Annals o f Clinical & Laboratory Science, vol. 30, no.

2, 2000 191

Chromatographic Measurements of Hemoglobin A2 in Blood

Samples that Contain Sickle Hemoglobin

1 2 3
Masih Shokrani, Falicia Terrell, Ernest A. Turner, and Maria del Pilar Aguinaga
Comprehensive Sickle Cell Center, Meharry Medical College, Nashville, Tennessee; 1Department of
Microbiology, Immunology & Genetics, Department of Pediatrics, 3 Department of Obstetrics & Gynecology

Abstract. In the sickle cell syndromes, Hb A2 measurements aid in the differential diagnosis of sickle cell anemia
from sickle-beta-thalassemia. The purpose of this study is to assess the Hb A2levels in samples containing sickle
hemoglobin (Hb S) by the use of an automated high performance liquid chromatography system (HPLC-
Variant (3-thalassemia Short Program). The blood samples analyzed were from individuals of African descent
living in the state ofTennessee who had either sickle cell trait (Hb AS), sickle cell disease (Hb SS), or sickle cell-
hemoglobin C disease (Hb SC). Interestingly, the Hb A2 levels determined by HPLC were found elevated in
samples containing Hb S. The Hb A^ mean in Hb AS samples (n=l46) is 4.09% (SD + 0.42, range 2.20 to
5.20%); in Hb SS samples (n=33) it is 3.90% (SD 1.08, range 0.60 to 5-90%); and in Hb SC samples (n=27)
it is 4.46% (SD 0.70, range 2.30 to 5.91%). The H b A2 mean by HPLC in normal individuals (Hb AA,
n=70) is 2.57% (SD 0.25, range 2.1 to 3.0%), and the Hb A 2 range in (3-thalassemia carriers is 4 to 9%. Our
results show that the Hb A2 levels in Hb S-containing samples partially overlap with those expected from (3-
thalassemia carriers. The hemoglobinopathy laboratory should be aware of this apparent elevation in Hb A2
levels determined by HPLC in individuals carrying Hb S. Other factors, such as family history and clinical
symptoms, should be taken into account before a diagnosis of sickle cell trait, sickle-beta-thalassemia, or sickle
cell anemia is made.

Keywords: Hemoglobin A2, hemoglobin S, high-performance liquid chromatography, thalassemia diagnosis

Introduction (3-thalassemia when determined by capillary IEF. The

The quantitation of H b A2 is im portant for the
accurate diagnosis of hemoglobin disorders such as
the (3-thalassemias, or hem oglobin variants in
combination with (3-thalassemia (eg, sickle-beta-
thalassemias). Hb A2 is known to be elevated in (3-
thalassemia trait [1], sickle-beta-thalassemia [2], and
sickle cell disease with a-thalassemia [3]. In sickle
cell trait, Hb A 2 determined by microchromatography
has been found slightly elevated [4], and significantly
elevated when determined by capillary isoelectric
focusing electrophoresis (IEF) [5]. However, no
overlap in H b A2 levels was seen between Hb S -
containing samples and samples from patients with
Address correspondence to Maria del Pilar Aguinaga, Ph.D.,
Comprehensive Sickle Cell Center, Meharry Medical College,
1005 Dr. D . B. Todd Jr. Bvld., Nashville, T N 37208; tel 615 327
6528; fax 615 327 5511; e-mail maguinaga@mmc.edu.
purpose of this report is to assess Hb A 2 levels by HPLC
in liquid blood samples containing Hb S. We have
found that Hb A2 levels quantitated by HPLC are
elevated in samples from non-thalassemic individuals
that carry Hb S. We have not observed this increase
in Hb A2 levels in samples that contain only Hb AA.
HPLC reference ranges for Hb A2 in sickle cell trait,
sickle cell anemia, and Hb SC disease should be
established in order to prevent an erroneous association
of these disorders with (3-thalassemia.
Methods
Blood specimens were collected in heparinized
capillaries or in ethylene diaminetetraacetare (EDTA)
microvettes (Sarstedt, Inc., Newton, NC). The criteria
for sample inclusion were: (1) presence of Hb S, (2)
sample age < 7 days after collection, (3) symmetrical

0091-7370/00/0200-0191 $1.00; 2000 by the Association of Clinical Scientists, Inc.

192 Annab o f Clinical & Laboratory Science
Hb A 2 chromatogram peak (Fig. 1), and (4) subjects
from 1 to 60 years of age. Samples from subjects being
treated with hydroxyurea, a medication prescribed for
treatment of sickle cell disease that increases Hb F
levels, were excluded from this study, as were samples
diagnosed as sickle-beta-thalassemia (Fig. 1) or samples
with asymmetrical Hb A2 peak (Fig. 2). Blood
specimens were received in the laboratory by postal
service, courier, or walk-in. Red cell hemolysates were
run by alkaline electrophoresis (pH 8.6) (Helena
Laboratories, Inc., Beaumont, TX) and by isoelectric
focusing electrophoresis (IEF) (Wallace Laboratories,
Inc., Akron, OH) before HPLC analysis. The Bio-
Rad Variant (3-thalassemia Short Program was used
for Hb quantitations. The Variant (Bio-Rad Labora
tories, Inc., Hercules, CA) is a fully automated HPLC
system that separates a hemoglobin mixture into its
individual components by pumping liquids through a
column at high pressure. It is a fast and reproducible
method that can be used to separate and determine
area percentages for various hemoglobins, including
Hb F and Hb A r It can also provide qualitative
determinations of abnormal hemoglobins. The (3-
thalassemia Short Program utilizes the principle of
E lu tio n T im e (m in.)
Fig. 1. Bio-Rad HPLC chromatogram showing a typi
cal Hb A2 peak (symmetrical peak). This patient was
diagnosed as sickle-beta-thalassemia.
cation exchange HPLC. The resin of the column has
a negatively charged surface group that will bind the
positively charged protein (Hb), while the more
negatively charged molecules will elute readily from
the column. A higher cation concentration will be
required to elute the positively charged hemoglobin
from the column. The ionic strength of the elution
buffer mixture is increased by raising the percent
contribution of elution buffer #2. As the ionic strength
of the mixture increases, more strongly retained
hemoglobins elute from the column [8],
Results
For this study, 146 Hb AS samples, 33 H b SS samples,
and 28 Hb SC samples were analyzed by HPLC.
Results showed that in Hb AS samples, the mean for
Hb A^ is 4.09% (SD 0.42, range 2.20 to 5.2%). In
Hb SS samples, the mean for Hb A 2 is 3.90% (SD +
1.08, range 0.60 to 5.90%. In H b SC samples, the
mean for Hb A^ is 4.46% (SD 0.70, range 2.30 to
5.91%). The corresponding hemoglobin quantitations
for Hb F, Hb A, Hb S, and Hb C are shown in Tables
1, 2, and 3 for the sample populations mentioned
Elution Time (min.)
Fig. 2. Bio-Rad HPLC chromatogram showing an
atypical Hb A2 peak (asymmetrical peak, having a
shoulder). This patient was diagnosed as sickle cell
trait.
 HPLC analysis ofH b A 2 in blood that contains H b S 193

above. The reference range for Hb A2in normal adults It has been postulated that, in the presence of
is 2.1 to 3.0%, with a mean of 2.57% [8] using the P-chain hemoglobin variants, the association of the
Bio-Rad Variant system. OC-globin chain with the (3-globin chain variant is
impaired, which fosters the association of CX-chains with
Discussion 8-chains, increasing the Hb A^ levels [10,11]. Other
factors that play a role in this increase are the
The sensitivity of HPLC has significantly contributed accumulation of Hb S adducts and sample aging. The
to the accurate quantitation of hemoglobin variants. Hb S adducts include glycohemoglobin [11] and other
In this report we have established reference ranges for post-translational modification products of Hb S. The
H b A2 quantified by HPLC (Bio-Rad Variant (3- effect of sample aging is observed in the accumulation
thalassemia Short Program) in samples containing of glycohemoglobin in blood samples over time [7].
Hb S. We found that blood samples containing H b S If Hb Scontaining blood samples are stored for 2
usually have increased Hb A2levels when compared to weeks, Hb A2 levels are increased, possibly due to
previously published HPLC reference ranges [8,9]. The accumulation of glycohemoglobin [7,12]. We did not
HPLC manufacturer states that Hb S byproducts could determine glycohemoglobin in our samples. Since the
coelute with Hb A2 [8]. We have not observed increased blood samples used for this study were relatively fresh
H b A2 values by HPLC in normal Hb AA samples. (1 week) and we still observed a marked increase in
Hb A2, we postulate that other factors may be involved
in the increase of Hb A2 in Hb S-containing samples
Table 1: Hemoglobin percentages in Hb AS samples when determined by HPLC.
(sickle cell trait), based on HPLC analysis. Concurrently elevated levels of Hb A2 and Hb F
Hb N Mean SD (%) Range (%) offer a practical way to diagnose carriers o f the
p-thalassemia gene or other Hb disorders that lead to
A 146 53.802.93 (47.70 - 63.60) an increase in this hemoglobin. The HPLC manu
S 146 34.663.71 (21.2 0 - 42.10) facturer indicates that a range for Hb A^ of 4 to 9% is
a2 146 4.090.42 (2.20 - 5.20) typical of heterozygous P-thalassemia [8]. O ur results
F 98 0.980.67 (0.20 - 3.90) show that in samples containing Hb S, the Hb A^ values
are elevated above the normal reference ranges given
by the manufacturer [8] and also above those that we
previously reported in a healthy African American
Table 2: Hemoglobin percentages in Hb SS samples population (Hb A^ range = 0.05 to 3.4%, mean 1.2%)
(sickle cell disease), based on HPLC analysis. [9]. W ith the system used in the present study, Hb A2
Hb N Mean SD (%) Rang; (%) levels in samples containing Hb S partially overlap with
those expected for P-thalassemia carriers. When Hb
S 33 86.743.62 (80.40 -- 94.70)
A2 quantitations in Hb Scontaining samples are
A, 33 3.901.08 (0.60 -- 5.90)
reported, it should be recognized that the apparently
F 33 6.802.95 (1 .3 0 - 11.40)
elevated Hb A2 values may well be within the normal
range for those samples. The elevated Hb A2 values
should be correlated with the clinical picture and family
Table 3: Hemoglobin percentages in Hb SC samples
studies to help establish the diagnosis.
(hemoglobin SC disease), based on HPLC analysis.

Hb N Mean SD (%) Range(%) Conclusions

S 28 46.461.27 (43.70 - 49.20)
C 28 44.251.74 (40.40 - 46.90) The HPLC Hb A2 levels in subjects older than 1 year
27 4.46+0.70 (2.3 0 -5 .9 1 ) and younger than 60 years of age having sickle cell
A,
F 27 2.17+1.51 (0.5 0 -6 .1 0 ) trait (Hb AS) or sickle cell disease (Hb SS and Hb SC)
have been assessed. These results confirm the elevation
194 Annals o f Clinical & Laboratory Science
of Hb A2 levels in individuals with Hb AS, Hb SS,
and Hb SC, as determined by the BioRad Variant
HPLC System. We conclude that in Hb AS, levels of
H b A2lower than 5.2% are normal. In Hb SS and Hb
SC samples, levels of Hb A2lower than 5.9% may also
be normal if there is no other sign of P-thalassemia.
If further increased Hb A 2 levels are found in a sickle
cell disease patient, one should also consider family
history, clinical symptoms, drug treatment, and sample
aging before the diagnosis o f an associated Pi-
thalassemia disorder is made.
Acknowledgments
The authors thank Dr. Mark J. Koury for critical
reading of the manuscript, Ms. Mai Williamson for
technical assistance, Ms. Deborah Majors for statistical
analyses, and Ms. Tasha White and Ms. Carolyn Tansey
for typing the manuscript. Supported by grants from
the Tennessee Department of Health and Environment
(GR-6 10510-600), and the National Heart, Lung and
Blood Institute (P60 HL-38737 and K14 HL-03141).
References
1. Huisman TH J. Levels of Hb A2 in heterozygotes and
homozygotes for beta-thalassemia mutations: influence
o f mutations in the CACCC and ATAAA motifs of the
beta-globin gene promoter. Acta Haematologica 1997;
98:187-194.
2. Sweeting I, Serjeant BE, Thomas PW, Serjeant GR.
Microchromatographic quantitation of H b A2 levels
in phenotypes o f sickle cell-beta(+) thalassemia. J
ChromatogB 1997;700:269-274.
3. Balias SK, Gay RN, Chehab FE Is H b A2 elevated in
adults with sickle-alpha-thalassemia (beta(S)/beta(S);
-alpha/-alpha)? Hemoglobin 1997;21:405-450.
4. Huisman THJ, Shroeder WA, Brodie AN, Mayson SM,
JakwayJ. Microchromatography of hemoglobins, III:
a simplified procedure for the determination of Hb Aj-
J Lab Clin Med 1975;86:700-702.
5. Craver RD, Abermanis JG, Warrier RP, Ode DL,
Hempe JM. Hemoglobin A2 levels in healthy persons,
sickle cell disease, sickle cell trait, and P-thalassemia by
capillary isoelectric focusing. Am J Clin Path 1997;
107:88-91.
6. Ou CN, Buffone GJ, Reimer GL, Alpert AJ. High-
perform ance liquid chrom atography o f hum an
hemoglobins on a new cation exchange. J Chromatog
1983;266:197-205.
7. Kutlar A, Kutlar F, Wilson JB, Headlee M G, Huisman
TH J. Quantitation of hemoglobin components by
high-performance cation-exchange liquid chrom
atography: its use in the diagnosis and in the assessment
of distribution of hemoglobin variants. Am J Hematol
1984;17:39-53.
8. BioRad. Variant Beta-Thalassemia Short Program
Procedure Manual. BioRad Laboratories, Inc., Hercules,
CA, 1994.
9. Roa PD, Turner EA, Aguinaga MdP. Reference ranges
for hem oglobin variants by H P L C in A frican
Americans. Ann Clin Lab Sci 1995;25:228-235.
10. W hitten WJ, Rucknagel DL. The proportion of Hb
A2 is higher in sickle than in normal homozygotes.
Hemoglobin 1981;5:371-378.
11. Huisman, THJ. Chromatographic separation of Hb
A2 and quantities of Hb A j in patients with AC trait,
CC disease, and C-beta thalassemia. Clin Chim Acta
1972;40:159-163.
12. Suh DD, Krauss J S, Bures K. Influence of Hb S adducts
on hemoglobin Hb A2 quantification by HPLC. Clin
Chem 1996;42;1113-1114.