard. Thus, our conclusions remain valid
for physiologic salt concentrations. Second. the present studies used rather dilute Hb solutions ( 1 mg/mL). so it might
be questioned whether the accompanying relative increase
in Hb dimerization might yield results that are misleading
vis-a-vis the physiologic environment of the red cell. However, calculations indicate that under physiologic Hb concentrations the absolute amount of dimer actually would be
37-fold higher than reflected under our experimental constraints,* so our results should be quite relevant to physiologic Hb concentrations. While it is not technically possible
to conduct such experiments using actual physiologic Hb
and membrane concentrations, we were able to perform an
experiment using a physiologic ratio of Hb to membrane
(30 mg/mL Hb and I mg/mL Hb) in the higher-salt buffer
From www.bloodjournal.org by guest on June 18, 2017. For personal use only.
3951
MetHb FORMATION CATALYZED BY MEMBRANE IRON
previously described. At this 30-fold higher Hb concentration, the amount of observed metHb formation was lower,
as expected, but the effect of L1 is unchanged: 3.9% membrane-dependent metHb formation at 6 hours, with only
1.9%for iron-stripped membranes, the latter being the same
as the spontaneous rate of metHb formation in membranefree solution. This strongly argues that our results also are
relevant to physiologic Hb concentrations.
Companion studies similarly identified an exaggerated effect of sickle membranes on actual deposition of additional
heme and globin onto the red cell membrane. Since this
material does not wash off these membranes it must be denatured hemoglobin, presumably hemichrome. Its accumulation, too, is clearly iron-mediated, although the excess influence of sickle membranes over normal membranes in this
regard is only partially (about 30%) ascribable to membranefree iron content. Thus, even the iron-stripped sickle membrane has significantly greater influence on hemichrome deposition than does the normal RBC membrane.
The discrepancy between the degree of influence of the
free iron on methemoglobin formation on the one hand and
on hemichrome deposition on the other is explained, in part,
by the fact that the methemoglobin formation experiments
examined formation of methemoglobin only in solution
while the hemichrome experiments documented what
becomes deposited on the membrane. The exaggerated influence of even iron-stripped sickle membranes on actual
hemoglobin denaturation could derive if the hemoglobindenaturing effect of membrane lipid12 is enhanced for sickle
cells. This, in turn, might be accounted for by the topographical abnormalities of these membranes, whereby membrane
proteins tend to be clumped together,’ which presumably
could allow some areas of lipid to become more exposed.
Alternatively, the very presence on sickle membranes of
excess denatured hemoglobin, which is somewhat hydrophobic in character, may itself be promotive of further surface
interaction with soluble hemoglobin. In this regard, however,
it has been reported that these abnormal membrane hemoglobin aggregates actually are covered with ph~spholipid.’~
If
this is true, this would certainly be expected to have a denaturation-promoting effect.
In any case, our studies indicate that membrane free iron
can exert a promotive effect on oxidation of soluble oxyhemoglobin and thereby facilitate deposition of hemichrome
on the red cell membrane. The use of L1 in these in vitro
studies to strip chelatable iron from the membranes punctuates its potential efficacy in vivo since this drug can penetrate
membranes of intact red cells and remove iron therefrom.
Indeed, the present results establish the basis for our recent
* Our present experimental conditions using 1 mg/mL Hb would
provide 12.5% dimer, compared with 0.008% at 30 g/dL Hb. However, these values correspond to absolute dimer concentrations of
1.9 x
m o m for our dilute solution versus 71.3 X
mom
for the physiologic Hb concentration. This calculation assumes that
the tetramer-dimer equilibrium can be expressed by D2R = K, where
m o m at 37.0”C. T = W4-D/2, H = [heme], D =
K = 1.1 X
[dimer], and T = [tetramer].”
observation that amount of red cell membrane hemichrome
diminishes in thalassemic humans treated with L1.”
ACKNOWLEDGMENT
We thank Zmira Taubext for assistance in manuscript preparation.
REFERENCES
1. Hebbel Rp: Membrane associated iron, in Embury SH, Hebbel
RP, Mohandas N, Steinberg MH (eds): Sickle Cell Disease: Basic
Principles and Clinical Practice. New York, Raven Press, 1994, p
163
2. Brugnara C, Bunn HF, Tosteson DC: Regulation of erythrocyte
cation and water content in sickle cell anemia. Science 232:388,
1986
3. Sugihara T, Hebbel Rp: Exaggerated cation leak from oxygenated sickle red blood cells during deformation: Evidence for a unique
leak pathway. Blood 80:2374, 1992
4. Rank BH, Moyer NL, Hebbel RP: Vesiculation of sickle erythrocytes during thermal stress. Blood 72:1060, 1988
5. Hebbel RP, Leung A, Mohandas N: Oxidation-induced changes
in microrheologic properties of the red blood cell membrane. Blood
76:1015, 1990
6. Low PS, Waugh SM, Zinke K, Drenckhahn D: The role of
hemoglobin denaturation and band 3 clustering in red blood cell
aging. Science 227:531, 1985
7. Waugh SM, Willardson BM, Kannan R, Labotka RJ, Low PS:
Heinz bodies induce clustering of band 3, glycophorin, and ankyrin
in sickle cell erythrocytes. J Clin Invest 78:1155, 1986
8. Schliiter K, Drenckhahn D: Co-clustering of denatured hemoglobin with band 3: Its role in binding of autoantibodies against
band 3 to abnormal and aged erythrocytes. Proc Natl Acad Sci USA
83:6137, 1986
9. Turrini F, Arese P, Yuan J, Low PS: Clustering of integral
membrane proteins of the human erythrocyte membrane stimulates
autologous IgG binding, complement deposition, and phagocytosis.
J Biol Chem 266:23611, 1991
10. Mohandas N, Hebbel Rp: Pathogenesis of hemolytic anemia,
in Embury SH, Hebbel RP, Mohandas N, Steinberg MH (eds): Sickle
Cell Disease: Basic Principles and Clinical Practice. New York,
Raven Press, 1994, p 327
11. Hebbel RP, Morgan WT, Eaton JW, Hedlund BE: Accelerated
auto-oxidation and heme loss due to instability of sickle hemoglobin.
Proc Natl Acad Sci USA 85:237, 1988
12. Marva E, Hebbel RP: Denaturing interaction between sickle
hemoglobin and phosphatidylserine liposomes. Blood 83:242, 1994
13. Eguchi LA, Saltman P: The aerobic reduction of Fe(II1) complexes by hemoglobin and myoglobin. J Biol Chem 259:14337, 1984
14. Eguchi LA, Saltman P Kinetics and mechanisms of metal
reduction by hemoglobin. 1. Reduction of iron(II1) complexes. Inorg
Chem 26:3665, 1987
15. Kuross SA, Hebbel R P Nonheme iron in sickle erythrocyte
membranes: Association with phospholipids and potential role in
lipid peroxidation. Blood 72:1278, 1988
16. Repka T, Shalev 0,Reddy R, Yuan J, Abrahamov A, Rachmilewitz EA, Low PS, Hebbel RP: Nonrandom association of free iron
with membranes of sickle and P-thalassemic erythrocytes. Blood
82:3204, 1993
17. Hartley A, Rice-Evans C: The nature of membrane-bound
iron-species involved in radical-mediated damage to sickle erythrocytes. Biochem Soc Trans 17:488, 1989
18. Hebbel RP, Eaton JW, Balasingam M, Steinberg MH: Spontaneous oxygen radical generation by sickle erythrocytes. J Clin Invest
70:1253, 1982
19. Repka T, Hebbel R P Hydroxyl radical formation by sickle
From www.bloodjournal.org by guest on June 18, 2017. For personal use only.
3952
erythrocyte membranes: Role of pathologic iron deposits and cytoplasmic reducing agents. Blood 78:2753, 1991
20. Shalev 0, Repka T, Goldfarb A, Grinberg L, Abrahamov A,
Olivieri NF, Rachmilewitz EA, Hebbel RP: Deferiprone (LI) chelates pathologic iron deposits from membranes of intact thalassemic
and sickle RBC both in vitro and in vivo. Blood 86:2008, 1995
21. Winterboum CC: Reactions of superoxide with hemoglobin,
in Greenwald RA (ed): CRC Handbook of Methods for Oxygen
Radical Research. Boca Raton, CRC Press, 1985, p 137
SHALEV AND HEBBEL
22. Kuross SA, Rank BH, Hebbel RP: Excess heme in sickle
erythrocyte inside-out membranes: Possible role in thiol oxidation.
Blood 71:876, 1988
23. Shalev 0, Hebbel RP: Extremely high avidity association
of fe(II1) with the sickle red cell membrane. Blood 1996 (in
press)
24. Liu SC, Nichols PE, Yi S, Ballas SK, Yacono PW, Golan
DE, Palek J: Phospholipid labeling of Heinz bodies in sickle cell
erythrocytes. Blood 80:78a, 1992 (suppl)
From www.bloodjournal.org by guest on June 18, 2017. For personal use only.
1996 87: 3948-3952
Catalysis of soluble hemoglobin oxidation by free iron on sickle red
cell membranes
O Shalev and RP Hebbel
Updated information and services can be found at:
http://www.bloodjournal.org/content/87/9/3948.full.html
Articles on similar topics can be found in the following Blood collections
Information about reproducing this article in parts or in its entirety may be found online at:
http://www.bloodjournal.org/site/misc/rights.xhtml#repub_requests
Information about ordering reprints may be found online at:
http://www.bloodjournal.org/site/misc/rights.xhtml#reprints
Information about subscriptions and ASH membership may be found online at:
http://www.bloodjournal.org/site/subscriptions/index.xhtml
Blood (print ISSN 0006-4971, online ISSN 1528-0020), is published weekly by the American
Society of Hematology, 2021 L St, NW, Suite 900, Washington DC 20036.
Copyright 2011 by The American Society of Hematology; all rights reserved.