Instability of the Oxy Form of Sickle Hemoglobin and of
Methemoglobin in Isopropanol
THOMAS A. BENSINGER,M.D.,LTC,MD, AND ERNEST BEUTLER, M.D.
Bensinger, Thomas A., and Beutler, Ernest: Instability of the
oxy form of sickle hemoglobin and of methemoglobin in
isopropanol. Am J Clin Pathol 67: 180-183, 1977. Oxygenated
hemoglobin from water-lysed sickle erythrocytes precipitated
more rapidly than hemoglobin from normal cells when exposed to 17% isopropanol in O.lM tris, p\\ 7.4, buffer (the
isopropanol solubility test of Carrell and Kay). Precipitates
became visible after approximately 20 minutes of incubation
with SS, AS, or SC blood, while AC hemoglobin and normal
adult A hemoglobin required 40-60 minutes to precipitate.
Methemoglobin comprised less than 1.5% of the total hemoglobin present in all samples under study. Significant precipitation of methemoglobin did not occur unless methemoglobin
levels were greater than 8%. Solubilization of the precipitate
in 6M urea-barbital butter and subsequent electrophoresis
demonstrated that the insoluble material was sickle hemoglobin (;SS chain). The results of these experiments confirm
previous findings indicating that even fully oxygenated sickle
hemoglobin is unstable. (Key words: Isopropanol; Sickle
hemoglobin; Methemoglobin; Instability.)
CONSIDERABLE ATTENTION has been focused
on the solubility and instability of the deoxy form of
sickle hemoglobin. Recently, Asakura and co-workers1
demonstrated that the oxygenated form of sickle
hemoglobin is more unstable than normal hemoglobin
under the stress of mechanical agitation. This finding
has renewed interest in pursuing other possible differences in the stability of the oxygenated form of
sickle hemoglobin compared with normal hemoglobin.
Unstable hemoglobins cause congenital Heinz-body
hemolytic anemias.14 These traditionally have been
demonstrated by observing increased precipitation on
heating a hemolysate at 50 C in appropriate buffer.5
Recently, Carrell and Kay introduced a screening
procedure that uses 17% isopropanol in tris buffer,
pH 7.4, to induce precipitation of unstable hemo-
Department of Surgery,
Letterman Army Institute of Research,
Presidio of San Francisco, California 94129,
and Department of Medicine,
City of Hope National Medical Center,
Duarte, California 91010
globins.3 Utilizing the isopropanol system in our
laboratory, we found on several occasions that fully
oxygenated hemoglobin from sickle-cell patients
produced a brown to white flocculent precipitate in
approximately 20 minutes, compared with hemoglobin
from normal controls, which did not precipitate until
40-60 minutes had elapsed. This study was undertaken to explore more fully the etiology of this precipitate formation.
Methods
Solutions
1. Isopropanol buffer: isopropanol, 17% by volume,
was prepared in 0.1M tris HCI, pU 7.4 (25 C).:i This
buffer was prepared approximately every week and
kept in a stoppered bottle to prevent evaporation.
2. Barbital buffer: 0.1 mol barbital was dissolved in
700 ml of boiling water, the pH was adjusted to 8.0
with 1 N NaOH, and the solution brought to one
liter with additional deionized water.
3. Urea-barbital buffer: urea, 6 M, was added to the
barbital buffer and 50 p\ of 2-mercaptoethanol per
5 ml of buffer were then added.'3
Experimental Procedures
Blood samples from donors, with various hemoglobin compositions (AA, AS, SS), were anticoagulated with EDTA. It was usually used the day of collection, but if not, it was stored as whole blood at
4 C for as long as five days before being washed
Received March 10, 1976; accepted for publication April 5, 1976.
Address reprint requests to Dr. Bensinger: Department of Surgery,
Letterman Army Institute of Research, Presidio of San Francisco,
California 94129.
The ideas and expressions expressed herein are those of the authors
and are not to be construed as those of the Department of the Army.
180
ISOPROPANOL— UNSTABLE S AND METHEMOGLOBIN
and hemolyzed. A 10% hemolysate was prepared from
thrice-washed erythrocytes according to the method of
Carrell and Kay.:i Membranes were removed by
centrifugation at 20,000 Xg for 20 minutes at 4 C.
The freshly prepared hemolysate was exposed to air
by gentle manual inversion in a small test tube,
followed by adding 200 jtxl of the hemolysate to 2 ml of
isopropanol buffer previously incubated at 37 C for
15 minutes in stoppered tubes. Subsequently, the mixture was inverted gently several times to insure
oxygenation and mixing. The tubes were replaced in a
37 C bath and observed at frequent intervals for the
formation of a flocculent precipitate. Duplicate controls were monitored occasionally to insure all samples
were >97% saturated with oxygen.* Monitoring of
flocculation at 650 nm was also undertaken, utilizing
a Gilford 2400-2 spectrophotometer in which an isopropanol blank was compared with isopropanol-diluted
hemolysates of AA, SS, and SC blood in capped
cuvettes. The loss in light transmission due to turbidity
was then observed over 40-60 minutes.
The precipitate formed in the isopropanol tests was
concentrated by centrifugation at low speed for 3
minutes. The button was solubiljzed in 100 /xl of
urea-barbital buffer and recentrifuged at low speed for
3 minutes. Control globin chains were obtained by the
procedure of Chernoff and Pettit,4 in which cold acid
acetone causes their precipitation, and were resolubilized in approximately 100 /i\ of urea-barbital
buffer.
Urea-barbital starch gels were prepared according
to the technic of Weatherall and Clegg.13 Electrophoresis of the material obtained from isopropanol
and the globin chain precipitates was performed for 22
hours at 250-300 volts and 35-45 milliamps at 4 C.
The gels were sliced in the cold, stained with amido
black for 10 minutes, and destained in methanolacetic acid-water (5:1:5) for 48 to 72 hours.
Methemoglobin solutions were prepared in two
ways. The first method utilized 20 minutes of incubation of A A blood from six donors at room temperature:
once-washed erythrocytes were incubated with an
equal volume of freshly prepared 0.145 M sodium nitrite
solution. 2 The cells were then washed five to six times
in a tenfold excess of 0.9% NaCl, followed by lysis
in distilled water.
In the second method, AA erythrocytes obtained
from four donors were distilled water-lysed and the
hemoglobin produced was washed with Whatman DE
11 cellulose 8 to remove methemoglobin reductase, followed by treatment with a 50% excess of K 3 Fe(CN) B
per mol of hemoglobin monomer. The substrate
* Co-Oximeter Model 182, Instrumentation Laboratory, Inc.,
Watertown, Massachusetts 02172.
70
60
SO
u
Z
40
<
IB
m
30
<
20
BLANK
20
30
TIME (MIN)
FIG. I. AA. SS or SC hemolysates were diluted in isopropanol buffer and incubated at 37 C in quartz cuvettes. The flocculation occurring over 40 minutes was monitored by measuring
the change in light absorption at 650 nm in a Gilford 2400-S
automatic recording spectrophotometer. The SC and SS hemolysates
produced more flocculation and absorbed more light than the AA
control.
was passed over a column of Sephadex G25 equilibrated with 0.1 M tris, pH 7.4, to remove excess
K:!Fe(CN)(i. Methemoglobin in control and study
samples was determined according to the method of
Evelyn and Malloy. 7
Various methemoglobin concentrations (from I to
100%) were made by adding the methemoglobin substrate to freshly prepared hemolysate (all originally
containing less than 1% methemoglobin). The mixture
of methemoglobin and fresh hemoglobin substrate was
added to the tris-isopropanol solution as described
above and incubated for 30-60 minutes in a 37 C
water bath. After incubation, the tubes were centrifuged and the hemoglobin concentration in the supernatant solution was determined optically as cyanmethemoglobin.
Hemoglobin types were identified by utilizing cellulose acetate electrophoresis a t p H 8.9 in t r i s - E D T A borate buffer10 and the results confirmed by starch
gel electrophoresis a t p H 8.6 and 6.8. 13 Fetal hemoglobin was determined by an alkali denaturization
method. 12
Results
Hemoglobin obtained from both sickle-cell disease
patients and donors with sickle-cell trait precipitated
more rapidly in the isopropanol-tris buffer than did
normal control hemoglobin. This finding was consis-
182
A.J.C.I'. • Fchn,;n v l'J77
BENS1NGER AND BEUTLER
a
Glob
1
2
3
4
5
AS
AS
ss
ss
AS
Acet ppt
Isop ppt
Acet ppt
Isop ppt
Acet ppt
FIG. 2. Urea-barbital electrophoresis of isopropanol precipitate or acid acetone extract of hemoglobin. Channels arc numbered
from left to right: (I), AS. acid acetone ppt; (2). AS. isopropanol ppt; (3). SS, acid acetone ppt; (4). SS. isopropanol ppt;
(5), AS. acid acetone ppt. The isopropanol precipitation of S hemoglobin (channels 2. 4) has selective /3s chains precipitated,
in contradistinction to acid acetone precipitation of AS or SS hemoglobin (channels 1. 3. 5). which has random precipitation
of all globin chains (a. ji, /3 s . and -y).
100
o
£
<
3
u
z
90.
80-
2 701
u.
< 60
30 MIN
5
60 MIN
5
°r
o
«> 40
30
0
O
S
n = 6
± SEM
SODIUM NITRITE METHEMOGLOBIN
10
4
6
8
10
25
50
100
% METHEMOGLOBIN ADDED
FIG. 3. Various amounts (0-100%) of sodium nitrite methemoglobin were added to freshly prepared hemolysates of normal adult A
hemoglobin. The amount of hemoglobin remaining in solution after
incubation for 30 or 60 minutes at 37 C in isopropanol buffer is
plotted on the Y axis against the percentage of original methemoglobin present in the solution of the X axis. As can be seen,
as much as 8% of methemoglobin leads to very slight precipitation.
tently observed in hemolysates prepared from blood
samples from five patients with sickle-cell disease and
five donors with sickle-cell trait. Hemolysates from
two C hemoglobin trait (AC) donors did not precipitate
faster than normal A hemoglobin control, and hemoglobin from one SC donor precipitated as rapidly as
those from sickle-cell trait donors. Figure 1 shows the
decrease in light transmission that occurred when
sickle hemoglobin instead of AA hemoglobin from a
normal donor was present in the cuvette at 650 nm.
Figure 2 demonstrates that the precipitate was
composed primarily of the sickle /3 chains (channels
2, 4) and was not the result of random precipitation
of all hemoglobin, as occurs in the acid acetone globin
precipitate (channels 1, 3, 5). Methemoglobin levels in
samples from all patients under investigation were less
than 1.5% of total hemoglobin.
Methemoglobin has been reported to cause falsepositive isopropanol tests 3 ; however, quantitative data
regarding the methemoglobin concentration necessary
for precipitation are not available. Therefore, it was
felt important to determine the effects of increasing
concentrations of methemoglobin. Figures 3 and 4
Vol. 67 . No. 2
ISOPROPANOL—UNSTABLE S AND METHEMOGLOBIN
illustrate the effects of adding increasing concentrations of methemoglobin obtained by sodium nitrite or
K:iFe(CN)(i oxidation to a freshly prepared normal
hemolysate, followed by 30 and 60 minutes of incubation in isopropanol buffer. Little effect is seen until
more than 8% methemoglobin is present, after which
point increasing flocculation occurs. When all of the
pigment is in the methemoglobin form, precipitation
occurs almost instantaneously, and less than 15% of the
hemoglobin remains in solution at the end of the 30minute incubation period. Fetal hemoglobin was 6% or
less in concentration in samples from all donors
except one SS donor, who had 8.2% fetal hemoglobin.
Discussion
This study demonstrates that oxygenated sickle
hemoglobin is less stable than normal hemoglobin
when exposed to isopropanol. Until 1973, it had not
been widely appreciated that sickle hemoglobin in
the oxygenated state is more mechnically unstable
than normal hemoglobin.1 Substitution at the f3K position" may lead to a shift in the overall conformation
of the molecule in such a manner that stress is
placed on it when it is dissolved in a less polar
solution. This stress, being greater than the internal
stability of the molecule, leads to its precipitation.
Substitution of the fin position glutamic acid with lysine," as occurs in hemoglobin C, does not lead to
decreased stability in tris- isopropanol buffer. It is
interesting that hemoglobinu,i(ien, which has been
shown to be a deletion of the /3(! or /37 position
glutamic acid," leads to a clinically significant instability of the hemoglobin molecule with hemolytic
anemia.
While methemoglobin concentrations present during
incubation of the sickle and C hemoglobins in isopropanol were not increased to levels that cause
significant precipitation, it was shown conclusively
that increasing concentrations of methemoglobin will
lead to increased precipitation in isopropanol. It should
be noted that increasing the methemoglobin concentration by adding various amounts of 100% methemoglobin to essentially 0% methemoglobin solutions
may not be the same as a situation in which the
various methemoglobin levels are achieved by direct
oxidation of the hemolysate.
Acknowledgment. Mrs. Suzanne McLaughlin provided technical
assistance.
183
100
X 90
?. 80
70
60
Z
o
;= so
o
z
z
40
30
2 20
O
O
S 10
4
6
8
10
25
% METHEMOGLOBIN ADDED
100
FIG. 4. Various amounts (0- 100%) of K3Fe(CN)B-prepared methemoglobin were added to freshly prepared heniolysates of normal
adult A hemoglobin. Plotting is the same as in Figure 3.
References
1. Asakura T, Agarwal PL. Relman DA, et al: Mechanical
instability of the oxy-form of sickle hemoglobin. Nature
244:437-438, 1968
2. Beutler E, Baluda MC: Methemoglobin reduction: Studies of
the interaction between cell populations and of the role of
methylene blue. Blood 22:323-333, 1963
3. Carrell RW, Kay R: A simple method for the detection of
unstable haemoglobins. Br J Haematol 23:615-619. 1972
4. Chernoff AI, Pettit NM: The amino acid composition of hemoglobin: III. Qualitative method for identifying abnormalities
of the polypeptide chains of hemoglobin. Blood 24:750-756.
1964
5. Dacie JV, Grimes AJ. Meisler A. et al: Hereditary Heinzbody anemia. A report of studies on five patients with
mild anemia. Br J Haematol 10:388-402. 1964
6. deJong WWW, Went LN. Bernini LF: Haemoglobin Leiden:
Deletion of ,86 of 7 glutamic acid. Nature 220:788-790.1968
7. Evelyn KA. Malloy HT: Microdetermination of oxyhemoglobin, methemoglobin and sulfhemoglobin in a single sample
of blood. J Biol Chem 126:655-662. 1938
8. Hegesh E. Calmanovici N. Avion N: New method for determining ferrihemoglobin (NADH methemoglobin reductase)
in erythrocytes. J Lab Clin Med 72:339-344. 1968
9. Ingram VM: A specific chemical difference between the globins
of normal human and sickle-cell anemia haemoglobin. Nature
178:792-794. 1956
Lehmann H, Huntsman RG: Man's Haemoglobins. Second
edition. Philadelphia. J. B. Lippincott, 1974. pp 398-399
Ibid, p 177
Pembrey ME, McWade P, Weatherall DJ: Reliable routine
estimation of small amounts of fetal haemoglobin by alkali
denaturation. J Clin Pathol 25:738-740. 1972
13 Weatherall DJ, Clegg JB: The Thalassaemia Syndromes.
Second edition. Philadelphia. F. A. Davis, 1972, pp 315316
14 White JM: The unstable haemoglobin disorders. Clin Haematol
3:333-356, 1974