From www.bloodjournal.org by guest on June 17, 2017. For personal use only.
Permeability Characteristics of Deoxygenated Sickle Cells
By Margaret R. Clark and Mary E. Rossi
This study investigated the effect of acute deoxygenation
on membrane permeability characteristics of sickle cells.
Measured fluxes of Na+ and K+ in ouabain-inhibitedcells, of
chlorideand sulfate exchangein 4,4'-diisothiocyanostilbene
2.2'-disulfonate (DIDSI-inhibited and untreated cells, and of
erythritol, mannitol, and arabinose in cytochalasin B-inhibited cells indicatedthat a deoxygenation-inducedpermeability change occurred in sickle cells only for cations and
chloride. Monovalent cation permeabilities increased fivefold, and chloride influx into DIDS treated cells was
enhanced nearly threefold on sickle cell deoxygenation. In
contrast, no detectable increase in permeability to the
other solutes was found. To gain perspective on these
findings, similar measurements were performed in normal
cells treated with diamide, an agent shown by others to
induce a coupled increase in membrane permeability and
phospholipid translocation, reminiscent of deoxygenationinduced changes in sickle cells. Although the increase in
cation permeability was no greater than that in sickled
cells, treatment with 2 mmol/L diamide also produced a
twofold increase in the first order rate constantsfor sulfate
exchange and mannitol efflux, indicating a relatively nonselective permeability increase that permitted flux of larger
solutes than in the case of deoxygenated sickle cells. These
results suggest that the deoxygenation of sickle cells
induces a permeability increase that is relatively insensitive
to charge, but is restrictive with respect to solute size.
0 1990 by The American Society of Hematology.
T
sickle cells was the result of mechanical disruption of bilayer
structure, the cells would exhibit a nonselective increase in
membrane permeability to a variety of solutes, similar to that
induced by diamide treatment of normal RBCs.
An alternative explanation for increased monovalent cation fluxes in sickled cells could be a sickling-induced activation of specific cation transport pathways. Studies have been
performed by others in an effort to detect sickling-associated
abnormalities in Na+/K+ cotransport, the volume-sensitive
KCL transport system, or Na+/K+ countertransp~rt>'~~'~
but no effect of deoxygenation on flux through these routes in
sickle cells was found. However, Joiner and Dew did find that
an inhibitor of anion exchange 4,4'-diisothiocyanostilbene2,2'-disulfonate (DIDS) blocked the increase in cation fluxes
in deoxygenated sickle cells, and they originally suggested
that the increased cation fluxes might possibly be mediated
by an altered form of the anion transport protein."
In an effort to test whether the sickling-inducedmembrane
permeability defect represents a nonselective breakdown of
the membrane barrier, we have measured the effect of
deoxygenation on the permeability of sickle cells to a variety
of solutes, including the sulfate and chloride anions and three
neutral solutes of different sizes. Direct comparison of
permeability changes in sickled cells with the observations of
Deuticke et a1 on diamide-treated cells seemed inappropriate, because those experiments were performed at levels of
oxidant treatment that gave larger changes in permeability
and lipid translocation than are seen in deoxygenated sickle
HE PRIMARY DEFECT in sickle cell disease is the
abnormal structure of the hemoglobin S (HbS) molecule, whose polymerization at low oxygen tension causes the
major clinical manifestation, vasoocclusion. Nevertheless, it
is thought that some of the factors that initiate vasoocclusive
events and promote rapid destruction of sickle cells may be
related to effects of Hb S on red blood cell (RBC) membrane
structure and function.'-3Most of the investigation of membrane abnormalities in sickle cells has focused on long-term,
cumulative damage such as that seen in irreversibly sickled
cells (ISC). However, for a complete understanding of the
nature and pathophysiologic implications of membrane damage in sickle cells, it is important to define each stage of the
process. From this perspective, we have begun to investigate
the nature of the early changes in RBC membrane permeability that are associated with sickling. More than 30 years ago,
Tosteson et a1 reported that acute deoxygenation of sickle
cells produced an increase in membrane permeability to
alkali cation^.^.' Subsequent work showed that in addition to
monovalent cations, Ca2+also could enter sickle cells at an
increased rate when they were deoxygenatedP The relationship of these observations to the extreme dehydration of a
subpopulation of sickle cells has been debated,'-'' but as yet
there is insufficient information about the nature of the
permeability change to determine how it fits into the development of sickle cell membrane injury. We previously postulated that the extreme distortion of the membrane at the tips
of spicules on sickled cells might result in a breach in the
membrane permeability barrier. In accord with this hypothesis, we found that the magnitude of the sickling-induced leak
for Na+ and K+ varied in parallel with the extent of cell
distortion." One might expect that such a mechanically
induced leak would show relatively little selectivity for
different types of solutes. Such a lack of specificity has been
observed by Deuticke et all2 and Schwister and Deuticke13
for permeability defects induced by treating normal cells
with oxidizing agents or electric discharge. Interestingly,
these investigators also found that the permeability increase
was accompanied by an increase in the rate of translocation
of phospholipids across the membrane bilayer, qualitatively
similar to the increased transbilayer movement that is
induced by deoxygenation of sickle cells.I4 Thus, we hypothesized that if the permeability change in acutely deoxygenated
Blood, Vol76, No 10 (November 15). 1990: pp 2139-2145
From the Department of Laboratory Medicine and Cancer
Research Institute. University of California, San Francisco.
Submitted February 20.1990; accepted July 27, 1990.
Supported in part by USPHS Grants HL20985 and AM32095.
Preliminary experiments in this study were reported in Blood
70:60a, 1987 (suppl 1).
Address reprint requests io Margaret R . Clark, PhD, Box 01 28,
Cancer Research Instiiute. University of California, San Francisco.
CA 94143.
Thepublication cosis of this article were defrayed in part by page
charge payment. This article must therefore be hereby marked
"advertisement" in accordancewith 18 U.S.C. section 1734 solely to
indicate this fact.
0 1990 by The American Society of Hematology.
0006-4971/90/761O-oO28$3.O0/0
2139
From www.bloodjournal.org by guest on June 17, 2017. For personal use only.
2140
CLARK AND ROSS1
cells. Therefore, we performed additional flux measurements, using normal cells treated with lower concentrations
of diamide to provide a model of nonselective permeability
increase for comparison with the sickle cells.
MATERIALS AND METHODS
Blood samples. Blood samples were obtained from sickle cell
patients at the Northern California Comprehensive Sickle Cell
Center, San Francisco General Hospital, after obtaining informed
consent as approved by the Committee on Human Research of the
University of California at San Francisco. All blood samples were
from patients who were homozygous for Hb S, as determined by
thin-layer isoelectric focusing and citrate agar electrophoresis.The a
globin genotype was not determined.
Blood samples from patients and control subjects were drawn into
acid citrate dextrose (ACD; Becton Dickinson, Rutherford, NJ) or
CPD-A (Fenwal Laboratories, Deerfield, IL) anticoagulant and
were normally used the day after they were obtained, although on
occasion they were used after storage for 3 to 5 days. To obtain a
more homogeneous population of sickle cells, arabinogalactan
gradients'* were used to isolate an intermediate density cell population from which the youngest reticulocytes and most of the ISC had
been removed. The cells collected usually had densities of 1.077 to
1.103 g/mL, but occasionally a narrower density range was selected
when a sufficient number of cells for the experiment could be
isolated.
Diamide treatment of normalcells. At the 5 mmol/L concentration used by Deuticke et a1,I9 diamide produced much greater
increases in the rates of phospholipid translocation and cation flux
than those induced by deoxygenation of sickle
Thus, in the
current studies, a lower concentration was used, to give a more
meaningful comparison of the effects of these two manipulations. A
concentration of 2 mmol/L diamide was chosen, because treatment
of normal cells at this level gave an increase in K+ efflux comparable
with that of deoxygenated sickle cells. The rate of lipid translocation
at this level has been shown to be approximately three times that in
deoxygenated sickle cell^.'^.'^ For diamide treatment, the cells were
incubated for 60 minutes at 37OC, in 90 mmol/L KCI, 45 mmol/L
NaCI, 44 mmol/L sucrose, and 10 mmol/L Na phosphate, pH 8.0.
They were then washed and prepared as described below for the flux
measurements.
General method for flux experiments. Before initiation of all
efflux experiments involving sickle cells, paired samples of preloaded
cells were placed on ice, and one of each pair was deoxygenated by
equilibration of the cell suspension with humidified nitrogen gas.
Because at 0°C sickling does not occur, and baseline permeability is
low, this prevented a large contribution to the measured fluxes
during the relatively slow deoxygenation process. Once the samples
were deoxygenated, they were sealed, and all samples were rapidly
brought to the temperature at which the flux was to be measured,
usually 37OC. For influx experiments, samples were first deoxygenated at room temperature. Then they were warmed to 37OC for at
least 5 minutes, the isotope was added (under a nitrogen atmosphere
for the deoxygenated samples), and the zero time sample removed
immediately after mixing. For all flux experiments, samples were
removed at intervals depending on the rate of the flux process to
provide at least four sequential determinations. Deoxygenated samples were removed under nitrogen, and the incubation vessels
resealed after each sampling. Experiments using diamide-treated
normal cells were performed without prior deoxygenation.
For efflux measurements, suspension samples were centrifuged,
and the supernatant analyzed for the effluxing solute. For influx
measurements, cells were separated from the medium containing the
test solute and assayed for its accumulation. In all but the sulfate flux
experiments, measurements of the hematocrit and Hb concentration
(measured as the cyanomethemoglobin complex) on an initial cell
sample were used in conjunction with measurements of the Hb
concentration in the incubation mixture to determine the hematocrit
of the incubation mixture, thereby permitting expression of measured fluxeson the basis of original cell volume. Efflux measurements
were corrected for hemolysis using measured values of the Hb
concentration in the supernatant. In the sulfate flux experiments, the
sulfate in the medium interfered with measurement of Hb as the
cyanomethemoglobin complex, and the absorbance of the supernatant at the Soret band (415 nm) was used to determine the hemolysis
correction. The hematocrit during incubation was determined from a
RBC count using a Coulter Model B cell counter (Coulter Electronics, Hialeah, FL) and the RBC count and hematocrit of the initial
washed cell suspension.
Sodium and potassiumpuxes. The net fluxes of Na+ and K+
were determined during incubation of the cells for 4.5 hours, 37OC,
at approximately 4% hematocrit in isotonic NaCl buffered with 10
mmol/L Na phosphate, pH 7.4. Cell samples were washed three
times in ice-cold, isotonic Tris buffered (10 mmol/L, pH 7.4) MgCI,
and then assayed for Na+ and K+ content by flame photometry. In
addition, samples of the suspending medium were assayed for K+ to
provide an alternate measurement of K+ efflux. To determine the
effect of DIDS on the cation fluxes, paired samples were incubated
with and without the addition of 100 pmol/L DIDS (Sigma
Chemicals, St Louis, MO).In all cation flux experiments presented
here, the DIDS was present during the flux measurements. Another
approach was tried in which the cells were pretreated with DIDS and
then washed before proceeding with the flux measurements, but this
method caused increased hemolysis in the DIDS-treated cells during
subsequent incubation and was therefore abandoned.
Sulfate fluxes. Sulfate flux was determined as the efflux of
"SO:- (New England Nuclear, Natick, MA) from cells that had
been previously loaded with the radiolabeled anion. The experiments
were performed after the cells had been incubated at 37OC for at
least 2 hours in medium containing 20 mmol/L sulfate, which was
found to be adequate for sulfate equilibration in the absence of
DIDS; thus, the subsequently measured sulfate fluxes in the same
medium (0.13 mol/L sucrose, 40 mmol/L NaCI, 20 mmol/L
Na,S04, and 20 mmol/L Tris-HC1, pH 7.4 at 37OC) represented the
rates of sulfate exchange, rather than net flux. The isotope concentration was adjusted to give a specific activity of approximately 0.5
mCi/mmol/L of sulfate. For the efflux experiments, the isotope was
added during the last 2 hours of a 4-hour preincubation. Before
measuring efflux of the label from preloaded cells, half of each cell
sample was treated with DIDS (50 or 100 pmol/L) to inhibit
carrier-mediated anion exchange, and the cells were then washed in
ice-cold incubation medium to remove extracellular isotope. The
intracellular sulfate concentration at the beginning of the flux
experiments was determined from the radioactivity of the washed,
loaded cells, together with the specific activity of the loading
medium. Subsequent loss of labeled sulfate from the cells was then
quantitated from the increase of isotope counts in the supernatant of
centrifuged samples and the hematocrit of the incubation suspension. Sulfate efflux was measured over a 1-hour period of incubation
at 37OC.
Because the effluxmeasurements represented the redistribution of
the isotopically labeled anion in cell suspensions that had already
been equilibrated with sulfate, the rates were expected to follow
first-order kinetics. Therefore, the data were analyzed by plotting
the natural log of the ratio of intracellular sulfate concentration to
the initial concentration versus time. This procedure compensates for
the variation in intracellular sulfate concentration after cell loading.
In particular, diamide-treated cells never retained as much sulfate as
From www.bloodjournal.org by guest on June 17, 2017. For personal use only.
2141
PERMEABILITY CHARACTERISTICS OF SICKLE CELLS
untreated cells while they were being washed after the loading
procedure.
In addition to the efflux experiments, we also performed a series of
experiments in which we determined the rate of influx of 35S0,2-into
cells preequilibrated in nonradioisotopic sulfate solutions. The precision in measuring uptake of the slowly equilibrating sulfate into
DIDS-treated cells is inherently much lower than that of the efflux
experiments, so a modest increase in influx would not have been
detected. This approach was therefore abandoned.
Chloride inJuxes. Because chloride fluxes in cells with a functional anion transport system are too rapid to be measured at 37OC,
we performed influx e;periments for DIDS-inhibited cells. Washed
sickle cells were pretreated with 100 amol/L DIDS in BSKG
(isotonic phosphate-buffered saline containing 10 mmol/L sodium
phosphate, pH 7.4, 5 mmol/L potassium chloride, and 11 mmol/L
glucose) for 30 minutes at room temperature, and one of each paired
sample was deoxygenated. The flux experiment was initiated after
placing the samples in a 37OC water bath for a 5-minute equilibration period. Then radiolabeled chloride was added, and four consecutive samples were taken, to provide 0-, 5-, lo-, and 20-minute values
for chloride uptake. The duplicate aliquots were immediately introduced into ice-cold incubation medium layered over dibutylphthalate and centrifuged to separate the cells from the e&racellular
isotope.2' The cell pellet was then lysed and a sample taken for Hb
determination. Aliquots of the remaining lysate were then treated
with 10% trichloroacetic acid to precipitate protein, centrifuged, and
aliquots of supernatant counted to determine the cell content of
labeled chloride.
Neutral species efluxes. Three solutes were used to determine
the effect of deoxygenation on permeability to neutral species. They
were I4C-labeled erythritol (Amersham Corp, Arlington Heights,
IL), mannitol, and arabinose (ICN Biomedicals, Inc, Costa Mesa,
CA). For these experiments, cells were preloaded under room air by
incubation in 4 mmol/L solute (0.5 pCi/mL) at 37OC for 2 hours, in
medium containing 100 mmol/L KCI, 36 mmol/L NaC1, 6.5
mmol/L sodium phosphate, pH 7.2, and 44 mmol/L sucrose, as
described by Deuticke et a1.I2 The cells were then washed and
resuspended in the same medium, to which cytochalasin B (10
pmol/L) was added to inhibit facilitated diffusion. Sickle cell
samples were maintained on ice during the deoxygenation period,
then the samples were warmed to 37OC, and efflux rates were
determined by measuring the accumulation of radioactivity in the
supernatant of centrifuged samples. In experiments using erythritol
and mannitol, rate constants were calculated from the apparent first
order efflux over the first 40 to 60 minutes. For arabinose efflux in the
diamide experiments, the efflux during the first 20-minute time
interval did not fit into a first order process with subsequent points.
Therefore, first order rate constants for all arabinose experiments
were calculated for the data between 20 and 120 minutes, which did
give a good fit to first order kinetics. To be sure that the effect of
diamide on mannitol efflux from normal cells was not due to the
presence of contaminating white blood cells, one experiment was
performed in which leukocytes and platelets were rigorously removed by filtration of the whole blood through cellulose.22
Analysis of results. A Statview program (Brain Power, Inc,
Calabasas, CA) on a Macintosh I1 computer (Apple Computers,
Inc, Cupertino, CA) was used to calculate the mean and standard
errors for replicated experiments, and to perform paired and
unpaired t-tests as appropriate. A Cricket Graph program (Cricket
Software, Malvern, PA) was used to obtain first order rate constants
for efflux measurements by performing linear regression of the log of
the intracellular concentration of the effluxing solute versus time.
RESULTS
Our initial experiments involved the measurement of
''SO,2- efflux from sickle cells in the deoxygenated versus
oxygenated state. The first few experiments showed a substantial deoxygenation-induced increase in sulfate efflux from
DIDS-treated
even after correction of the total
supernatant radioactivity for the slight amount of hemolysis
that occurred. In marked contrast to these initial results,
subsequent experiments using a new lot of DIDS failed to
show any deoxygenation-induced increase in sulfate efflux
from DIDS-treated or untreated cells (Figs 1 and 2, Table
1). In the analysis of these experiments, the total sulfate
efflux was corrected for hemolysis, which ranged from 0.4%
to 3% total, and 0.2% to 1.5% net increase during the efflux
experiments.
In contrast to the lack of effect of deoxygenation on sulfate
efflux from DIDS-treated sickle cells, treatment of normal
cells with 2 mmol/L diamide caused a twofold increase in
DIDS-resistant sulfate efflux (Fig 2B, Table 1). No effect of
2.0 -
1.0
Oxygenated
-
Deoxygenated
0.0
u.0
0.2
0.4
0.8
0.6
1 .o
Time - hours
f
Diamide Treated
"-1
/
/
I
Untreated Control
V."
0.0
0.2
0.4
T
0.6
0.8
1.o
Time - hours
Fig 1. Sulfate exchange effluxes in the absence of DIDS. (AI
Oxygenated (0)
and deoxygenated (HI sickle cells. (SI Normal
control RBCs untreated ( 0 )and treated with 2 mmol/L diamide
(HI. In both graphs, the natural log of the ratios of the intracellular
radioisotope concentration to its initial value are plotted versus
time. Values are the means from four experiments; error bars
indicate the standard error of the mean (SEMI at each time point.
Indicated ratio on the ordinate should be inverted for Figs 1 end 2.
From www.bloodjournal.org by guest on June 17, 2017. For personal use only.
2142
CLARK AND ROSS1
A
70
-
0.10
z=
0
Y
rI
T
60 -
0.08 -
0.06 -
Oxygenated
50
-
40
.
I
+ oxy
* deoxy
a
c
=
m
3
cn
Y
c
-
0.02
0
0.000.0
0.2
.
0.4
0.60
Deoxygenated
0.8
14.o
~
30
Time - hours
B
-
0.10
20
0
5
10
15
20
Diamide Treated
-
T
Time - min
h
0
a,
v
0.08 -
Fig 3. Effect of deoxygenation on chloride uptake by DIDS
treated sickle cells. Oxygenated ( 0 )and deoxygenated (m) sickle
cells were treated with 250 pmol/L DIDS, incubated at 37°C. and
the accumulation of "Cl determined. Values plotted are the means
from seven experiments. and the error bars represent the SEM.
c
=
m
-.
2
%
0.06
-
0.04
-
c
Y
a,
5
=
3
cn
Y
-c
0.02 -
0.00
0.0
Untreated Control
0.2
0.4
0.6
0.8
1 .o
Time - hours
Fig 2. Sulfate exchange effluxes in the presence of DIDS. (A)
Oxygenated ( 0 )and deoxygenated (W) sickle cells treated with 50
pmol/L DIDS. (B) Normal control RBCs untreated (0)
and treated
with 2 mmol/L diamide (W) both treated with 50 pmol/L DIDS.
diamide could be detected in cells that were not treated with
DIDS (Fig 1B).
Although deoxygenation had no effect on efflux of sulfate
from sickle cells, it produced a consistent increase in the rate
of 36C1accumulation by sickle cells that had been treated
with DIDS to suppress active anion exchange (Fig 3). The
magnitude of the effect varied from experiment to experiment, giving an average increase of between twofold and
threefold in the rate of chloride uptake over a 20-minute
period (.005 < P 5 .01). In only one of seven experiments
was there no significant difference in chloride uptake by
deoxygenated as compared with oxygenated cells.
The effect of DIDS on the deoxygenation-induced cation
leak was also studied. As reported by Joiner,24DIDS reduced
both K+ efflux and N a + influx in deoxygenated sickle cells to
approximately half the values in the absence of DIDS (Fig
4). The difference in fluxes between oxygenated and deoxygenated cells was decreased by an even larger amount,
because DIDS treatment actually increased the Na+ and K+
fluxes in oxygenated sickle cells. It should be noted that early
experiments, conducted at the same time as the preliminary
sulfate efflux experiments that had suggested a deoxygenation-dependent sulfate leak, showed a much smaller effect
of DIDS on the cation fluxes as well. Presumably the DIDS
used in these early experiments was not fully active.
Cation fluxes were also measured for normal cells treated
with 2 mmol/L diamide. Diamide treatment induced a large
increase in cation permeability, comparable with the permeability increase seen on deoxygenation of sickle cells (Fig 4).
However, treatment with DIDS did not reduce the enhancement of K+ and N a + fluxes. In fact, when the effect of DIDS
on fluxes from diamide-treated and untreated cells was
analyzed using a t-test for paired samples (with and without
DIDS), only the N a + influx into diamide-treated cells was
significantly different from the flux in DIDS-free cells, and it
was increased (.01 < P 5 .025). The Pvalues for the effect of
Table 1. Summary of Rate Constants for DIDS-Resistant Exchange of Sulfate in Sickle and Diamide-Treated Normal RBCs
Normal Cells
Sickle Cells
Rate constant-h-'
on,
DEOW
Control
0.034 (0.013)
0.035 (0.010)
0.042 (0.006)
2 mmol/L Diamide
0.089 (0.013).
First order rate constants were calculated by linear regression analysis of the data from four experiments with each type of cells. All cells were treated
with 5 0 pmol/L DIDS.
* . 0 0 5 < P 5 . 0 1 for diamide-treated as compared with control normal cells.
From www.bloodjournal.org by guest on June 17, 2017. For personal use only.
PERMEABILITY CHARACTERISTICS OF SICKLE CELLS
2143
Table 2. First Order Rate Constants for the Efflux of Neutral
Species From Oxygenated and Deoxygenated Sickle Cells (SS)
T
c
x
-
5
0.12
Y
.-zc
-2
s
0.08
m
(r
Arabinose
Mannitol
0.070
0.050
0.0085
0.010
0.0086
0.0085
0.045
0.050
0.056
0.016
0.013
0.018
0.0065
0.0068
0.0044
All cells were pretreated with cytochalasin B (10 pmol/L) to suppress
facilitated diffusion. Units for rate constants are min '. All experiments
were performed at different times.
0.04
0.00
SS Oxy
S S Deoxy
AA Control
AA Diamide
T
"
oxy ss
Deoxy SS
Experiment set 2
oxy ss
Deoxy SS
Normal control (one experiment)
e
Erythritol
Experiment set 1
S S Oxy
S S Deoxy
AA Conlrol
AA Diamide
Fig 4. Effect of DIDS on K ' and Na' fluxes in sickle and
diamide treated normal cells. Shown are oxygenated and deoxygenated sickle cells and normal cells untreated and pretreated with 2
mmol/L DIDS. (A) First order rate constants for K ' efflux. (W),
Cells in the absence of DIDS; (El),cells treated w i t h 250 pmol/L
DIDS. Values are the means from five experiments; bars represent
the SEM. (B) Rates of Na ' influx for cells w i t h (W) and without (0)
DIDS.
DIDS treatment for all other flux components were between
. I and .375.
To determine whether deoxygenation of sickle cells made
them more permeable to uncharged solutes, we used three
different ''C-labeled probes, erythritol, mannitol, and arabinose. The rates of flux of these molecules across the RBC
membrane vary over a broad range, reflecting their different
sizes. As shown in Table 2, the first order rate constants for
all three species were unaffected by deoxygenation of sickle
cells. Rates of mannitol efflux from both deoxygenated and
oxygenated sickle cells were higher than those observed for
normal control cells, but the number of experiments is too
small to determine whether this difference was consistent.
When the effect of diamide on the permeability of normal
cells was examined, we found that treatment with 2 mmol/L
diamide did produce an approximately twofold increase in
the efflux of mannitol (.025 c P 5 .005). There was no
detectable alteration in the permeability to erythritol or
arabinose induced by this treatment (Table 3). Because
Deuticke et al had found a greater permeability increase for
erythritol than for mannitol with diamide treatment at 5
mmol/L concentration, we compared erythritol efflux from
cells treated with 0,2, and 5 mmol/L diamide, but found no
difference (data not shown). We also considered the possibility that the increased mannitol efflux from diamide-treated
samples might be due to the presence of contaminating
leukocytes. Cells were prepared with and without prefiltration through cellulose to remove white blood cells and
platelets, and were then treated with 2 mmol/L diamide.
There was no difference in the loading of mannitol or its
subsequent efflux (data not shown). Because of the very low
permeability of the membrane to mannitol, the initial concentrations of this solute were substantially lower than those of
the other two, 0.8 mmol/L on the average, as compared with
2.5 and 2.4 mmol/L for erythritol and arabinose. However,
there was sufficient loading to observe first order efflux
kinetics over a 40-minute period.
DISCUSSION
The results of these studies indicate that the increase in
membrane permeability induced by deoxygenation of sickle
cells has some selectivity. Acutely deoxygenated sickle cells
showed the characteristic fivefold increase in ouabainresistant Na' and K' fluxes under the conditions of these
experiments, and a smaller but consistent twofold to threefold elevation in chloride influx when the anion transporter
was inhibited with DIDS. However, they showed no detectable increase in permeability to sulfate anion or to three
different neutral species, erythritol, mannitol, or arabinose.
This pattern of permeability changes was different from the
pattern of permeability changes induced by treatment of
Table 3. Effect of 2 mmol/L Diamide on Efflux of Neutral Species
From Preloaded Cells
Erythritol
Arabinose
Mannitol
~~
Untreated control 0.044 (0.001) 0.013 (0.002) 0.0024 (0.0002)
Diamibtreated 0.047 (0.003) 0.013 (0.003) 0.0040 (0.0006).
Values given are mean first order rate constants in units of min-'.
Numbers in parentheses indicate the standard error. Four experiments
each were performed to measure erythritol and mannitol fluxes; five were
performed to measure arabinose fluxes.
*.005 e P _c .05.
From www.bloodjournal.org by guest on June 17, 2017. For personal use only.
2144
CLARK AND ROSS1
normal cells with 2 mmol/L diamide. Although diamide
treatment a t this level gave the same increase in Na’ and K+
permeability observed in deoxygenated sickle cells, it also
produced a twofold increase in the rate constants for efflux of
mannitol and sulfate anion exchange in DIDS-inhibited
cells, increases that were not produced by deoxygenation of
sickle cells. Thus, it appears that the membrane defects in
sickle cells and diamide-treated normal cells are different,
despite the coupling of a permeability increase and an
increase in transbilayer movement of phospholipids in both
instances.
Joiner et a1 have also studied anion permeability in sickle
cells, but they found no evidence for an effect of deoxygenation on either sulfate or chloride
However,
the experiments they performed to determine the effect of
deoxygenation on chloride permeability at 37OC involved
indirect measurements of K+ efflux from Valinomycintreated cells. The assumption is that when K+ permeability is
enhanced by Valinomycin treatment, the flux of K+ is limited
by membrane permeability for the chloride counter-ion. A
lack of effect of deoxygenation on the K’ efflux from
Valinomycin-treated sickle cells was thus taken as an indication of a lack of effect on chloride permeability. At present, it
is not clear how these results relate to our direct observations
of an increase in chloride influxes associated with sickling.
Assuming that our observations reflect a real increase in
chloride diffusional permeability in deoxygenated sickle
cells, it suggests that the membrane leak is based primarily
on solute size, with little discrimination on the basis of
charge. This would be consistent with the observed increases
in permeability to all the alkali metal cations, Ca2+, and
chloride, which have similarly small hydrated radii, and the
absence of increased permeability to erythritol, which is
uncharged but substantially larger. The diamide-induced
leak behaves rather differently, showing an increase in
permeation of much larger solutes. It has been proposed that
it can be represented by transient pores, of a size that admits
passage of rather large solutes, albeit at a slower rate than
small solutes. We speculate that the leak induced in sickle
cells places a more stringent limit on the size of molecules
that can pass through it.
Our results on the pattern of permeability changes induced
by diamide differ somewhat from those reported by Deuticke
et al.I9 They found that the increase in permeability to
erythritol, arabinose, and mannitol varied inversely with
solute size, giving the largest magnitude of leak for erythritol. In contrast, we found increased permeability only for
mannitol, the largest solute. Although we made an effort to
use the same experimental conditions described by these
investigators, it appears that in our experiments a much
smaller change in fluxes occurred, such that the increase was
only detectable for the solute with an extremely low background flux rate.
The present results confirm the observations of Joiner that
DIDS treatment reduces the cation fluxes in deoxygenated
sickle cells.24Because DIDS, an amino reactive agent, reacts
primarily with the anion transport protein band 3, it was
originally proposed that the deoxygenation-induced cation
leak might involve an altered form of the band 3 p r ~ t e i n . ’ ~
However, subsequent work showing that the deoxygenationinduced leak and anion transport can be differentiated in
their DIDS concentration and temperature dependence suggests that the two fluxes may use different pathways.24
Interestingly, the inhibitory effect of DIDS in deoxygenated
sickled cells contrasts with its effect of enhancing cation
fluxes in oxygenated sickle cells, and in both diamide-treated
and untreated normal RBCs, suggesting a relatively specific
effect.
In initiating these experiments, we hypothesized that the
deoxygenation-induced leak was due to localized areas of
mechanical stress on the membrane overlying spicules of
polymerized H b S. Our current observations, suggesting a
charge-insensitive pathway with a stringent size limitation,
are compatible with this possibility. However, Joiner has
shown that DIDS suppresses most of the deoxygenationdependent cation permeability change without reducing
morphologic distortion. As proposed by Joiner, this may
suggest an indirect effect of membrane stretching on permeability. Ney et
have reported that deformation of sickle
cells by application of fluid shear stress also causes an
increase in K permeability, which is synergistically enhanced
under peroxidative stress by t-butylhydroperoxide. However,
this type of mechanical stress is qualitatively different from
the stress induced by polymer in the spicules. When cells are
undergoing stable, shear-induced deformation, the stress is
distributed over a large portion of the membrane surface as
the cell membrane “tank-treads” around the interior. In
contrast, in the sickled cell, stress appears to be focused on a
relatively small and constant area. This difference could be
the basis for the much larger K leak seen in deoxygenated
sickle cells as compared with that in oxygenated sickle cells
subjected to high shear stress. Finally, another recent finding
should be considered in proposing a model involving permeability changes in the spicules. Liu et a1 have demonstrated
that the membrane covering long spicules is essentially
devoid of spectrin, although it retains the band 3 protein.27
Thus, in considering possible mechanisms for inducing a
membrane permeability change, one should probably think
of an RBC membrane stripped of its skeletal underpinnings
rather than an intact RBC membrane. At present, little is
known about the influence of the membrane skeleton on
membrane permeability, and it is difficult to judge how
DIDS treatment might interact with any such influence.
ACKNOWLEDGMENT
The authors are grateful to Dr William C. Mentzer and Klara
Kleman for providing samples of sickle cells. This is paper no 101
from the MacMillan-Cargill Hematology Research Laboratory,
University of California, San Francisco.
REFERENCES
1 . Hebbel RP: Auto-oxidation and a membrane-associated ‘Fenton reagent’: A possible explanation for development of membrane
lesions in sickle erythrocytes.Clin Haematol 14:129, 1985
2. Wagner GM, Schwartz RS, Chiu DT-Y, Lubin BH: Membrane phospholipid organization and vesiculation of erythrocytes in
sickle cell anemia. Clin Haematol 14:183, 1985
From www.bloodjournal.org by guest on June 17, 2017. For personal use only.
PERMEABILITY CHARACTERISTICS OF SICKLE CELLS
3. Green GA, Kalra V K Sickling-induced binding of immunoglobulin to sickle erythrocytes. Blood 7 1:636, 1988
4. Tosteson DC, Carlsen E, Dunham ET: The effects of sickling
on ion transport: I. Effect of sickling on potassium transport. J Gen
Physiol39:3 1,1955
5. Tosteson DC: The effects of sickling on ion transport. 11. Effect
of sickling on sodium and cesium transport. J Gen Physiol 3955,
1955
6. Palek J: Red cell calcium contenband transmembrane calcium
movements in sickle cell anemia. J Lab Clin Med 89:1365,1977
7. Masys DR, Bromberg PA, Balcerzak S P Red cells shrink
during sickling. Blood 44:885, 1974
8. Fabry ME, Nagel RL: The effect of deoxygenation on red cell
density: Significance for the pathophysiology of sickle cell anemia.
Blood 601370,1982
9. Berkowitz LR, Orringer E P Passive sodium and potassium
movements in sickle erythrocytes. Am J Physiol249:C208,1985
10. Joiner CH, Platt OS, Lux S E Cation depletion by the sodium
pump in red cells with pathologic cation leaks. Sickle cells and
xerocytes. J Clin Invest 78:1487, 1986
11. Mohandas N, Clark MR, Rossi M E Association between
morphologic distortion of sickle cells and deoxygenation-induced
cation permeability increase. Blood 68:450, 1986
12. Deuticke B, Liitkemeier P, Sistemich M: Ion selectivity of
aqueous leaks induced in the erythrocyte membrane by crosslinking
of membrane proteins. Biochim Biophys Acta 775:150, 1984
13. Schwister K, Deuticke B: Formation and properties of aqueous leaks induced in human erythrocytes by electrical breakdown.
Biochim Biophys Acta 816:332, 1985
14. Mohandas N, Rossi M, Bernstein S, Ballas S, Ravindranath
Y, Wyatt J, Mentzer W: The structural organization of skeletal
proteins influences lipid translocation across erythrocyte membrane.
J Biol Chem 260:14624,1985
15. Canessa M, Fabry ME, Blumenfeld N, Spalvins A, Nagel
R L Volume dependent and NEM stimulated K+CI- transport in
sickle cells: Effect of deoxygenation. Clin Res 34:718A, 1986
2145
16. Joiner CH, Dew A, Ge DL: Deoxygenation-induced cation
fluxes in sickle cells: Relationship between net potassium efflux and
net sodium influx. Blood Cells 13:339,1988
17. Joiner CH, Dew A: Deoxy cation fluxes in sickle cells are
inhibited by DIDS. Fed Proc 45446, 1986
18. Clark MR, Unger RC, Shohet SB: Monovalent cation composition and ATP and lipid content of irreversibly sickled cells. Blood
51:1169,1978
19. Deuticke B, Poser B, Liitkemeier P, Haest CWM: Formation
of aqueous pores in the human erythrocyte membrane after oxidative
cross-linking of spectrin by diamide. Biochim Biophys Acta 73 1: 196,
1983
20. Mohandas N, Wyatt J, Me1 SF, Rossi ME, Shohet SB: Lipid
translocation across the human erythrocyte membrane. Regulatory
factors. J Biol Chem 257:6537, 1982
21. Young JD, Ellory JC: Flux measurements, in Ellory JC,
Young JD (eds): Red Cell Membranes. A Methodological A p
proach. San Diego, CA, Academic, 1982, p 119
22. Beutler E: Red cell metabolism (2nd ed). Philadelphia, PA,
Grune and Stratton, 1970, p 10
23. Clark MR, Rossi ME: Deoxygenation-induced increase in
sulfate efflux from DIDS treated sickle red cells. Blood 7060a, 1987
(SUPPI 1)
24. Joiner CH: Deoxygenation-induced cation fluxes in sickle
cells: 11. Inhibition by stilbene disulfonates. Blood 76:212, 1990
25. Joiner CH, Frolicb 0, Gunn RB: Anion transport in sickle
red cells (SSRBC). Blood 74260a, 1989 (suppl 1)
26. Ney PA, Christopher MM, Hebbel R P Synergistic effects of
oxidation and deformation on erythrocyte monovalent cation leak.
Blood 75:1192,1990
27. Liu SC, Derick LH, Zhai S, Palek J: Ultrastructural anatomy
of the red cell membrane lesion in sickle cells: Penetration of the
hemoglobin S polymers through the membrane skeleton and reorganization of the skeletal lattice. Blood 74:44a, 1989 (suppl 1)
From www.bloodjournal.org by guest on June 17, 2017. For personal use only.
1990 76: 2139-2145
Permeability characteristics of deoxygenated sickle cells
MR Clark and ME Rossi
Updated information and services can be found at:
http://www.bloodjournal.org/content/76/10/2139.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.