CLIN.
CHEM.
20/6,
666-669
(1974)
Screening Cord Bloodsfor Detection of
Sickle Cell Disease in Jamaica
Beryl E. Serjeant,1 Miriam Forbes,1 Leslie L. Williams,2 and Graham R. SerjeanU
A cord-blood screening program, designed primarily
for detecting sickle cell disease, has been in operation for seven months (8000 samples)
at a large
maternity
unit in Kingston, Jamaica.
We describe
techniques
of cord-blood
collection and electrophoretic investigation on both cellulose acetate and agar
gel. These methods appear to give rapid, valid results at minimal expense and are well adapted to
screening large populations.
Additional Keyphrases: screening
diagnostic
electrophoresis on solid media
gene frequency
I
aids
I
Early diagnosis of genetic disease affords an opportunity for early therapy and an approach to prevention
of further cases through genetic counselling.
Detection of sickle cell disease at birth and its prospective study gives valuable information
on the development of hematological
changes and their clinical sequels. In this way, detection of sickle cell disease from symptoms alone can be avoided, and valid
data can be collected on the natural history of the
disease.
Since a-chain synthesis accounts for only a small
percentage
of the total chain synthesis at birth, i3chain variants
such as hemoglobins
S and C are
present in small amounts in umbilical
cord blood.
Furthermore,
identification
of the correct genotype
requires
that techniques
be capable of detecting
small amounts of the Hb A that would distinguish
sickle cell trait from homozygous sickle cell disease.3
These considerations
preclude the use of sickling and
solubility tests for Hb S. Standard methods of hemoglobin electrophoresis
will demonstrate
Hb S, but
may not adequately
detect small amounts of Hb A in
the presence of large amounts
of Hb F. Electrophoresis on agar gel will demonstrate
the presence of both
hemoglobins
S and A, which clearly separate from
Hb F, but this has generally been a slower, more
time-consuming
technique,
not well suited to processing large numbers of samples, and rarer hemoSickle Cell Section, MRC Epidemiology
West Indies,1 and Victoria JubileeHospital,
Nonstandard
abbreviations
tris(hydroxymethyl)aminomethane;
used:
Hb,
EDTA,
traacetate.
Received Feb. 27, 1974; accepted
666
CLINICAL
CHEMISTRY,
Unit, University of the
Kingston, Jamaica.2
hemoglobin;
Tris,
ethylenediaminete-
April 3, 1974.
Vol. 20, No.6,
1974
globin variants may go undetected.
Previous studies
of screening cord bloods for sickle cell disease have
worked with these limitations,
by using either agar
gel electrophoresis
alone (1, 2) or agar gel electrophoresis preceded by screening with paper electrophoresis (3).
Recently,
Schneider
(4) has developed
a buffer
system that separates
hemoglobins
A, F, and S on
cellulose acetate, but small amounts of Hb A, indicative of sickle $-thalassemia,
may not be distinguishable and other hemoglobin variants may be mistaken
for Hb S and Hb C. The procedure we used entails
preliminary
screening of all samples on cellulose acetate membrane
with a Tris-EDTA-borate
buffer at
pH 8.4 (4), followed by investigation
of all abnormal
hemoglobin
bands on agar gel. A system of agar gel
electrophoresis
have been developed that allows investigation of as many as 100 specimens in 2 h.
Materials and Methods
Apparatus
Hemoglobin
electrophoresis
was performed
in a
standard
horizontal tank (Kohn Model U77, Shandon Southern Instruments
Ltd., Camberley,
Surrey,
UK) on 78 mm X 150 mm cellulose acetate membranes (Shandon
Celagram),
with a Vokam power
pack (Shandon Southern).
Reagents
Hemolyzing
reagent
(4). The stock solution
is
tetrasodium
ethylenediaminetetraacetate,
40 g/dl.
For the working solution, mix 0.3 ml of stock solution and 120 ml of distilled water, and add 2 ml of
potassium
cyanide solution (50 g/liter) to prevent
methemoglobin
formation
during
electrophoresis.
Store the solution in a brown glass bottle at room
temperature.
Prepare it freshly each week, because
the cyanide deteriorates.
Tris-EDTA-borate
buffer
(pH 8.4)
(4). Tris(hydroxymethyl)methylamine,
10.2 g; ethylenediaminetetraacetic
acid, 0.6 g; boric acid, 3.2 g; and distilled water to 1.00 liter.
Store the buffer at toom temperature.
It may be
used for about 10 runs in the electrophoresis
tanks,
then should be changed because of loss from evaporation.
case did this cause diagnostic confusion. Inadequate
mixing of the cord blood with the anticoagulant
occasionally resulted in clotted specimens,
but these
Wash for the membrane.
Dilute 50 ml of glacial
could still be used satisfactorily
for making hemolysacetic acid to 1.00 liter with water.
ates. In each labor ward, duplicate,
consecutively
Store the fixative and wash solution at room temnumbered,
self-adhesive
labels were available,
one
number being applied to the blood tube and the
perature.
Fixative can be re-used until exhausted,
other to the data card, which was then detached
but discard the wash after use.
from the record. Blood specimens were stored in a
Benzidine
stain: Benzidine
base, 1 g; sodium nitray at room temperature,
with the data cards, and
troprusside,
1 g; glacial acetic acid, 50 ml; and discollected early each morning. Specimens
and cards
tilled water to 1.00 liter.
were then checked against the daily records of delivStore stock solution in a brown bottle at 4 #{176}C.
Preeries, and missing samples were collected (by heel
pare the working stain by adding three drops of hyprick) into small plastic tubes containing
dry hepadrogen peroxide (3% solution) to 25 ml of stock solurin or potassium oxalate.
tion immediately
before use. If the working stain is
Electrophoretic
procedures.
Hemolysates
were
then warmed to room temperature,
staining is more
made by adding one drop of whole blood to 0.5 ml of
rapid.
working hemolyzing reagent, mixing well, and allowCitrate buffer for agar gel (pH 6.0) (3). Trisodium
ing the mixture to stand for 10 minutes. The concencitrate, 147 g; citric acid, 4.3 g; and distilled water to
tration of hemolysates
(about 0.25 g/dl) was matched
1.00 liter.
visually
and
more
blood
was required from samples
Store the stock buffer at 4 #{176}C.
Dilute 60 ml to 600
with
low
hemoglobin
concentrations.
Less concenml with distilled water to make a working buffer; use
trated hemolysates
separated well, but minor bands
50 ml of this to make the agar gel, and adjust the rewere not always distinct, whereas more concentrated
maining 550 ml to pH 6.2 with citric acid solution
hem#{243}lysates
ran slowly and separated poorly.
(200 g/liter).
Use this buffer for electrophoresis.
It
may be recovered and re-used for as many as 10
Electrophoresis
was performed in a standard horizontal tank containing 500 ml of Tris-EDTA-borate
times without deterioration.
Agar gel. Dissolve 0.5 g of agar (Difco Bacto-Agar;
buffer at room temperature.
A cellulose
acetate
membrane
was soaked in Tris-EDTA-borate
buffer
Difco Laboratories,
Detroit, Mich. 48232) in 50 ml of
in a shallow tray. It was laid on the surface before it
working buffer by boiling gently until the solution is
was submerged,
because premature
submerging
of
clear. Cool to about 60 #{176}C
and then pour onto a
the membrane
before it had been permeated
comlevel, warmed plate of heavy glass (20 X 23 cm) to
pletely from the lower side led to trapping of air (inproduce an agar gel that is about 1 mm thick. When
by a mottled
appearance)
rendering
the
cold, wrap the gel in plastic film and store it at 4 #{176}Cdicated
membrane useless. After soaking, the membrane was
until needed. Before use, warm it to room temperablotted thoroughly between layers of Whatman No. 3
ture, taking care to ensure that the surface is not
chromatography
paper and left between the layers to
wet. Plates may be prepared in batches, and give exprevent further drying. The hemolysates
were taken
cellent results for as long as four weeks if wrapped
up on a 16-tooth
multiple
applicator
(Shandon
carefully and stored at 4 #{176}C.
Fixative
for agar gel. Methanol,
700 ml; and acetic
Southern
Instruments)
and applied 1 cm to the cathodal side of the center of the membrane,
thus alacid (200 mi/liter),
300 ml. Store at room temperalowing a longer path of migration.
Cord blood conture; may be re-used.
taining hemoglobins A, F, and S was applied in posiMethods
tions 1 and 16, as controls, allowing 14 specimens to
be run on each occasion. A nonwettable
surface such
Cord-blood
collection.
Cord-blood specimens were
collected at the Victoria Jubilee Hospital,
a large
as Parafilm underlying
the application
site preventmaternity
unit in the city of Kingston,
Jamaica,
ed the sample from diffusing through into the filterpaper blotter. The membrane was then placed in the
where 14 000-15 000 babies are delivered annually
electrophoresis
tank with a bridge width of 5 cm and
and which serves the entire metropolitan
area. During
10 X 23 cm filter paper wicks (Whatman
No. 3 chroadmission of the patient, the clerical staff filled out
matography
paper). A 1-cm strip of cellulose acespecially printed cards with essential information
for
tate, soaked in buffer and blotted,
was placed at
tracing the mother (name, address,
and hospital
number) and the cards were stapled to the front of
each end of the membrane
to prevent curving of the
the patients’ records.
electrophoretic
run and aided interpretation
of results.
A
voltage
of
350
V
was
applied
for
20
mm,
or
Immediately
after delivery, blood from the placental end of the umbilical cord was collected in 10-ml
until there was good separation.
The membrane
was
plastic tubes containing,
as anticoagulant,
dry lithithen fixed in the fixative for 3 mm, washed three
um heparin
(Turner-Staynes
Laboratories
Ltd.,
times in the wash solution
at 2-mm intervals,
Bishop Auckland, Co. Durham, England). Under the
drained well, and stained with benzidine until all hemoglobin bands were clearly visible. After staining,
conditions, the cord could not be cleaned carefully to
the membrane
was washed three times in distilled
avoid contamination
by maternal
blood, but in no
Fixative
for cellulose
50 g of trichloroacetic
1.00 liter.
acetate
acid
Dissolve
and dilute to
membrane.
in water
CLINICAL
CHEMISTRY,
Vol. 20, No.6,
1974
667
water acidified with acetic acid. Results were read
while the membrane
was wet, and it was then dried
between sheets of filter paper and mounted.
When
dry, the bands of hemoglobin
in the stained membrane faded somewhat, but contrast was restored by
soaking in acidified water.
This electrophoretic
system separates hemoglobins
A, 5, and C from Hb F, in most cases allowing distinction of AA, AS, SS, and SC genotypes.
Occasionally the presence of small amounts of Hb A was
in doubt and, to confirm the SS genotype and to
identify other variants, we also studied by agar gel
electrophoresis
all samples that evidenced
electrophoretic abnormality
on cellulose acetate. This was
done according to the method of’ Metters et al. (3),
with modifications.
A horizontal electrophoresis
tank
(Kohn Model U 77) was filled with 500 ml of working
buffer and refrigerated
at 4 “C. Hemolysates
were
made as described previously, although four drops of
blood were added to 0.5 ml of hemolyzing reagent to
give a concentration
of about 3-4 g of hemoglobin
per deciliter.
Wicks measuring
1 mm X 4 mm
(Whatman
No. 3 chromatography
paper)
were
dipped into the hemolysates
and placed on the surface of the agar gel at 1.0-cm intervals, starting at
least 1 cm from the edge of the plate. Two lines of
wicks, 5 cm apart, were applied, allowing 40-50 samples to be run, but as many as four lines of wicks and
100 samples may be applied in this manner.
After 20
mm, the wicks were removed, care being taken not
to break the agar surface. The plate was then placed
in the electrophoresis
tank at 4 #{176}C
with filter paper
wicks 10 x 23 cm overlapping
both ends to a depth
of 4.5 cm (3 cm from the lines of application).
A constant current of 50 mA was applied for 60 mm, which
separated
hemoglobins
F and C by about 2 cm. The
gel was removed to a shallow tray, fixed in acid
methanol for 10 mm, drained, and stained with benzidine. When hemoglobin
bands had stained clearly
(5-10 mm) the plate was washed in two changes of
acidified distilled water and left to soak overnight.
After a further wash, the gel was gently slid off onto
a piece of thin cardboard that had been soaked for a
minute in the acidified water. Slow drying at room
temperature
kept the cardboard from curling.
Results
Table 1 shows an analysis of gene frequency in the
first 8000 samples of cord blood. Prediction
of the
expected
phenotype
frequency
from this gene frequency is compared
with the observed
values in
Table 2. The close correlation between these figures,
especially in cases of sickle cell disease, indicates the
accuracy of the diagnostic methods used.
Figure 1 illustrates
a cellulose acetate strip. Hemoglobin S consistently
separated from Hb F into a
well-defined
band visible at a concentration
of less
than 5% of the total Hb. Usually, the Hb A band also
separated well from Hb F, although a trail of hemoglobin linked the two bands. Blood specimens from
cases of homozygous sickle cell disease have no Hb A
668
CLINICAL
CHEMISTRY,
Vol. 20, No. 6, 1974
Table 1. Hemoglobin Gene Frequencies in 8000
Newborn Infants in Jamaica
Genes
No. observed
A
14727
941
S
C
$.thal
309
2
1
20
HPHF
Variants
16000
Relative frequency
0.9204375
0.0588125
0.0193125
0.0001250(0.004)”
0.0000625
0.0012500
1.0000000
(0.0005)”
Because the techniques
used do not allow detection of the
traits for -thaIassemia
and hereditary
persistence
of fetal
hemoglobin
in the cord blood, it is almost certain that these
have been underestimated.
The bracketed
figures are those
predicted by assuming
a 4.thalassemia
trait frequency
of 0.8%
(5) and an hereditary
persistence
of fetal hemoglobin
trait
frequency of 0.1% (6).
Table 2. Observed Phenotype Frequencies
Compared with Values Predicted from
Gene Frequency
Phenotype
Expected
relative frequency
AA
AS
AC
0.847205
A/thaI
0.000230(0.007364)’
A/HPHF
A/Variant
SS
SC
S/thaI
S/HPHF
S/Variant
CC
0.000115 (0.000920)”
0.108267
0.035552
0.002301
0.003459
0.002272
No.
observed
No.
predicted
6782
862
6777.64
280
?
866.14
284.42
20
(58.91)”
(7.36)”
18.41
28
19
27.67
18.18
?
0.000015(0.000471)”
2
0.000007(0.000059)”
0.000147
1
1
0.000373
5
1.18
2.98
C/thaI
0.000005(0.000155)”
0
(1.24)”
C/HPHF
C/Variant
Others
Total
0.000002(0.000019)”
0.000048
0.000002
1.000000
0
0
(0.15)”
0.39
“See
0
(3.77)”
(0.47)”
0.01
8000
footno te in Table 1.
and also a heavier Hb S band than in cases of the
trait.
Occasionally
diagnostic
confusion
arose in
cases where a weak Hb S band was present and no
Hb A was visible. In these cases, electrophoresis
on
agar gel almost always confirmed
the presence of
small amounts of both Hb A and Hb 5, and we assumed that these cases represented
the sickle cell
trait with late onset or a low extent of i9-chain synthesis. This problem was more common in premature infants and in one infant, born two months prematurely,
repeated
electrophoreses
on agar gel at
four and eight weeks were required before significant
amounts of Hb S were clearly present in the absence
of Hb A. In cases of the trait, the concentrations
of
hemoglobins A and S were usually equal on agar gel
electrophoresis,
but occasionally
the Hb S band was
clearly stronger and these cases were assumed to be
developing the Hb A type of sickle 3-thalassemia.
‘!t!TiVfflflflt’T’t?-
-‘
-origin
(A)
C-.,-
Baris-
c-
‘!!flflTtt!!ttTlE!’
J.I
(B)
21 211219 212121
+
+
Fig. 1. Hemoglobins
on a cellulose acetate electrophore-
sis strip
Control AFS specimens denoted at each end of the strip are c. From the
left, specimens 1-14 are Barts AF, AF, AFC, AF. AFS. AFS, AF, F5,
AF, AF, Barts AF, AF, AF, and A2, and AFS
Fig. 2. Agar gel electrophoresis
demonstrated
acetate
electrophoretic
of 42 specimens that
abnormality on cellulose
Sickle celltrait pattern of SAF at (A) 3-7. Hb C trait pattern of CAF at
(A) 2, SS disease represented by SF pattern at (B) 16, SC disease represented by CSF pattern at (B) 19. SIckle 4-thalassemia with Hb A appears at (A) 10 (note that the Hb S band Is stronger than the Hb A
band). The AF patterns at (A) 8 and (A) 17 appeared as AFS on cellu-
lose acetate and represent hemoglobin variants
Electrophoresis
on agar gel was invaluable for confirming the absence of Hb A and for detecting the
very small amounts of Hb A in cases presumed to
represent the Hb A-containing
type of sickle -thalassemia.
All specimens
demonstrating
electrophoretic abnormalities
were examined by this technique,
to confirm the AS and AC genotypes and to detect
other variants. Examples of two such variants, which
appeared
as the sickle cell trait when examined on
cellulose acetate, are shown in Figure 2.
Discussion
Homozygous sickle cell disease, the non-Hb A type
of sickle $-thalassemia,
and the syndrome of double
heterozygosity
for the gene for Hb S and the hereditary persistence
of Hb F (Hb S/HPHF)
could not be
differentiated
in cord-blood specimens by these techniques. When possible, electrophoretic
investigation
and sickling tests on the blood from both parents
were performed to confirm the diagnosis. When this
is not possible the Hb S/HPHF
syndrome
should
be confirmed by acid elution studies, and the nonHb A type of sickle 3-thalassemia
by Hb A2 quanti#{149}
tation at a later age.
The present system combines the advantages
of
previous techniques.
A presumptive
diagnosis may
be made rapidly with the cellulose acetate method,
so that the parents can be informed and blood taken
for baseline hematology within 24 h of birth. The diagnosis was confirmed by electrophoresis
on agar gel
at weekly intervals:
this revealed other hemoglobin
variants traveling in the position of hemoglobins
S
and C on cellulose acetate, which would be missed in
a program relying solely on screening by agar gel.
Furthermore,
rapid processing of many samples in the
agar gel method is economical of technician time.
We have investigated
8000 samples during the first
seven-month
period, an average of 35 cord-blood
specimens daily. The cost of specimen collection and
cellulose acetate
electrophoresis,
including
capital
equipment
and consumables,
has been less than 6c
(U. S.) per specimen and the labor of one part-time
technician,
including special weekend coverage, has
averaged 15 per specimen.
In communities
where sickle cell disease constitutes a significant
health problem, the diagnosis at
birth is feasible and reasonable, and offers benefits in
the subsequent
management
of patients
with the
disease.
We thank Mrs. Tern Hilton and Mrs. Helen Trought for technical assistance,
especially weekend coverage, and the clerical
and nursingstaff of Victoria Jubilee Hospital, Kingston,fortheir
cooperation and interest in the study.
References
1. ‘Van Baelen, H., Vandepitte.,
J., and Eeckels, R., Observations
on sickle cell anaemia and haemoglobin Bart’s in Congolese neonates. Ann. Soc. Beig. Med. Trop. 49, 157 (1969).
2. Pearson, H. A., O’Brien, R. T., McIntosh, S., Aspnes, G. T.,
and Yang, M-M., Routine screening of umbilical cord blood for
sickle
celldiseases. J. Amer. Med. Ass. 227,420(1974).
3. Metters,J. S.,Huntsman, R. G., and Yawson, G. I.,The use
of the cord blood sample for the detection of sickle cell anaemia
in the newborn. J. Obstet. Gynaecol. Brit. Commonw.
77, 935
(1970).
4. Schneider,R. G., Developments in laboratory
diagnosis. In
Sickle Cell Disease, Diagnosis, Management,
Education, and Research, H. Abramson, J. F. Bertles, and D. L. Wethers, Eds. C.
V. Mosby Co., St. Louis, Mo., 1973, p 230.
5. Ahem, E. J., Swan, A. V., and Ahern, V. N., The prevalence
of the rarer inherited haemoglobin
defects in adult Jamaicans.
Brit. J. Haematol. 25, 437 (1973).
6. Conley, C. L., Weatherall, D. J., Richardson, S. N., Shephard,
M. K., and Charache, S.,Hereditaryper;istenceof fetal hemoglobin: A study of 79 affected persons in 15 Negro families in Baltimore. Blood 21, 261 (1963).
CLINICAL
CHEMISTRY,
Vol. 20, No. 6, 1974
669