Phenotypic Expression of Known and Novel Hemoglobin A2-Variants, Hemoglobin A2-Mae Phrik [Delta 52(D3) Asp > Gly, HBD:c.158A > G], Associated with Hemoglobin E [Beta 26(B8) Glu > Lys, HBB:c.79G > A] in Thailand

The interactions of δ-globin variants with α- and β-thalassemia or other hemoglobinopathies cause complex thalassemic syndromes and potential diagnostic problems. Understanding the molecular basis and phenotypic expression is crucial. Four unrelated Thai subjects with second hemoglobin (Hb) A2 fractions were studied. A standard automated cell counter was used to acquire initial hematological data. Hb analysis was carried out by capillary electrophoresis (CE) and high-performance liquid chromatography (HPLC) assays. Globin gene mutations and haplotype were identified by appropriate DNA analysis. An allele-specific polymerase chain reaction method was developed to provide a simple molecular diagnostic test. Hb analysis revealed a Hb A2 variant in all cases. DNA analysis of the δ-globin gene identified the Hb A2-Melbourne [δ43(CD2)Glu > Lys] variant in combination with Hb E in three cases. Analysis of the remaining case identified a novel δ-Hb variant, namely Hb A2-Mae Phrik [δ52(D3)GAT > GGT; Asp > Gly], found in association with Hb E and α+-thalassemia, indicative of the as yet undescribed combination of triple heterozygosity of globin gene defects. An allele-specific PCR-based assay was successfully developed to identify this variant. The β-haplotype of the Hb A2 Mae-Phrik allele was strongly associated with haplotype [+ − − − − ± +]. This study advanced our understanding of the phenotypic expression of known and novel δ-Hb variants coinherited with other globin gene defects, routinely causing problems with diagnosis. Therefore, knowledge and recognition of this Hb variant and molecular assessments are crucial to improving diagnosis.


Introduction
Hemoglobin A 2 (Hb A 2 ) is one of the forms of hemoglobin (Hb) that is produced in normal adult red blood cells. The Hb A 2 tetramer is composed of two α chains and two δ chains (α 2 δ 2 ) and is naturally expressed at a low level of 2.5-3.5% of the total hemoglobin in healthy adult life [1]. An increase in Hb A 2 levels, usually to above 4%, is a potential marker for diagnosing the β-thalassemia (β-thal) trait [2]. δ-Hb variants (Hb A 2 variants) result from nucleotide substitutions or deletions in either coding or non-coding regions of the δ-globin gene, structural alteration, or instability and may consequently affect the δ-globin chain synthesis. Since the δ-globin chain is synthesized at a reduced rate, the number of variants within it that are detected by several methods is far lower than that of the αor β-globin genes.
Recently, about 97 structural δ-globin gene variants have been described worldwide, varying in different regions and different ethnic populations (http://globin.cse.psu.edu (accessed on 17 April 2022)) [3]. Nearly all are silent and have an innocuous clinical presentation in heterozygous and even homozygous cases. Although most are non-pathological variants, coinheritance with βor α-thalassemia or other variants results in diverse clinical syndromes. This may lead to diagnostic problems, as the amount of Hb A 2 observed is often reduced to a value within the range usually observed in normal individuals [4][5][6]. These δ-Hb variants are especially prevalent in the Mediterranean region and are occasionally seen in other regions such as Southeast Asia [7][8][9]. In Thailand, where thalassemias and Hb variants are prevalent, about 20-30% of the population carry the α-thalassemia gene, reaching 30-50% in the northern part of the country, whereas about 3-9% are β-thalassemia carriers. The average frequency of Hb E (HBB:c.79G > A) is 20-30%, reaching 50-60% at the junction of Thailand, Laos, and Cambodia. The prevalence of the Hb constant spring (CS) (HBA2:c.427T > C) or Hb Paksé (HBA2:c.429A > T) variants is 1-8% [10].
Despite the fact that these variants are readily detected by high-performance liquid chromatography (HPLC) and capillary electrophoresis (CE) by the observation of the extra Hb A 2 fraction split from normal Hb A 2 , these fractions show similar mobility when assessed by both HPLC and CE assays. Thus, accurate identification and differential diagnosis should be performed by DNA analysis. Moreover, the clinical consequences and significance of many δ-Hb variants when coinherited with αor β-thalassemia or other hemoglobinopathies remain unrecognized. To fully understand the variants' hematological manifestations and molecular basis, we determined the mutations responsible for the novel δ-Hb variant, namely Hb A 2 -Mae Phrik, detected based on different Hb A 2 fractions on electropherograms or chromatograms, and evaluated their influence on the hematological profile alongside the development of a simple molecular diagnostic tool.

Subjects and Hematological Studies
After informed consent was obtained, peripheral blood anticoagulated with ethylenediaminetetraacetic acid was collected from four unrelated Thai individuals (S1-S4) and the parents of subject 1 (S1). The subjects were initially diagnosed as Hb E heterozygotes and were also suspected of having δ-globin structural variants due to the presence of a second Hb A 2 fraction, referring to the School of Allied Health Sciences, University of Phayao, for further Hb and DNA analyses. Hematological data were collected using an automated blood cell counter (Coulter UniCel DxH 800; Beckman-Coulter Co., Miami, FL, USA). Hemoglobin analysis was performed by capillary electrophoresis (Capillarys 2 Flex Piercing; Sebia, Lisses, France) and cation exchange HPLC on two dedicated devices (VARIANT II TM , β-Thalassaemia Short Program; Bio-Rad Laboratories, Hercules, CA, USA; and Premier resolution from Trinity Biotech, Bray, Ireland, USA). Iron deficiency status was determined by measuring the levels of serum iron (SI), serum ferritin, total iron-binding capacity (TIBC), and transferrin saturation (Tsat) using an automated immunoassay analyzer (ARCHITECT ® i2000SR, Abbott Laboratories, Abbott Park, IL, USA).

β-Globin Gene Haplotype Analysis
Globin gene haplotypes based on seven polymorphic restriction sites within the β-globin gene cluster, including HincII at 5 -ε, HindIII at G γ, HindIII at A γ, HincII at ψβ, HincII at 3 ψβ, AvaII at 5 β, and BamHI at 3 β, were determined using PCR and then restriction digestions, as described previously [23]. The haplotypes were documented as + or − according to the presence or absence of the respective restriction enzyme sites.

Development of An Allele-Specific Polymerase Chain Reaction Assay (ASPCR) for Identification of the Hb A 2 -Mae Phrik Variant
A novel δ-globin gene mutation (GAT > GGT) at codon 52 of exon 2 was identified in subject 1 (S1) and his mother, generating the Hb A 2 -Mae Phrik variant. This variant was confirmed using a newly developed ASPCR technique. The δ 52A > G allele-specific reverse primer SP35 (5 -AGGGTTGCCCATAACAGCAC-3 ) was used with forward primer G58 (5 -AGGGCAAGTTAAGGGAATAG-3 ) located 5 upstream to produce a 500 bp fragment specific to the δ 52A > G allele. As an internal control for the PCR amplification, the 713 bp fragment was generated using primers G58 and F16. The PCR reaction mixture (50 µL) contained 50-200 ng of genomic DNA, 30 pmoles of primers G58 and F16, 7.5 pmoles of primer SP35, 200 µM of dNTPs, and 1.0 unit of Taq DNA polymerase (Vivantis Technologies, Selangor Darul Ehsan, Malaysia) in 10 mM Tris-HCl (pH 9.1) buffer, 50 mM KCl, and 0.1% TritonTM X100. The amplification reaction was carried out on a thermal cycler (Cyclerus personalis, Bio-Rad, USA). After an initial denaturation step of 94 • C for 3 min, the reaction was followed by 35 cycles of 94 • C for 1 min, 67 • C for 45 s, 72 • C for 1 min, and final extension at 72 • C for 10 min. The amplified product was analyzed by electrophoresis on a 1.5% agarose gel and visualized under UV light after ethidium bromide staining.

Results
The four unrelated subjects were initially diagnosed with heterozygous Hb E, coinherited with Hb A 2 defects. Hematological characteristics, iron profiles, and globin genotypes of the four subjects (S1-S4) are summarized in Table 1. Hemoglobin analysis of S2, S3, and S4 using automated HPLC on a VARIANT II TM system reveals a major peak of Hb A and Hb E and a small peak resembling the Hb A 2 -variant at a retention time of 4.73-4.75 min ( Figure 2F), with the amounts ranging from 1.3 to 1.7%. Hb analysis using a capillary electrophoresis system showed similar patterns in all cases, in which the Hb A 2 -variant was clearly separated in zone 1 ( Figure 2C), with the amounts ranging from 1.3 to 1.7%, while the Hb A 2 levels were 1.0-1.4%, lower than those observed in heterozygous Hb E (3.8 ± 0.3%). Automated HPLC using the Premier Resolution system, which can separate Hb E and Hb A 2 , was recently installed in our laboratory during the study period. Hb analysis with this system was performed on the blood specimens of S2, S3, and S4. As shown in Figure 2G, Hb A, E, A 2 , and A 2 -variant were clearly separated on this system, allowing calculation of their amounts. The Hb A 2 -variant was eluted from the column at a retention time of 6.401-6.428 min, occupying the S zone of the chromatogram. The levels of Hb A 2 -variants were 1.4% in S2, 1.8% in S3, and 1.9% in S4. In all cases, DNA analysis by ASPCR revealed a G-A substitution at codon 43 of the δ-globin gene to be responsible for Hb A 2 -Melbourne, which has been previously documented in Southeast Asia [6,11,17]. Additionally, the analysis found that the subjects were positive for the β E -globin gene, while none of the common α-thalassemia genes mentioned above were observed. Consequently, the subjects were double heterozygotes at Hb A 2 -Melbourne and Hb E.    S1 was a 34-year-old male who attended an antenatal care clinic. Complete blood count and iron deficiency profiles revealed slightly decreasing mean corpuscular volume (MCV) and mean corpuscular hemoglobin (MCH) but no anemia or iron deficiency ( Table 1). The Hb analysis using HPLC on the VARIANT II TM system revealed a major peak of Hb A (64.6%) and Hb E (24.2%) and a small peak resembling Hb A 2 -variant (1.4%), with a retention time of 4.46 min on the HPLC chromatogram ( Figure 2E). Additionally, analysis using capillary electrophoresis showed 75.5% Hb A, 22.2% Hb E, 1.1% Hb A 2 , and 1.4% probable HbA 2 -variant migrating toward the cathode lower than Hb A 2 , which was at the electrophoretic zone 1 ( Figure 2B). The marked appearance of a small Hb peak on both the electropherogram and chromatogram in the subject prompted us to investigate structural abnormalities in the δ-globin chain. Analysis of the δ-globin gene performed using PCR and direct DNA sequencing revealed a novel single-nucleotide substitution in exon 2, HBD:c.158A > G (codon 52, GAT > GGT), shown in Figure 1B, resulting in a substitution of aspartate for glycine in the δ-globin chain. This mutation has not yet been described in the δ-globin gene [3], and we decided to name the new variant Hb A2-Mae Phrik after the region of origin of the subject. Further DNA analysis of the αand β-globin genes revealed 3.7 kb deletional α + -thalassemia and also β 26G > A mutation, corresponding to Hb E; therefore, this subject was a triple heterozygote for Hb A 2 -Mae Phrik, Hb E, and deletional α + -thalassemia.
To determine whether the two variants were in cis or in trans, his parents were invited to participate in testing. Their hematological and iron deficiency profiles and globin genotypes are presented in Table 1. The father showed no anemia or alteration of the red blood cell indices, whereas mild anemia and slightly low MCV and MCH were observed in the mother. However, her iron profiles indicated normal iron storage. It was, therefore, less likely that she had iron deficiency but that the true reason for her anemia was not verified, and she may have had chronic disease anemia. Hb and DNA analysis revealed that the patient's father carries a Hb E variant, accompanying α + -thalassemia. The mother's Hb type according to HPLC was A (85.8%), A 2 (2.0%), and A 2 -variant (0.9%) (Figure 2A), while techniques using CE demonstrated a similar Hb pattern (97.9% Hb A, 1.3% Hb A 2 , and 0.8% Hb A 2 -variant) ( Figure 2D), constituting double heterozygosity for Hb E and Hb A 2 -variants.
δ-globin gene sequence analysis for the same condition investigated in S1 was performed in the mother. The analysis detected A and G at the same position in the δ-globin gene; she was therefore heterozygous for Hb A 2 -Mae Phrik. Further DNA analysis of the α-globin genes illustrated no abnormality and thus a possible genotype as αα/αα. The A-G mutation at codon 52 occupied the D helix of the δ-globin chain positioned on the molecular surface in the crystal structure of hemoglobin, which is not involved in contact with heme or between chains [24]. The β-globin gene haplotype analysis demonstrated that the presence of the Hb A 2 -Mae Phrik allele is associated with the β-globin gene haplotype [+ − − − − ± +] ( Table 2).
We developed a new allele-specific PCR technique to confirm the presence of and provide a rapid diagnostic tool for Hb A 2 -Mae Phrik. As a result, a 500 bp amplified fragment specific to the δ 52A > G allele was observed in S1 and his mother, while a 713 bp fragment particular to the δ A allele was observed in all cases (Figure 3). This result confirms that the patients have an A to G mutation of codon 52 of the δ-globin gene and that the newly developed ASPCR technique can be used successfully for the rapid diagnosis of the Hb A 2 -Mae Phrik variant. We developed a new allele-specific PCR technique to confirm the presence of and provide a rapid diagnostic tool for Hb A2-Mae Phrik. As a result, a 500 bp amplified fragment specific to the δ 52A > G allele was observed in S1 and his mother, while a 713 bp fragment particular to the δ A allele was observed in all cases (Figure 3). This result confirms that the patients have an A to G mutation of codon 52 of the δ-globin gene and that the newly developed ASPCR technique can be used successfully for the rapid diagnosis of the Hb A2-Mae Phrik variant.

Discussion
Many δ-Hb variants (Hb A 2 -variants) have been documented worldwide. Patients who are heterozygous or even homozygous for this variant usually have no clinically significant problems, since Hb A 2 constitutes only a small proportion of the total Hb content. However, the interaction of δ-Hb variants with other variants or other thalassemias can lead to an erroneous interpretation of routine laboratory diagnostic test results [4,6,12,25]. This study reported known and novel Hb A 2 -variants coinherited with Hb E in Southeast Asia. Hb A 2 -Melbourne (HBD:c.130G > A), hitherto reported in patients from Thailand and Laos [11,17], was identified in the three Hb E heterozygote subjects (S2-S4). Mild microcytosis was observed in S2 and S4, while a remarkable reduction in MCV (69.2 fL) together with high red cell distribution width (RDW) was observed in S3, indicating masked α-thalassemia 1 gene. On the other hand, iron deficiency parameters and common α-thalassemia defects were not detected in all subjects (Table 1), demonstrating that the red cell alterations may affect Hb E. It is noteworthy that hematological parameters noted in these subjects were not likely to be more severe when compared with heterozygous Hb E, despite coinheritance of this variant [26]. The replacement of negatively charged glutamate with positively charged lysine results in an increased positive charge on the δ-globin chain surface. Thus, the net charge of the α 2 δ 43(G > A) 2 tetramer (Hb A 2 -Melbourne) was more positive compared to that of normal Hb A 2 , such that Hb A 2 -Melbourne was completely separated from Hb A 2 by CE and HPLC methods, with Hb A 2 -Melbourne migrating more slowly than Hb A 2 ( Figure 2). In addition, the high resolution of the new HPLC device could distinguish this variant and allow calculation of its amount, showing a comparable result to the two methods mentioned above ( Figure 2G). This facilitates easy detection of this variant during routine Hb analysis. Remarkedly, the amount of Hb A 2 -Melbourne observed in these three subjects, as measured using two assays (1.3-1.7%), was likely increased when compared to that observed using the same assay in definite cases of heterozygous Hb A 2 -Melbourne (0.9 ± 0.3%) [14]. This result is concordant with that previously noted in a pregnant Laotian woman [17]. The exact Hb A 2 concentration could be calculated by integrating both of the split peaks of HbA 2 , generating a HbA 2 concentration of 2.3% in S1, 2.9% in S2, and 3.1% in S3 for CE, which was the level usually observed in a group of normal individuals (2.5-3.5%) [1] but not in a group with heterozygous Hb E without concomitant iron deficiency or α-thalassemia; the Hb A 2 concentration normally observed in such a group was 3.8-3.9% [26], the accepted range for diagnosing the β-thalassemia trait. Our findings demonstrate that the Hb A 2 levels did not increase in production in compound heterozygous Hb E with the δ-Hb variant, indicating that the low total Hb A 2 level reported in this condition might result from reduced synthesis or instability of the abnormal δ-globin chain [11]. We confirmed the low total Hb A 2 value and that the second Hb A 2 fraction is also visible in the presence of Hb E on Hb-CE analysis; the δ-globin gene mutation is the most likely diagnosis.
Hb A 2 -Mae Phrik (HBD:c.158A > G) is a novel δ-Hb variant characterized in S1, where it has been coinherited with Hb E (HBB:c.79G > A) and a deletional α + -thalassemia, with heterozygosity of this variant in the subject's mother. Molecular analysis showed that S1 inherited the HBD:c.158A > G variant from his mother, while the HBB:c.79G > A variant was inherited from his father. This finding suggests that the βand δ-globin gene mutations were carried in trans by the mother. The expression of hematological parameters of the two conditions was similar (Table 1) and did not significantly differ from those of the compound heterozygous Hb E and Hb A 2 -Melbourne. The replacement of glycine results in loss of one negative charge at the side chain of the δ-chain per molecule, leading to a marked change in the net charge of the tetramer of Hb A 2 -Mae Phrik. Consequently, the mobility of the α 2 δ Mae Phrik 2 tetramer on CE and HPLC is slower than that of Hb A 2 (α 2 δ A 2 ), leading to it being completely separated from this variant and allowing measurement of the concentration of this variant (Figure 2). Electropherograms and chromatograms obtained from CE and HPLC, respectively, showed comparable results to HbA 2 -Mae Phrik. The level of HbA 2 -Mae Phrik, as measured using capillary electrophoresis, was 0.8% and 1.4% in a heterozygote and a compound heterozygote, respectively, while using HPLC revealed 0.9% and 1.4%, respectively. This demonstrated that HbA 2 -Mae Phrik levels were raised when combined with Hb E. In addition, we observed equal proportions of Hb A 2 (1.3%) and Hb A 2 -Mae Phrik (0.8%), as measured using capillary electrophoresis, in the mother of S1, who was heterozygous for the variant. In addition, the variant combined with Hb E (S1) and Hb A 2 and Hb A 2 -Mae Phrik was present in consistently equal proportions of 1.1% and 1.4%, respectively, indicating similar assembly rates of the two δ-globin chains to an α-globin chain. Of interest, the Hb A 2 (α 2 δ A 2 ) level was observed to be 1.3% and 1.1%, although the latter was a value observed in S1 who had coinherited the variant with Hb E, which was likely to lead to a lower value when compared to the range normally observed in heterozygous Hb E without an αor β-thalassemia allele [19]. However, in order to calculate the actual Hb A 2 value in these cases, the combined percentage of Hb A 2 and the Hb A 2 -Mae Phrik variant must be assessed. Therefore, total Hb A 2 as measured by CE was 2.5% (1.1 + 1.4) in S1 and 2.1% (1.3 + 0.8) in his mother, who was heterozygous for this variant. This showed a consistently lower value than those usually observed for heterozygous Hb E (3.8 ± 0.3%) [19]. This result demonstrates a distinct decrease in total Hb A 2 when combined with other Hb variants, which might result from the reduced synthesis of the abnormal δ-globin chain. Nevertheless, an aspartate amino acid at codon 52 of the δ-globin chain is in the second of the D helices of the Hb molecule and is not involved in contact with heme or between chains. Therefore, the replacement of glycine at this position could compromise the instability of the Hb molecule.
The characteristic HPLC retention times seen in Hb A 2 -Mae Phrik ranged from 4.44 to 4.46 min ( Figure 2D,E), compared to the 4.73-4.75 min observed for HbA 2 -Melbourne. It seems that Hb A 2 -Mae Phrik eluted from the cation exchange column markedly faster than HbA 2 -Melbourne. This may be explained by the fact that glutamate was substituted for lysine in HbA 2 -Melbourne, providing increased affinity with the column than when aspartate was replaced with glycine in Hb A 2 -Mae Phrik. Therefore, we suggest that the characteristic HPLC retention times obtained for Hb A 2 -Melbourne and the novel variant Hb A 2 -Mae Phrik could be used to provide a presumptive identification.
Recently, another Hb variant caused by a mutation at the same position was identified, the non-pathological Hb A 2 -Walsgrave (δ52(D3) Asp > His; HBD:c.157G > C). This variant was found in the United Kingdom in a person who had emigrated from India [27] and had also been characterized in a pregnant Thai woman [15]. Moreover, five other δ-Hb variants have been documented in Thailand [12][13][14]16]. These variants presented on an electropherogram in zone 1, similarly to Hb A 2 -Mae Phrik, potentially causing issues in the routine diagnosis of this variant. Thus, a differential diagnosis of these conditions requires DNA analysis in order to provide an accurate diagnosis and appropriate genetic counseling. The newly developed allele-specific PCR method can readily detect the Hb A 2 -Mae Phrik mutation in δ-globin genes. Moreover, the second small Hb peaks appear in the electropherogram, not only indicating Hb A 2 -variants but also possibly unstable or degraded Hb. Identifying the Hb A 2 -variant is crucial for diagnosis and the exclusion of β-thalassemia or α-thalassemia, especially in areas where thalassemias are prevalent.
Haplotype information permits a better understanding of the origin and evolutionary relationships between populations worldwide. β-haplotypic analysis demonstrated that the Hb A 2 -Mae Phrik allele was greatly associated with haplotype [+ − − − − ± +] ( Table 2). These findings indicate that the β-haplotype is very likely similar to that observed in the ancestral Thai 5' haplotype [23]. This may imply that the herein-described mutation in the δ-globin gene has the same evolutionary origin. Moreover, the β-haplotype associated with Hb A 2 -Mae Phrik is likely that observed in Hb A 2 -Melbourne [11,14]; the Hb A 2 -Mae Phrik allele certainly has the same evolutionary origin.

Conclusions
This is the first report of the novel Hb A 2 -Mae Phrik variant, with two different genotypes in Thai people. This variant is clinically silent, and coinheritance of α-thalassemia does not change its phenotype. Hb A 2 -Melbourne combined with Hb E is also described. Interaction of Hb E with the two δ-globin variants leads to a decrease in total Hb A 2 level and therefore could lead to misdiagnosis when inherited together with a β-thalassemia allele. A second small Hb fraction from these or other δ-globin mutations seen in the same zone of the electropherogram might lead to problematic interpretation during routine Hb analysis. The HPLC retention times of the δ chain variants are characteristic and useful in order to initially distinguish these variants. A newly developed ASPCR assay is able to readily detect and distinguish this variant from others, facilitating accurate diagnosis, appropriate treatment, and genetic counseling.
Author Contributions: A.P. performed Hb analysis, laboratory data analysis and interpretation, and drafting of the manuscript; M.M. assessed the iron deficiency profile; N.J. performed specimen collection and clinical data acquisition; S.S. performed the DNA extraction and laboratory investigation; S.P. is the main investigator, responsible for conception and design of the work, and also performed laboratory investigation, wrote the manuscript, critically revised and approved of the article, and acquired the research grant. All authors have read and agreed to the published version of the manuscript.