Red Cell Membrane and Cation Deficiency in Rh Null Syndrome

By Samir K. Ballas, Margaret R. Clark, Narla Mohandas, H. F. Colfer, M. S. Caswell, M. 0. Bergren,

H. A. Perkins, and S. B. Shohet

A 52-yr-old multiparous white female was found to have Rh the increased osmotic fragility of whole blood. This was
null blood type. She had macrocytic anemia. with reticulo- confirmed by a density-dependent reduction in cell choles-
cytosis (1 5%-20%). of long duration. Although stomato- terol content. suggesting membrane instability in vivo. Rh
cytes in peripheral blood were numerous and osmotic null subpopulations showed a twofold increase in both
fragility was increased, suggesting increased cell water, ouabain-sensitive and -insensitive Na-K ATPase activity
the RBC cation content, and thus cell water, was and ‘Rb transport. even in the dense fraction with the
decreased. Cell dehydration was confirmed by an increased fewest reticulocytes. No membrane protein or glycopro-
proportion of high density RBC on Stractan density gra- tein abnormality was detected by SDS-PAGE. The asso-
dients. The deformability of RBC from four gradient sub- ciated deficiencies of both membrane surface area and

Downloaded from http://ashpublications.org/blood/article-pdf/63/5/1046/1638737/1046.pdf by guest on 09 September 2021

populations was measured in the ektacytorneter as a cation content in Rh null cells, as well as increased Na-K
function of suspending medium osmolality. Analysis of pump activity. suggest a pleiotropic functional interrela-
these measurements showed an abnormal reduction in cell tionship among Rh antigen. membrane stability. and cation
surface area with increasing cell density. thus explaining regulation.

E RYTHROCYTES FROM Rh NULL individuals Rheologic and biochemical measurements showed a

lack all Rh-Hr determinants including the Land- reduction in both membrane surface area and intracel-
steiner-Weiner (LW) factor.”2 Individuals who bear lular ion and water content, suggesting some role for
this rare phenotype have an undefined red cell mem- the antigen in both membrane stability and volume
brane defect associated with variable degrees of sto- regulation capacity.
matocytosis, spherocytosis, increased osmotic fragility,
and hemolytic anemia.35 In addition, Lauf and Joiner MATERIALS AND METHODS

found increased rates of Na, K transport and of Na-K Erythrocyte Preparation

ATPase activity, which were higher than could be Erythrocytes from the proposita, her relatives, and from healthy
explained on the basis of reticulocytosis.6 These asso- volunteer human donors were drawn into heparin or acid-citrate
ciated defects, considered as indications of abnormal dextrose, cooled to 4#{176}C,
and processed within 24 hr of collection. All
membrane function, provide support for the idea that procedures, unless otherwise noted, were performed at 4#{176}C.
Aliquots
of cells were washed at least 3 times in buffers, as specified for the
the Rh antigen, whose molecular identity has not yet
various procedures described below, and the bully coat was removed
been firmly established, is an integral part of the red by aspiration after each wash.
cell membrane.5 Although the Rh null disease is rare,
the estimated incidence being I /6 x I 06 individuals,7 it Preparation of Resealed Erythrocyte Ghosts
is, nevertheless, of considerable interest because it Erythrocytes, previously washed in 140 mM NaCl, 10 mM
indicates that certain blood group antigens may play Tris-HCI, pH 7.4, were lysed in 40 volumes of ice-cold hypotonic
an important role in the maintenance of normal struc- medium containing 7 mM NaCI, 5 mM Tris-HCI, pH 7.4. After
hemolysis was complete, the hemolysate was centrifuged at I 5,000
ture and function of the erythrocyte and its membrane.
rpm for 10 mm in a Sorvall RC-5 centrifuge. The supernatant was
In an effort to further define the role of the Rh antigen,
removed, and the ghosts were suspended in I 0 volumes of “resealing
we have studied selected properties of Rh null red cells. buffer,” after which they were incubated at 37#{176}Cfor I hr to
complete resealing. The resealing butTer contained 10 mM Tris-
HCI, pH 7.4, to which NaCI was added to adjust the osmolality to
From the Cardeza Foundation for Hematologic Research.
290 mosmole/kg. A subsequent centrifugation at 15,000 rpm for 5
Department of Medicine, Jefferson Medical College. Thomas
mm produced a concentrated ghost suspension for the fragmentation
Jefferson University, Philadelphia, PA; and the Departments of
measurement described below. This procedure yielded ghosts
Laboratory Medicine and Medicine. and the MacMillan-Cargill
impermeable to Dextran, but probably not to small cations, such as
Hematology Research Laboratory, University of California. San
Na or K. However, the ghost fragmentation assay only requires
Francisco, and Irwin Memorial Blood Bank, San Francisco. CA.
resealing to Dextran in order to produce laser diffraction patterns
Presented in part at the 24th Annual Meeting of the American
from the ghosts suspended in Dextran solutions. Thus, it was
Society ofHematology. Washington. DC. December 6, /982.
unnecessary to add chelators, such as EGTA or adenosine triphos-
Supported in part by the Deans c Overage Research Program of
phate (ATP). to the “resealing buffer” to insure resealing to small
Jefferson Medical College, Philadelphia. PA, and in part by
molecules. When examined by phase-contrast microscopy, the
USPHS Grants AM /6095, AM 26263, and HL 20985. The major
resealed ghosts had a smooth discoid morphology.
part ofthis work was done during the sabbatical leave ofS. K. B.
Submitted February 24. 1983; accepted November 8. 1983.
Ektacytometric Studies
Address reprint requests to Dr. Samir K. Ballas. Cardeza Foun-
dation, 1015 Walnut Street. Philadelphia. PA 19107. The ektacytometer, a viscodifTractometer, designed and pre-
© I 984 by Grune & Stratton, Inc. viously described by Bessis and Mohandas,89 was used to determine
0006-4971/84/6305-O0I0$03.00/0 red cell deformability under various experimental conditions. This

1046 Blood. Vol. 63, No. 5 (May). 1984: pp. 1046-1055

ABNORMALITIES OF Rh NULL ERYTHROCYTES 1047

device imposes a well-defined laminar shear stress field on erythro- containing 5 mM glucose, and the buffy coat was removed after each
cytes. while simultaneously monitoring the extent of cell deforma- wash. Washed cells were resuspended at a hematocrit of 10% in the
tion. The instrument consists of a concentric cylindrical viscometer, same medium, containing 0.1 mM ouabain to define the ouabain-
in which the outer cylinder can be rotated to produce the desired insensitive component of K efflux. After preincubation at 37#{176}C
for
level of shear stress. A laser beam is directed through the cell 10 mm, triplicate aliquots ofcell suspension were removed at I 5-mm
suspension, which is contained within a 0.5-mm gap between the two intervals, centrifuged in a Beckman microfuge, and the supernatants
cylinders. The cells are oriented by the shear stress field and diffract were reserved for subsequent analysis of K4 by flame photometry.
the laser beam to produce a coherent diffraction pattern, from which Initial intracellular cation concentrations were determined on hemo-
cell deformation is determined, using a photometric image analysis lysates of cell aliquots washed in isotonic Tris-buffered MgCl2 ( I0
system.89 A “deformability index” (DI) is obtained, which is equiva- mM Tris-HCI, pH 7.4, 290 mosmole/kg). First-order K effiux rate
lent to a measure of the ellipticity of a uniformly deforming cell constants were determined by plotting the log of the declining
population. In the standard mode of operation, the DI is recorded intracellular K concentrations, calculated from the initial and
continuously as a function of shear rate.9’#{176} For measurement of supernatant concentrations, versus time.
intact cell deformability. 10 zl of 40% red cell suspension is To define the chloride-dependent component of the ouabain-
thoroughly mixed with 3.0 ml of polyvinylpyrrolidone (PVP, mol wt insensitive K effiux,’5 experiments were performed in which paired

Downloaded from http://ashpublications.org/blood/article-pdf/63/5/1046/1638737/1046.pdf by guest on 09 September 2021

360,000, 4g/dl, w/v, 32.6 cp viscosity. 290 mosmole/kg, pH 7.4). At samples were run with the addition of I mM Piretanide (3-N-
20#{176}C,
this suspension produces a maximum shear stress of I 70 pyrrolidino-4-phenoxy-5-sulfamoylbenzoic acid). One such experi-
dynes/sq cm at the maximum applied shear rate of 100 rpm. ment was performed on freshly drawn cells, and a second was
Numerical values of the maximum deformability index reached, performed on cells from a separately drawn sample of heparinized
defined as Dl max, were used to compare the deformability of whole blood stored for 5 days at 4#{176}C
in the presence of 0.5 volume
different samples. modified Alsever’s solution (Gamma Biologicals, Inc., Houston,
The ektacytometer was also used to measure whole cell deforma- TX) and 0.5 volume of phosphate-buffered saline containing 5 mM
bility of erythrocytes as a continuous function of the suspending K and 10 mM glucose (BSKG).’4 Storage had no effect on the K
medium osmolality at a constant applied shear stress of 170 dynes/ effiux rate constants obtained.
sq cm (Osmotic Gradient Ektacytometry). For these studies, the Dl
of red cells was continuously recorded as the suspending medium Adenosine Triphosphatase Assays
osmolality was linearly increased from 50 to 900 mosmole/kg, as
Hemoglobin-free erythrocyte membranes were prepared from red
previously described.”
cells by hypotonic lysis and washing in 7 mM NaCI, 5 mM Tris-HCI,
I .0 mM EDTA, pH 7.4.
Fragmentation Assay
The Mg2-dependent adenosine triphosphatase (ATPase) activity
The ektacytometer was also used to evaluate the mechanical was assayed in a medium containing 3.0 mM ATP (disodium salt
stability of cell membranes by determining the relative tendency of neutralized to pH 7.1), 20 mM Tris-HCI, pH 7.4, 1.0 mM EGTA,
red cell membranes to fragment into undeformable spherical vesicles 0.2 mM ouabain, 3 mM MgCl2, 1 16 mM NaCI, and 16 mM KCI.
under shear stress.’2 For each measurement, 100 .il of resealed Ouabain was omitted when the total Na -K ATPase activity was to
ghosts, prepared as described above, were thoroughly mixed with 3.0 be determined. The Na-K ATPase was defined as the ouabain-
ml of Dextran solution (mol wt 40,000, 35 g/dl, w/v, 97.5 cp, 290 sensitive component, and the ouabain-insensitive component as the
mosmole/kg. pH 7.4). Samples were brought to the maximum shear Mg2-dependent ATPase activity. Each assay tube contained 100-
stress within 20 sec and maintained at that level until the DI fell to a 300 zg of membrane protein in the presence or absence of 0.2 mM
constant plateau (3-10 mm). The Dl signal decay was recorded ouabain in a total volume of 2.0 ml. ATP and membrane blanks were
continuously as a function of time, and the time required for it to run simultaneously. Incubations were at 37#{176}Cfor I hr. The
decay to half its maximum value was designated t’/2, the half-time inorganic phosphate (Pi) liberated was measured by the method of
for fragmentation. The applied shear stress was 575 dynes/sq cm, as Fiske and Subbarow,’6 as modified by Bartlett.’7 Each ATPase
calculated from the solution viscosity and the maximum shear rate of activity was expressed as nmole Pi liberated/mg membrane protein/
590 sec ‘ provided by the ektacytometer.8’#{176} mm.
In some experiments. the Na-K ATPase activity of open mem-
Separation ofRed Cells by Density brane preparations was defined as the rate of hydrolysis of adeno-
sine-5’ (‘y-32P)-triphosphate, as previously described.9’8
Subpopulations of erythrocytes of uniformly defined densities
were isolated by centrifuging normal and Rh null cells on discontin-
uous Stractan II (St. Regis Paper Co., Tacoma, WA) density Membrane Lipid Extraction and Quantitation
gradients.’3 The gradients covered a density range from I .065 to
Erythrocyte lipids were extracted according to the method of Rose
I .1 39 g/ml, in increments of about 0.004 g/ml. and Oklander’9 using isopropanol-chloroform. Phospholipid-phos-
phorus was determined by the method of Fiske and Subbarow,’6 as
Active and Passive Influx of 86Rb modified by Bartlett.’7 Cholesterol was measured by the method of
Influx of Rb ions was determined by measuring the intracellular Wybenga et al.#{176}For analysis of phospholipid classes in red cells,
accumulation of 86Rb added to the cell suspension medium in the lipid extracts were dried with N2 and the various phospholipid
presence or absence ofO. I mM ouabain, as previously described.’4 In classes were separated by thin-layer chromatography (TIC) on
these experiments. erythrocytes were washed and suspended in silica gel H plates, using the solvent system described by Skipski et
buffered saline containing potassium and glucose (BSKG), as al.2’ Lipid spots on the TIC plates were visualized using iodine
described before.’4 vapor, and their phosphorus content was determined.’7

Potassium Efflux Analytical Methods

Red cells, drawn into heparin, were washed 3 times in a HEPES- Hemoglobin (Hb), mean corpuscular volume (MCV), and mean
buffered (10 mM, pH 7.50) NaCI solution (290 mosmole/kg) corpuscular hemoglobin concentration (MCHC) were determined
1048 BALLAS ET AL.

by standard methods and the hematocrit (Hct) was determined by

centrifugation. Red cell counts were determined using a Coulter
Counter Model F. Immunohematologic studies were performed Ri Ri
according to standard Blood Bank procedures. Red cell sodium and
potassium concentrations were determined by flame photometric
analysis of red cell aliquots washed in ice-cold, isotonic Tris-buffered ECTOPIc
PREGNANCY
McCl2 solution (10 mM Tris-HCI, pH 7.4). Membrane proteins
were analyzed by polyacrylamide gel electrophoresis in sodium
dodecyl sulfate (SDS-PAGE), as previously described.22 Red cell Ri Rz

membrane protein content was determined by the method of Lowry

et al.23
Fig. 1 . Family pedigree. The arrow indicates the proposita.

RESULTS members studied are indicated in Fig. I . Her husband

Case History typed as group A, Rh0(D) positive (R,R, or DCe/

Downloaded from http://ashpublications.org/blood/article-pdf/63/5/1046/1638737/1046.pdf by guest on 09 September 2021

DCe) and her daughter typed as group 0, Rh0(D)
K.M. is a 52-yr-old white multiparous woman who
was transfused in 1958 with 2 U of blood during an positive (R,R, or DCe/DcE).

operation for oophorectomy for ectopic pregnancy. She Hematologic Studies

was typed, at that time, as blood group 0, Rh0(D)
negative. In April 1 982, she sustained a fracture of her The patient’s Hb varied between 8 and 10.8 g/dl,
left humerus and was found to have multiple blood Hct 26%-3l%, MCV 100-104 fi,and MCHC 31-35
g/dl. The reticulocyte count ranged from 14% to 20%,
group antibodies in her serum during routine blood
and the osmotic fragility of freshly drawn red cells was
ba nk testi ng. Comprehensive immunohematologic
minimally increased (50% hemolysis at 0.45% NaCI;
investigations, described below, were diagnostic of the
normal range for this laboratory 0.39%-0.43% NaC1).
Rh null phenotype.
Serum haptoglobin was reduced to 5 mg/dl, and Hb
The patient is known to have had macrocytic anemia
electrophoresis revealed no abnormalities. Examina-
of moderate severity and of unknown etiology for a
tion of the peripheral smear showed numerous sto-
long time. Serum iron, total iron binding capacity, folic
matocytes, some spherocytes, and polychromasia (Fig.
acid, and vitamin B12 levels were all within normal
2A). Wet preparations, examined by phase-contrast
limits. Results of Schilling tests I and II were normal.
Her anemia did not respond to therapeutic trials with
iron or vitamin replacement.

Serologic Investigations
The patient’s erythrocytes lacked all Rh antigens
and the LW factor. Her serum agglutinated all red
cells teated, except Rh null cells (two examples),
giving 2 + reactions at room temperature, 3 + reac-
tions at 37#{176}C,
and 4+ reactions at the antiglobulin
phase. The antibody was completely removed by two
absorptions with R1R,, R2R2, and rr cells. It was
identified as anti-Rh 29 antibody, which reacts with
the total Rh antigens.
Aliquots of the patient’s cells were used to absorb
anti-Rh0(D), anti-rh’ (C), anti-rh” (E), anti-hr’ (c), and
anti-hr”(e) antibodies. Eluates were prepared and
tested with the appropriate cells. The patient’s cells did
not reduce the titer of the antisera, nor was it possible
to elute antibody from them after exposure to the
antisera.
Taken together, all of the above serologic investiga-
tions are diagnostic of the Rh null phenotype. Based on
the above and on additional studies, the patient’s
erythrocytes typed as group 0, Rh null (D - ,C ,E - -,

c-,e-), LW-, M+, N+, 5+, s-, U-, Le(a-b-),

Fig. 2. (A) Photomicrograph of faxed blood smears from the
Pl+, 1+, Fy(a-b+), Jk(a-b+), K+, k-, proposita showing stomatocytes and spherocytes. (B) Scanning
K(a-b+), Js(a-b+), Vel+ and Lu(b+). Family electron microscopy of glutaraldehyde-fixed Rh null cells.
ABNORMALITIES OF Rh NULL ERYTHROCYTES 1049

Table 1 . Cation Concentration and Content of Normal and Rh Null Cells

meg/Liter ABC meq/lOg Hb

ABC Experiment Na K’ Na + K’ Na’ K’ Na’ + K’

Control 1 8.5 94.8 103.3 - - -

2 8.9 92.1 101.0 0.26 2.68 2.94

3 8.9 98.5 107.3 0.28 3.04 3.32
Rh null 1 4.6 75.0 79.6 - - -

2 6.6 85.4 92.0 0.19 2.46 2.65

3 6.4 92.1 98.5 0.19 2.66 2.85

For a series o f 20 control samp les, the following average (SD) cation contents were found: Na4 -0.26 (0.05). K’ -2.83 (0.17 ), Na + K -3.08
(0.18).

microscopy, showed that the red cells were bowl- deficiency of about 25% in total cation content, despite
shaped instead of biconcave. Scanning electron micros- the presence of 9.6% reticulocytes. These cells were,

Downloaded from http://ashpublications.org/blood/article-pdf/63/5/1046/1638737/1046.pdf by guest on 09 September 2021

copy of the red cells (Fig. 2B) confirmed the presence therefore, severely dehydrated. The density distribu-
of stomatocytes. Scintigraphic studies using 99mTc.. tion of the Rh null cells, shown in Fig. 3, indicates that
sulfur colloid showed marked splenomegaly (5-6 times the majority of the cells are the high-density subpopu-
normal size), but no hepatomegaly. lations (density >1.095), in which a deficiency of
cation content was found.
Special Hematologic Studies To provide further information about cation regula-
Because of reported heterogeneity in the hydration tion in the Rh null cells, we measured various compo-
status ofstomatocytes,24 we next determined the cation nents of monovalent cation fluxes. In the experiment
content of Rh null erythrocytes, as summarized in summarized in Fig. 4, 86Rb influx into red cells was
Table I . Both Na and K were slightly decreased in Rh determined in the presence or absence of 1 .0 mM
null red cells compared to control cells, suggesting ouabain. As indicated, both the total and ouabine-
modest cell dehydration. This cellular dehydration was insensitive (passive) influx of Rb was increased in Rh
confirmed by the observation of increased populations null erythrocytes. The ouabain-sensitive influx of Rb,
of high-density red cells on discontinuous Stractan generally taken to represent the active, pump-
density gradients (Fig. 3). Table 2 summarizes the mediated influx, was also increased (by 72%) in Rh
reticulocyte count, the MCHC, and the cation content null cells. Similar results were obtained when Rb flux
of five subpopulations of Rh null erythrocytes sepa- was determined in one of the dense Rh null Stractan
rated on the density gradient. As indicated, the most fractions (stopping density I . I 05; see Table 2), as
=

dense red cells had a very high MCHC and showed a shown in Fig. 4B. In this experiment, control cells had

35
A B
CONTROL Elli
30
Rh null

25

-J
< 20
I-
0
H
15
0

Fig. 3. Stractan gradient

I0
density distribution for control
and Rh null erythrocytes. The
latter show a shift in cell
population to higher density. 5
The inset shows the actual
density profile of Rh null (A)
and normal (B) erythrocyte
0
subpopulations separated on I 080 1.090 1.100 I 110 1.120 I 180
discontinuous Stractan density
gradients. STOPPING DENSITY (g/ml)
1050 BALLAS ET AL.

Table 2. Properties of Rh Null Subpopulations Isolated by Strac tan Gradient Centrifugation

. MCHC (g/dl) . meg/lOg Hb

Stopping’ MCV Aeticulocytes
Density Control Rh Null (fi) (%) Na’ K’ Na’ + K

1.075 - 24.8 123 60.4 0.21 4.19 4.41

1.083 30.2 - - - - - -

1.085 - 28.5 112 34.3 0.18 3.49 3.67

1.087 31.0 - 95 - - - -

1.092 32.8 - 93 9.3 0.22 2.95 3.17

1.095 - 34.9 101 18.0 0.14 2.61 2.75
1.096 33.8 - 92 0.5 0.25 2.86 3.11
1.101 34.9 - 90 0 0.26 2.67 2.93
>1.101 35.7 - 89 0 0.27 2.24 2.53
1.105 - 39.0 90 8.4 0.17 2.09 2.27
1.115 - 44.2 81 9.6 0.21 1.55 1.76

Downloaded from http://ashpublications.org/blood/article-pdf/63/5/1046/1638737/1046.pdf by guest on 09 September 2021

The density of t he Stractan layer on which each cell subpopulation rested.

an MCV of 89 fi and a reticulocyte count of 1.2%,

whereas the fractionated Rh null cells had an MCV of
(6 90 fi and a reticulocyte count of 8.4%. Thus, both cell
types used had the same volume, although other mea-
14
surements indicated that the Rh null cells had less
surface area (see below). Measurement of ATPase
activity in membrane preparations from Rh null cells
U

and their subpopulations showed a twofold increase in

10

0
both Na-K ATPase and Mg ATPase activities, as
0
measured by colorimetric determination of the rate of
x inorganic phosphorus released from ATP (Table 3).
a.
U
Although the increased number of reticulocytes would
be expected to cause an increase in ATPase activity,
the percentage of reticulocytes appeared to have little
effect on the ATPase activities in the various Rh null
subpopulations. An alternative method of measuring
ATPase activity, in which we determined the total and
ouabain-sensitive hydrolysis of y-32P-ATP by unsepa-
TIME (mm)
rated red cell membranes, also showed increased
25 ATPase activity in Rh null membranes. Total hydroly-
sis of radioactive ATP was 6.45 nmole/mg/min by Rh
20 null membranes versus 3.03 in controls; ouabain-
sensitive hydrolysis was 2.45 and 1 .26 by Rh null and
0 normal membranes, respectively. It should be noted
0 that the values for the ATPase activities measured by
‘y-32P-ATP hydrolysis are lower than those shown in
Table 3, probably because of differences in the condi-
Rh null FRACTION.....
tions of incubation in the two methods used. Most
CONTROL
---&-
importantly, the concentration of ATP in the colon-
metric method was 3.0 mM (see above), whereas in the
60
TIME (minI
-y-32P-ATP hydrolysis method it was less than 1.0
mM.9”
Fig. 4. (A) ‘Rb influx into control and Rh null erythrocytes. In order to further investigate the nature of the
Cells are incubated at 37’C in glucose phosphate-buffered saline
containing 5 mM K and Rb in tracer quantities. in the presence
abnormality in cation regulation in Rh null cells, we
and absence of 0.1 mM ouabain. (B) Rb influx into control RBC determined K efilux in the presence on absence of 1.0
and one of the Rh null subpopulations separated on Stractan mM piretanide, an inhibitor of Cl transport and
density gradient. The Rh null fraction (MCV 90 fi. retics 8.4%) was
that which rested on the Stractan layer with 1 .1 05 g/ml density
chloride-dependent K flux.’5 Figure 5 shows that Rh
(see Table 2). MCV of unseparated control RBC = 89 fi. null cells had a slightly increased K efflux in the
ABNORMALITIES OF Rh NULL ERYTHROCYTES 1051

Table 3. Comparison of ATPase Activities in C ontrol Cells. Rh Null RBC. and Their Subpopulations

. . Na’ - K’ ATPase Actmvmty (nmoles

Stopping Aeticulocytes
Density 1%) Ouabain-Sensitive Ouabamn-lnsensitive

1.085 34.3 (2) 13.7 ± 0.17 (2)5.0 ± 0.16

1.095 18.0 (2) 13.5 ± 0.15 (2) 6.2 ± 0.61
1.105 8.4 (2) 15.4 ± 0.20 (2) 5.2 ± 0.38
1.1 15 9.6 (2) 12.7 ± 0.34 (2)6.0 ± 0.34
Unseparated ABC
RhnuIl 15.6 (6) 13.1 ± 3.10 (6)5.1 ± 1.20
Control 0.7 (6) 6.3 ± 1.44 (6)2.9 ± 0.47

Values are mean ± SD. The number of determinations is indicated in parentheses.

The density of the Stractan layer on which each cell subpopulation rested.

absence of piretanide. The first-order rate constant for optimal deformability of normal red cells under iso-

Downloaded from http://ashpublications.org/blood/article-pdf/63/5/1046/1638737/1046.pdf by guest on 09 September 2021

K efflux for Rh null cells was 0.024 hr, compared to tonic conditions. This optimum represents a balance in
0.017 in control cells (41 .2% increase). However, addi- the adjustment of surface area/volume ratio and intra-
tion of 1 .0 mM piretanide to the medium reduced the cellular viscosity. Previous work9” has shown that the
rate constant for K effiux for Rh null cells to that maximum value of DI that is attained is influenced by
found for normal cells (Fig. 5). several factors. A reduction in cell surface area, sub-
In order to further elucidate the hydration status stantial heterogeneity of cell geometry, or increased
and surface area/volume ratio in Rh null erythrocytes membrane stiffness can all cause a reduction in the
and their subpopulations, we studied these cells by the maximum. A second feature is that, as medium osmo-
method of osmotic gradient ektacytometry, described lality decreases below 290 mosmole/kg, the DI
previously.” The osmotic gradient deformability pro- declines, reaching a minimum at a characteristic hypo-
files for normal and Rh null cells are shown in Fig. 6. tonic osmolality indicated by 0 minimum (0 mm).
In this method, the deformability index of cells sub- This decline in the ability of the cells to elongate occurs
jected to constant shear stress is continuously mea- because of their osmotically increased volume at con-
sured while the suspending medium osmolality is stant surface area. The hypotonic osmolality at which
linearly varied over a broad hypo- and hypertonic the DI reaches a minimum coincides with the osmolal-
range (50-900 mosmole/kg). The normal curve con- ity at which 50% of the cells hemolyze in an osmotic
tains several important features. First, it reaches a I .0
maximum near 290 mosmole/kg, corresponding to the

o.e
(40
0 x
- w
4 0
z
(20 0.6
::2 >-
I-.
a. -J
‘00 n.2 51
n.2
0 4
I 0.4
a:
0
80 U.
Ui
-I 0
0 0.2

60
0
U
U.
0 40
0
0
,
00
0 MIN

200
.

300
u- ,

400 500

OSMOLALITY MOSMOL/KG
U) 20
U)
0 Fig. 6.
Osmotic deformability profile for normal and Rh null
-I
erythrocytes. The ektacytometric deformability index (Dl) was
Co N TROL Rh null recorded as a continuous function of suspension osmolality. The
profiles are characterized by the following features: (1 ) 0 mm. the
Fig. 5. K ‘ loss from freshly drawn normal and Rh null erythro- osmolality at which the Dl reaches a minimum in the hypotonic arm
cytes. Cells were incubated for 1 hr at 37’C in 10 mM HEPES (pH of the curve, (2) the maximum value attained by Dl. which occurs
7.5). 5 mM glucose. and NaCI to bring the final osmolality to 290 near isotonic osmolality for normal cells. and (3) 0’. the osmolality
mosmole/kg. in the absence or presence of 1 .0 mM piretanide. at which Dl equals half the normal maximum on the hypertonic arm
The values were corrected for the very small amount of hemolysis of the curve. The shaded area represents the range for normal
that occurred during the incubation. control cells.
1052 BALLAS ET AL.

.0
fragility assay, and the cells, on the average, have a
spherical geometry.” Thus, this critical osmolality
provides information about the critical hemolytic vol- 0.8

ume and the initial surface area/volume ratio. Finally, 3<

Ui
0
as the suspending medium osmolality increases above z
0.6
290 mosmole/kg, the DI decreases again, reaching 3-
I-
half the normal maximum DI at a characteristic -J

51
osmolality, 0’. This decrease in hypertonic deformabil- 4 0.4
a:
ity results from an increase in intracellular viscosity 0
U.
secondary to osmotic water loss and increasing hemo- Ui
0
0.2
globin concentration. The position of this declining
portion of the curve along the osmolality axis come-
lates with the initial MCHC ofthe cell sample.”

Downloaded from http://ashpublications.org/blood/article-pdf/63/5/1046/1638737/1046.pdf by guest on 09 September 2021

0.

The osmotic gradient deformability profile of Rh 0 00 200 300 400 500

null cells showed several abnormalities. The 0 mm was OSMOLALITY MOSMOL/KG

shifted to higher osmolality, which was consistent with
Fig. 7. Osmotic deformability profile for Rh null subpopula-
a slight increase in the osmotic fragility of Rh null cells tions isolated on Stractan density gradients. The shaded area
and was indicative of a reduction in surface area/ represents the range for normal controls. The numbers indicate
volume ratio under isotonic conditions. This could have the density of the Stractan layer above which the Rh null subpopu-
lation rested. All Rh null fractions have increased osmotic fragility.
reflected either a reduction in membrane surface area
as indicated by the shift in 0 mm to higher osmolality.
or an increase in cell volume. The observation of
increased numbers of high density cells and of reduced
potassium and total cation content indirectly impli- were lower in cholesterol than the most dense 2% of
cated a reduction in membrane surface area. The normal cells, which contained no reticulocytes. As
deformability profile provided more direct evidence in expected, the low-density, gradient-separated cell pop-
that the maximum DI value was reduced, an effect ulations had elevated cholesterol content because of
seen in surface-deficient red cells.” However, the their very high reticulocyte content. Analysis of the
abnormally gradual slope of the hypertonic arm of the phospholipid subclasses revealed no abnormality in the
profile indicated that the Rh null sample consisted of a phospholipid composition of unseparated Rh null cells
heterogeneous mixture of cells covering a broad or their subpopulations (Table 5).
MCHC range, and it was also possible that cellular To determine whether the presumed instability of
heterogeneity could have been responsible for the the Rh null membrane could be detected as an increase
reduced DI maximum. in mechanical fragility, we measured their vulnerabil-
In order to resolve this question, we also obtained ity to shear-induced fragmentation. These measure-
osmotic gradient deformability profiles for the Rh null ments showed that Rh null membranes fragmented at
subpopulations isolated on density gradients. As shown the same rate as control membranes. Analysis of
in Fig. 7, the two most dense, dehydrated subpopula- membrane proteins by SDS-PAGE showed no detect-
tions had the same 0 mm as the more hydrated cells able quantitative abnormality in either membrane
and were thus osmotically fragile despite their reduced protein or glycoprotein composition of Rh null cells.
water content. In addition, these dense cells showed
DISCUSSION
even more striking reductions in the maximum DI than
the whole blood. Taken together, these observations The immunohematologic data establish this patient
provided further support for a process of concomitant as another example of Rh null disease. Our findings
loss of membrane area and intracellular cation con- confirm, and, to an extent, expand the data described
tent. Further, direct measurement of red cell phospho- in previously reported cases.’ 7.25.26 The fact that the
lipid and cholesterol content showed that unseparated proposita transmitted the R2(DcE) gene complex to
Rh null cells, despite their high proportion of lipid-rich her daughter (Fig. 1) indicates that the mode of
reticulocytes, did not contain more cholesterol per cell inheritance in this case is due to a suppressor, not an
than normal cells (Table 4). Analysis of density- amorphic, gene. The negative findings of the absorp-
isolated cells showed that the high-density cells were tion and elution experiments indicate that the expres-
preferentially reduced in cholesterol content. Although sion of Rh determinants is completely suppressed in
it is true that normal cells also show a density- this patient. The technique used detects the weak Rh
associated reduction in cholesterol, the most dense 9% variant (D’s) and the weakest subtype of blood group
of Rh null cells, which contained 7% reticulocytes, A,27’28 and even at this level of sensitivity, no Rh or Hr
ABNORMALITIES OF Rh NULL ERYTHROCYTES 1053

TabI e 4. Membrane Lipid Composition of Normal Cells. Rh Null RBC. and Their Subpopulations

Percent of Total Cells Retmcu)ocytes (%) ChoesteroI (g/ 10 ‘ Celist)

Stopping Density Control Rh Null Control Rh Null Control Rh Null

1.083 3.9 - 7.5 - 1.37(0.03) -

1.087 35.5 - 0.8 - 1.40(0.06) -

1.092 33.5 30.8 0.5 23 1.48 (0.01) 1.58 (0.05)

1.096 21.1 - 0.5 - 1.39(0.02) -

1.101 4.4 24.2 0.5 14 1.15 (0.05) 1.49(0.09)

1.106 - 20.9 - 12 - 1.25 (0.06)
1.115 - 15.5 - 9 - 1.23(0.01)
1.168 1.6 8.6 0.5 7 1.11 (0.01) 1.01 (0.06)
Unseparated ABC 1 .0 1 3.4 1 .29 (0.04) 1 .36 (0.06)

The density of the Stractan layer on which each cell subpopulation rested.

tValues shown are the average of two determinations. The range of variation is given in parentheses.

Downloaded from http://ashpublications.org/blood/article-pdf/63/5/1046/1638737/1046.pdf by guest on 09 September 2021

determinants could be detected. Similarly negative membrane that serve as a common substrate for the
studies from elution experiments were reported in action of the Rh-Hr genes and of the genes of the
other cases of Rh null disease.”2629 It must be pointed MNSs and U systems.3
out that the clinical and hematologic picture in Rh null The clinical picture of our patient is similar to
disease is the same regardless of whether the mode of previously reported cases of Rh null disease.’7’25’26 She
inheritance is due to suppressor or amomphic gene. had anemia of unknown etiology of long duration
Thus, it appears that the absence of the Rh antigen before the diagnosis was established. Other causes of
binding site itself results in the membrane defect. anemia were looked for and ruled out. Unlike most of
Rhm,j S a different RBC abnormality, described in the previously reported cases, this patient presented
two North American families,30 in which expression of with macrocytic anemia due, most likely, to the high
Rh antigens is depressed, but not absent, because of reticulocyte count. Stomatocytes and spherocytes were
the influence of a suppressor gene. This disorder was noted in her peripheral blood.
also accompanied by stomatocytosis and hemolytic The present studies, like those of previous investiga-
anemia, which required splenectomy.3#{176} This variant tors,6’3335 found evidence for increased rates of both
was ruled out in the present case by the absorption and active and passive cation transport in Rh null cells. In
elution studies described. our patient, these cation flux abnormalities were asso-
Another point to be noted is that the patient’s ciated with apparent in vivo loss of cations and water
erythrocytes were found to be U-negative, although from the cells. Although the extent of dehydration was
they are S+. This aberrant U blood group behavior substantial, as indicated by a marked increase in cell
accompanying Rh null disease has been noted density for virtually the entire cell population, the
before.331’32 The absence of the Rh-Hr determinants observed leakage of K from the cells was only modestly
seems to cause attenuation of the Ss and U antigens. In increased. Thus, the dehydration may not simply be
other words, Rh null cells type falsely negative for the due to an increase in K permeability, as was proposed
U antigen. The exact mechanism leading to this aber- for other disorders involving red cell dedhyration.36
rant antigenic interrelation is not well understood. The One alternate (speculative) possibility is that an
proposed mechanism is homozygosity for a gene whose increased number of red cell Na pump sites, as was
normal allele produces components of the erythrocyte found in another Rh null patient by Lauf and Joiner,6

Table 5. Distributi on of Phospholipids in Rh Null Subpopulations

Phospholipid Class (%)t

Stopping
Density SM PC P1 + PS PE

1.085 27.1 26.3 17.4 29.2

1.095 25.9 27.1 18.8 28.3
1.105 28.1 29.5 15.4 27.0
1.115 26.1 28.3 16.2 29.4
Unseparated RBC
Rhnull (4)25.0± 1.96 (4)28.6±0.39 (4)17.3± 1.34 (4)29.1 ± 2.55
Control (4) 27.0 ± 0.40 (4) 28.9 ± 0.18 (4) 16.2 ± 0.82 (4) 28.0 ± 0.84

The density of the Stractan layer on which each cell subpopulation rested.
tSM, sphingomyelin; PC, phosphatidylcholine; P1 + PS, phosphatidylinositol and -serine; PE. phosphatidylethanolamine.
*Values are mean ± SD. Number of determinations is indicated in parentheses.
1054 BALLAS ET AL.

could contribute to a reduction in total cation content. observed differences in the fluorescence of a mem-
That is, because the Na-K pump expells three Na ions brane probe and in labeling of membrane sulfhydryl
for the uptake of two K ions, a high level of pump groups in Rh null cells that they interpret as indica-
activity would tend to depress both Na and total cation tions of abnormal membrane lipid-protein interac-
levels. The fact that our patient’s red cells consistently tions. Such an alteration could possibly be a secondary
contained subnormal amounts of Na is consistent with consequence of membrane loss and reorganization. As
this possibility. The patients studied by Lauf and Schmidt has pointed out,38 Rh null disease is similar in
Joiner6 had both normal Na and water content. How- many respects to hereditary spherocytosis. The present
ever, variations in the red cell permeability to Na evidence for membrane loss in vivo extends these
might account for differences in hydration status, as similarities. In hereditary spherocytosis, one or several
the steady-state ion and water content of the cell defects in spectnin394’ are thought to be responsible for
represents a dynamic balance of active and passive ion membrane instability. The biochemical nature of the
fluxes. It should be noted that, in our patient, reticulo- Rh antigen has not been identified with certainty, and

Downloaded from http://ashpublications.org/blood/article-pdf/63/5/1046/1638737/1046.pdf by guest on 09 September 2021

cytosis may have contributed to the increase in oua- at present, there is no information about its interaction
bain-sensitive ion transport. However, normal reticulo- with other membrane components, such as spectnin or
cytes do not have reduced cation content;35 thus, other possible components of the membrane skeleton.
reticulocytosis per se cannot explain the reduced cation Although several studies agree that the Rh0(D)
content in this patient. antigen is an integral membrane protein, estimates of
A second possibility is that red cell dehydration in its molecular weight vary from 7,000 to 1 80,000.424
our patient could be the result of abnormal Na-K Traditionally, it is believed to be a lipoprotein of low
cotransport. Our finding that the slight increase in K molecular weight.45 More recently, Victoria and col-
efflux was reduced to normal by piretanide, an inhibi- leagues44 have shown that anti-Rh0(D) IgG binds to
ton of the chloride-dependent cotransport process,’5 band 3 glycoprotein of the human red cell, suggesting a
strongly suggests an abnormality in this component of band 3 localization for the Rh antigen. SDS-PAGE
cation transport. However, because the possible role of analysis of Rh null membranes from our patient was
cotransport in red cell volume regulation has not yet similar to previous studies44 in showing no detectable
been defined, and because we have measured the abnormalities in either the amount or the molecular
cation flux in only one direction, the implications of weight localization of the band 3 protein. Thus, even if
this observation are not clear. the Rh0(D) antigen is localized on band 3, the Rh null
One point of interest is that substantial numbers of defect does not suppress overall protein synthesis of
reticulocytes were found throughout the density gra- band 3. The Ss and U antigens, which also showed
dients, indicating that even very young cells had abnormal expression in our patient, are carried on
become dehydrated. Thus, it appears that the process glycophonin B.46’47 PAS staining of the SDS-PAGE
that causes dehydration affects cells randomly and gels showed no abnormality in either glycophonin A or
does not represent a progressive, cumulative change B. The multiple abnormalities of Rh null cells suggest
over the cell’s lifespan. a functional interrelationship among Rh antigens,
In addition to a reduction in red cell cation and membrane stability, and cation regulation, or, perhaps,
water content, the present study provides unequivocal a fundamental alteration in membrane organization.
evidence for a loss of membrane surface area by these Firm identification and characterization of the Rh
Rh null cells. Both unseparated and density-separated antigens may well provide useful insights into struc-
cells had increased osmotic fragility, as determined tural requirements for membrane permeability and
from the ektacytometnic osmotic gradient profiles. In stability.
addition, the density-separated cells showed a pro-
gressive reduction in lipid content, which culminated ACKNOWLEDGMENT
in a lipid content below that of normal cells in the
We are grateful to the patient and her family members for their
denser fractions, even though the reticulocyte count
cooperation in performing these studies. We thank Dr. Jyothsna
remained substantially elevated. The decreased ‘Narla for bringing this patient to our attention, and Barbara J. Scott
osmotic fragility of these dense cells, whose reduced for her secretarial assistance. We also thank the Hoechst-Roussel

water content should have made them osmotically Pharmaceutical Corporation for a gift of piretanide.

resistant, indicates that the reduction of lipid in the

relatively more mature cell populations went beyond a REFERENCES
simple maturational remodeling. It therefore appears 1 . Vos GH, Vos D, Kirk RI, Sanger R: A sample of blood with no
that absence of the Rh(D) antigen rendered the mem- detectable Rh antigens. Lancet I :14, 1961
brane unstable in some way. Smith et al.37 have 2. Levine P, Celano MJ, Vos GH, Morrison J: The first human
ABNORMALITIES OF Rh NULL ERYTHROCYTES 1055

blood ---/--- which lacks the “D-like” antigen. Nature 194:304, 25. Levine P. Celano MJ, Falkowski F, Chambers JW: A second
I 962 example of ---/--- blood or Rh null. Nature 204:892, 1964
3. Sturgeon P: Hematologic observations on the anemia asso- 26. Ishimosi T, Hasekura H: A Japanese with no detectable Rh
ciated with blood type Rh null. Blood 36:310, 1970 blood group antigens due to silient alleles or deleted chromosomes.
4. Schmidt PJ, Holland PV: Rh null disease. BibI Haematol Transfusion 7:84, 1967
38:230, 1971 27. Pittenger RG, McQuiston D, Sturgeon P: The Rh0 variant-
5. Levine P. Tripodi D, Struck J Jr, Zmijewski CM, Pollack W: D’. Ill. A field trial of the DU tube test in typing donor bloods.
Hemolytic anemia associated with Rh null but not with Bombay Transfusion 4:361, 1964
blood. Vox Sang 24:417, 1973 28. Sturgeon P, Moore BPL, Weiner W: Notations for two weak
6. Iauf PK, Joiner CH: Increased potassium transport and A variants: A,,.. and At,. Vox Sang 9:214, 1964
ouabain binding in human Rh null red blood cells. Blood 48:457, 29. Levine P. Celano Mi, Falkowski F, Chambers JW, Hunter
I 976 OB Jr. English CT: A second example of ---/--- or Rh null blood.
7. Seidle 5, Spielman W, Martin H: Two siblings with Rh null Transfusion 5:492, 1965
disease. Vox Sang 23:182, 1972 30. Mallory DM, Rosenfield RE, Wong KY, Heller K, Rubin-
8. Bessis M, Mohandas N: A diffractometric method for the stein P. Allen FH Jr. Walker ME, Lewis M: Rh mod, a second

Downloaded from http://ashpublications.org/blood/article-pdf/63/5/1046/1638737/1046.pdf by guest on 09 September 2021

measurement of cellular deformability. Blood Cells I :307, 1975 kindred (Craig). Vox Sang 30:430, 1976
9. Mohandas N, Clark MR, Jacobs MS. Shohet SB: Analysis of 31. Schmidt PJ, Lostumbo MM, English CT, Hunter OB Jr:
factors regulating erythrocyte deformability. J Clin Invest 66:563, Aberrant U blood group accompanying Rh null. Transfusion 7:33,
1980 I 967
10. Groner W, Mohandas N, Bessis M: Newoptical technique for 32. Schmidt PJ, Vos GH: Multiple phenotypic abnormalities
measuring erythrocyte deformability with the ektacytometer. Clin associated with Rh null (---/---). VoxSang 13:18, 1967
Chem 26:1435, 1980 33. Lee P. Stevenson MM: Membrane permeability to sodium
11. Clark MR. Mohandas N, Shohet SB: Osmotic gradient and potassium in Rh null red blood cells. Proc mt Union Physiol Sci
ektacytometry: Comprehensive characterization of red cell volume 11:16, 1974 (abstr)
and surface maintenance. Blood 61:899, 1983 34. Wiley JS: Cation fluxes in Rh null red cells, Blood 5 1:555,
12. Mohandas N, Clark MR. Heath BP, Rossi M, Wolfe IC, 1978 (letter)
lux SE, Shohet SB: A technique to detect reduced mechanical 35, Wiley JS, Shaller CC: Selective loss of calcium permeability
stability of red cell membranes: Relevance to elliptocytic disorders. on maturation ofreticulocytes. J Clin Invest 59:1 113, 1977
Blood 59:768, 1982 36. Glader BE, Sullivan DW: Erythrocyte disorders leading to K
I 3. Clark MR. Guatelli JC, White AT, Shohet SB: Study on the loss and dehydration. Prog Clin Biol Res 30:503, 1979
dehydrating effect ofthe red cell Na7K-pump in Nystatin-treated 37. Smith JA, Lucas FV, Martin AP, Senhauser DA, Vorbeck
cells with varying Na4 and water contents. Biochim Biophys Acta ML: Lipid-protein interactions oferythrocyte membranes: Compar-
646:422, 1981 ison of normal 0, Rh(D) positive with the rare 0, Rh null. Biochem
14. Clark MR. Morrison CE, Shohet SB: Monovalent cation Biophys Res Commun 54:1015, 1973
transport in irreversibly sickled cells. J Clin Invest 21:329, 1978 38. Schmidt PJ: Hereditary hemolytic anemias and the null blood
15. Dunham PB, Stewart GW, Ellory JC: Chloride activated types. Arch Intern Med 139:570, 1979
passive potassium transport in human erythrocytes. Proc NatI Acad 39, Agre P. Orringer EP, Bennett V: Deficient red-cell spectrin in
Sci USA 77:171 1, 1980 severe, recessively inherited spherocytosis. N EngI J Med 306:1 15,
16. Fiske CH, Subbarow Y: The colorimetric determination of 1982
phosphorus. J Biol Chem 66:375, 1925 40. Goodman SR. Shiffer KA, Casoria LA, Eyster ME: Identifi-
I 7. Bartlett GR: Phosphorus assay in column chromatography. J cation of the molecular defect in the erythrocyte membrane skeleton
Biol Chem 234:466, 1959 ofsome kindreds with hereditary spherocytosis. Blood 60:772, 1982
I 8. Blostein R: Relationships between erythrocyte membrane 41. Wolfe IC, John KM. Falcone JC, Byrne AM, Lux SE: A
prosphorylation and adenosine triphosphate hydrolysis. J Biol Chem genetic defect in the binding of protein 4. 1 to spectrin in a kindred
243:1957, 1968 with hereditary spherocytosis. N EngI J Med 307: 1 367. 1982
19. Rose GH, Oklander M: Improved procedure for the extrac- 42. Plapp FV, Kowalski MM, Tilzer I, Brown PJ, Evans J, Chiga
tion oflipids from human erythrocytes. J Lipid Res 6:428, 1965 M: Partial purification of Rh0(D) antigen from Rh positive and
20. Wybenga DR. Pileggi VJ, Dirstine PH, DiGiorgio J: Direct negative erythrocytes. Proc Natl Acad Sci USA 76:2964, 1979
manual determination of serum total cholesterol with a single stable 43. Gahmberg CG: Molecular identification of the human
reagent. Clin Chem 16:980, 1970 Rh0(D) antigen. FEBS Iett 140:93, 1982
21 . Skipski VP, Peterson RF, Barclay M: Quantitative analysis of 44, Victoria EJ, Mohan IC, Masouredis SP: Anti-Rh0(D) lgG
phospholipids by thin-layer chromatography. Biochem J 90:374, binds to band 3 glycoprotein of the human erythrocyte membrane.
I 964 Proc NatI Acad Sci USA 78:2898, 1981
22. Iaemmli UK: Cleavage of structural proteins during the 45. Calvin M, Evans RS, Behrendt V. Calvin G: Rh antigen and
assembly of the head of bacteriophage T4. Nature 227:680, 1970 hapten: Nature and its isolation from erythrocyte stroma. Proc Soc
23. Lowry OH, Roberts NR, Leiner KY, Wu MI, Farr Al: Exp Biol Med 61:416, 1946
Protein measurement with the Folin phenol reagent. J Biol Chem 46. Anstee AJ: The blood group MNSs-active sialoglycoproteins.
207:1, 1954 Semin Hematol 18:13, 1981
24. Nathan DG, Shohet SB: Erythrocyte ion transport defects 47, Dahr W, Uhlenbruck G, Schamalisch R, Jassen E: Ss-blood
and hemolytic anemia: “Hydrocytosis” and “desiccytosis.” Semin type associated PAS-staining polymorphism of glycoprotein three
Hematol 7:381, 1970 from human erythrocyte membranes. Hum Genet 32:121, 1976