Distinct Distribution of HBB Variants in Two Cohorts of Beta Thalassemia Patients, and a Novel Variant from Turkey

Thalassemias are among the most frequently observed genetic diseases in the world [1]. The most common types are alpha and beta thalassemias [1, 2]. Beta thalassemia is an autosomal recessive single-gene disease caused by pathogenic or likely pathogenic variants in the β-globin (HBB) gene. It is characterized by the absence or reduction of the β-globin chain in the hemoglobin tetramer. The clinical manifestations of beta thalassemia are highly variable. It is classified into three types: thalassemia major, intermedia, and minor. Beta thalassemia major is characterized by early onset severe anemia requiring regular blood transfusions; intermedia and minor, usually moderately anemic or healthy [1].

The HBB gene is located on the short arm of the 11th chromosome (11p15.4) and contains three exons and two introns encoding 146 amino acids. In addition to beta thalassemia, the HBB gene pathogenic variants also cause other clinically significant hemoglobinopathies such as sickle cell anemia, HbC disease, HbE disease [3, 4]. Most beta thalassemia pathogenic variants are point type such as single nucleotide substitutions, deletions, and insertions of one or two nucleotides. Sequence variants are interpreted according to American College of Medical Genetics and Genomics (ACMG) standards and guidelines. The variants are classified into five tiers depending on the applied criteria: pathogenic, likely pathogenic, uncertain significance, likely benign, and benign [5]. To date, more than 400 HBB gene pathogenic variants have been identified [1, 2, 6]. Classification changes that may occur in variants of unknown or conflict clinical significance are of great importance.

Beta thalassemia is mostly seen in the Mediterranean region, the Middle East, Africa, East Asia, and Latin America [7, 8]. It has also become a global health problem due to population migrations. Migrations also influence the distribution, diversity, and frequency of genetic variants [9]. Consequently, population studies based on genetics may provide insights into historical social events such as migration patterns and wars over time [10]. Nevertheless, the existing literature on this subject remains inadequate.

Beta thalassemia is common in Turkey, due to its geographical location and the prevalence of consanguineous marriages. The overall incidence of beta-thalassemia in Turkey is 2.1%, but this rate varies regionally between 0.6% and 13% [2, 4, 6, 11]. There are about 1,600,000 thalassemia carriers and about 5,500 patients with homozygous thalassemia and other hemoglobinopathies in Turkey, while the population is about 85 million [4, 12]. Syria is also one of the regions where thalassemia is common. There are more than 8,000 registered thalassemia major patients in Syria, and the number of patients is constantly increasing every year [13]. Turkey has recently accepted approximately 3.5 million refugees and immigrants due to the civil war in Syria. Konya is the city located in Central Anatolia. Approximately 118 thousand (4.9% of the city population) Syrians are sheltered in also Konya. So, Syrian beta thalassemia patients followed in our center have increased recently.

Since the knowledge of the variant spectrum present in a population is fundamental for the molecular diagnosis and prevention of beta thalassemia and other hemoglobinopathies, determination of updated variant profiles in the populations is important. We aim to update the variant spectrum that has changed due to migrations in the HBB gene in our region. Our study has contributed to the literature and population genetic science on this subject. Also, we present the c.31_32insT variant that emerged for the first time in Turkey as a result of the migration.

Our study group consisted of 142 unrelated patients (46 Turkish and 96 Syrian) with 90 thalassemia minor and 52 thalassemia major prediagnoses. HBB sequence analysis of these patients was performed between May 2018 and December 2020 at Konya Training and Research Hospital, and between December 2020 and February 2022 at Konya City Hospital, with pathogenic and/or likely pathogenic variants detected in one or both alleles. For the detection of HBB gene variants; DNA was isolated from peripheral blood samples taken into 5 mL EDTA tube. PCR was established from the obtained DNA material with specific primers (NM_000518.5). The HBB gene was evaluated by applying sanger sequence analysis to the entire gene PCR product in the ABI 3500 DNA Sequencer device.

Our study was approved by Karatay University Ethics Committee with the date of March 29, 2022, and decision number 30576. Written informed consent was obtained from all patients for genetic diagnosis and retrospective study. Written informed consent was obtained from foreign patients accompanied by a sworn translator.

Statistical analysis was carried out via SPSS 17 for Windows (Chicago Inc., 2008). Categorical variables are demonstrated as frequencies and percentages. Analyses were conducted with Pearson χ² and Fisher exact tests. Values of p < 0.05 were accepted as significant.

The two cohorts consisted of 142 patients with hemoglobinopathy. The characteristics of the two cohorts such as age, gender, consanguinity, family history, and comorbidities are presented in Table 1. There was no significant difference between the two cohorts in terms of mean age and gender distribution. However, consanguinity, family history, and comorbidity were statistically significantly associated with ethnicity. Comorbidity was observed in 8 (66%) of the patients with two different variants, and in 34 (81%) of the patients with homozygous variant. The majority of comorbidities were hepatomegaly and/or splenomegaly and splenectomy. Heart failure, changes in facial bones, and short stature were also observed in a few patients. One of the patients with homozygous variant was discovered incidentally when investigated for the etiology of anemia in adulthood. MLPA of the HbA gene was also performed to this patient, and HbH disease was detected. The patient had a mild clinical course due to the reduction of the imbalance between the alpha and beta chains.

In this study, thalassemia minor was detected in 81, thalassemia major in 46, thalassemia intermedia in 7, and other hemoglobinopathies in 8 of 142 patients. 23 types of pathogenic or likely pathogenic variants were identified in the HBB gene. These variants involved 20 different beta-thalassemia variants and three other hemoglobinopathies (Table 2). We also identified c.−31C>T variant, commonly referred to as 5′UTR +20 (C>T), in 1 patient, but we did not include it in our study because ClinVar has recently classified this variant as likely benign.

The list, number, and distribution pattern of the variants are given in Tables 2 and 3. Three variants including IVS II-1G>A, IVS I-110G>A, Codon 39 (C>T) constituted 52.3% of all detected variants. Table 2 also includes the dbSNP number of the variants and their localization information on the gene. Majority of the variants we detected was localized in intron1 (38.6%).

The rate of c.−80T>A, in other name is −30 T>A, was found significantly higher in Turkish subjects compared to the Syrian patients (5.4% vs. 0%, respectively; p = 0.022, Fisher’s exact test). Similar to this, c.17_18delCT, in other terms Codon 5 (−CT), frequency was significantly higher in Turkish than that of Syrian subjects (12.5% vs. 3.5%; p = 0.028, Fisher’s exact test). In addition to these findings, c.20A>T, c.25_26delAA and c.93–21G>A variants were significantly higher in favor of Turkish patients (p = 0.024, p = 0.022 and p < 0.001, respectively). On the other hand, c.92+5G>C, c.118C>T and c.315+1G>A variants were found at a higher rates in Syrian subjects compared to the Turkish patients (p = 0.036, p = 0.002 and p = 0.040, respectively, see Table 3).

The distribution by zygosity and ethnicity of variants is also shown in Table 4. The term “compound heterozygosity” refers to beta-thalassemia compatible with autosomal recessive inheritance. This was confirmed by segregation analysis in some patients with two heterozygous variants, whereas in others it was interpreted on the basis of other laboratory and clinical information and family history. Syrian patients constituted the majority of homozygous and compound heterozygous patients; 80.9% and 83.3%, respectively. There was a significant difference in the distribution of homozygous mutations between the two groups (8 vs. 34; χ²(1) = 4.403, p = 0.028). No significance was detected in compound heterozygotes (2 vs. 10; p = 0.339, Fisher’s exact test). Another significance was detected in heterozygous distribution between two groups (36 vs. 52; χ²(1) = 7.661, p = 0.006). While 21.7% of Turkish patients (n = 10) were homozygous or compound heterozygous, this rate was 45.8% in Syrian patients (n = 44) and this distribution was significantly higher in favor of Syrian subjects (χ²(1) = 7.661, p = 0.006).

Homozygous and compound heterozygous genotypes are shown in Tables 5 and 6. We had 42 patients with homozygous genotypes, and 12 with compound heterozygous genotypes. Although consanguineous marriages are very common in both Turkey and Syria, homozygous and combined heterozygous genotypes were more common in Syrian patients in our study.

One of the patients had both homozygosity for Codon 39 (C>T) and heterozygosity for c.295G>A. It was regarded as homozygous in the Table 4, but all variants were considered in variant counting (Table 3). Therefore, we detected 197 variants, despite the number of patients is 142. While the most common variant in total and in Syrian patients was IVSII-1G>A, it was IVS I-110G>A in Turkish subjects.

The most frequent HBB variant in Turkey is IVS I-110G>A (40%), as in Syria and many other countries [4, 6, 11, 13‒16]. The HBB variant spectrum may show great molecular heterogeneity between different regions, even within the same country, like in Syria [13]. The most common pathogenic variant in every region of Turkey has been reported as IVS I-110G>A. The frequency of the IVS I-110G>A has been shown to vary from 20 to 30% in the Eastern and Southeastern Anatolia to 50% in the west, in the Marmara and the Aegean regions [6, 15, 17]. Tadmouri G.O. reported that the IVS I-110G>A variant was the most common in the Central Anatolia Region with 52.3% [18]. In a thesis study conducted in 2009, it was shown that the most common variant in Konya was also IVS I-110G>A with a 74% frequency [19]. In our study, IVS I-110G>A is the second common variant in total, while it is the most common variant in Turkish patients (Table 3).

IVS II-1G>A is the variant with the highest prevalence in this study. However, this variant has not been reported as the most commonly neither in Syria nor in Turkey [4, 12, 14, 16, 19, 20]. There have been studies reporting this variant with the third frequency [15, 19]. Homozygous and heterozygous individuals were evaluated together in these two studies, as in ours, unlike studies conducted only on patients with beta thalassemia major. It is understood from this comparison; the IVS II-1G>A variant may be more common in heterozygous individuals than homozygous ones. Indeed, this variant was also found in our study with the third frequency in Turkish patients. Besides, all of these patients had heterozygous genotype. 86.11% (31/36) of this variant was seen in Syrian patients. All of the homozygous and compound heterozygous genotypes containing IVS II-1G>A were seen in Syrian patients (Tables 5; Table 6).

Codon 39 (C>T) has been found the third most common variant in this study, and the second common variant between Syrian patients (Table 3). In fact, 93.75% of whole alleles we detected Codon 39 (C>T) variant are from Syria (Table 3). Furthermore, this variant is the most common among homozygous genotypes (Table 5). Codon 39 (C>T) orders third according to the Syrian variant distribution [13, 15, 20]. It was observed a clear deviation in the distribution of this variant between Syrian regions. It was seen the most frequent in northeast region (Raqqa) with 34.8% rate [13].

The IVS I-1G>A, which we detected with the fourth frequency in our study, is known as the second most common variant in Syria [13, 15, 20]. Like us, Karaer et al. [6] also found this variant in the fourth frequency. There have been studies reporting that it is the second and third in Turkey [12, 21]. Besides, this variant is in the third frequency in our Syrian patients. 86.7% of alleles include this variant has been found in Syrian patients (see Table 3).

We have detected Codon 5 (–CT) variant with the second frequency in our Turkish patients (Table 3). In various studies conducted in different regions of Turkey, the frequency of this variant has been reported between 2.6 and 4% [2, 6, 7, 19, 22, 23]. As a matter of fact, this rate was reported to be 4% in a study conducted in Konya and its surroundings in 2009 [19]. We have detected this variant by 4.5% in carriers, and 6.1% in total (Table 3). Three of these patients are Syrian, two of them homozygous (Tables 3; Table 5). Murad et al. [13] reported Codon 5 (–CT) with a frequency of 5.6% in Syria. It was mostly reported in the middle of Syria. Gunes et al. [14] found this variant significantly higher in Syrian refugees compared to Turkish citizens (15.7% vs. 0%, p = 0.023). Similar to this findings, in our study Codon 5 (−CT) frequency was significantly higher in Turkish than that of Syrian subjects (12.5% vs. 3.5%; p = 0.028).

IVS I-5G>C has been observed only in Syrian patients in our study (see Table 3). Of 81.8% alleles carrying IVS I-5G>C is situated in homozygous or compound heterozygous individuals (see Tables 5; Table 6). Gunes et al. [14] mentioned this variant with the third frequency among Syrian refugees. They also reported IVS I-5G>C and IVS I-1G>A as the most common variants among compound heterozygous genotypes. Murad et al. [13] submitted this variant with 4.1% frequency between the eight predominant variations in Syria. This variant was reported at very low rates in studies in Turkey [15]. In the study by Karaer et al. [6], this rate was 3.87% and the majority of these patients were also Syrian patients.

In addition to IVS I-5G>C; there were nine more variants seen only in Syrian patients in this study (Table 3). Hb C, c.52A>T, Codon 29 (C>T), IVS I-128T>G, and Codons 36/37 (−T) are very rare variants. Hb C has not been described in Syria or Syrian refugees so far [13, 14]. Codon 17 (A>T) was reported in 4 Syrian immigrants [6]. Codon 29 (C>T) and Codons 36/37 (–T) have been reported in very few patients in Syria [6, 13]. IVS I-130 G>C has been seen in Syria at a rate of 1.2% [13]. IVS I-130 G>C and Codons 36/37 (–T) have also been reported at low rates in Turkey [6, 11, 15, 24]. Altay [15] reported that IVS II-745C>G was the second frequent variant like IVS I-1G>A and mostly in thalassemia major patients, in Syria. This variant has been reported in middle and southern Syria [13]. In our study, it was also found in 3 Syrian patients, two of whom were compound heterozygous (see Tables 3, 6). In 2013, it was reported that the c.−31C>T variant should be considered as an innocuous single nucleotide polymorphism associated with the IVS II-745C>G variant in cis position [25]. As a matter of fact, ClinVar has classified the c.−31C>T variant as likely benign based on clinical-significance assessments of six clinical diagnostic laboratories and a database (ITHANET) in recent years.

The c.30_31insT variant was detected in 1 patient in a study in the Thrace region [2]. This variant was previously reported as pathogenic with rs34548294 dbSNP number (HGMD ID: CI951933). The c.31_32insT variant has not previously been described. This variant causes a frame-shift and greatly affects the normal structure and function of the protein. In silico genetic prediction tools evaluate this variant to be likely pathogenic according to the recommendations of the ACMG. Indeed the c.31_32insT variant has caused a similar change and frameshift (p.Ala11ValfsTer13) with the c.30_31insT variant (p.Ala11CysfsTer13). The 30_31insT variant was reported at a rate of 0.3% in middle and northeast of Syria [13]. In our study, 3 patients with c.31_32insT variant were also Syrian (see Table 3). Yalçıntepe S. declarated 3 patients with the c.364G>A (p.Glu122Lys) variant [2]. We also detected c.364G>C (p.Glu122Gln) genomic change at the same position in 2 patients, one of whom was from Syria. It has not been found in the study on the general population of Syria [13]. Glutamic acid (Glu)-Glutamine (Gln) missense variation has an intermediate score in terms of Grantham matrix score. Although Glu and Gln have similar chemical properties, its specific effects will depend on the structure and function of the protein [26] In silico prediction tools assess the effect of Glu-Gln variation on the beta-globin protein as likely pathogenic.

We also had variants present in only Turkish patients (see Table 3). −87 C>G was seen as homozygous in 1 patient in our study. It is a very rare variant. It was found a few cases previously [19]. C>T variation at the same position was also reported low rate [11, 15]. Altay Ç. reported that the −30 T>A and Codon 8 (−AA) variants were most common in Eastern Anatolia, and they probably originated from here and moved to Southeast and Central Anatolia, and then to the Mediterranean and Aegean Regions [15]. In our study, these two variants were found at a rate of 1.5%. Karaer et al. [6] reported these variants in Syrian and Turkish patients in southeastern Turkey. −30 T>A and Codon 8 (−AA) variants have been reported in Syria at a rate of 2.3% and 6%, respectively. The common region with no −30 T>A variant and the lowest number of Codon 8 (−AA) variants is the northeast [13, 20]. There are many studies reporting the frequency of c.25_26 delAA variant between 2% and 12% [2, 6, 7, 11, 15, 19, 23, 24, 27‒29]. Moreover, in researches conducted throughout Turkey, there are studies reporting this variant even in the 2nd and 3rd frequencies [4, 22]. IVS I-116T>G and Codon 98 (G>A) variants were observed at a frequency of less than 1%.

More than 40 types of variant have been found in Turkey [1, 15, 30]. The number and types of variants detected in each study are affected by the methods used for detection. In this study, we found 23 types of variant (see Table 2). Since some previous studies used strip-based method, known variants could be screened [7, 11, 19, 23]. We used Sanger sequencing method to identify possible new sequence variations in the HBB gene. Indeed, we were able to detect different genomic changes at the same position as some previously identified variants, such as c.31_32insT, c.364G>C. The diversity of variants we found was also higher than in previous studies [7, 11, 14, 23].

The mutation spectrum and allele frequencies are different for the Turkish and Syrian populations. Therefore, the mutation spectrum and allele frequencies observed in the combined population (Turkish and Syrian) are expected to be different from the Turkish or Syrian populations. However, although the most common variant was reported as IVS I-110G>A in both populations, IVS II-1G>A was found in this study. Likewise, the second common variant was also different from the data of the two populations. IVS I-1G>A in Syria and mostly IVS I-6T>C in Turkey, whereas IVS I-110G>A was found in our study (Table 3). The third common variant was consistent with the Syrian data. These results may be explained by population-specific founder effect. Some researchers have suggested that migration is the cause of gene flow and founder effect [10]. Indeed, we think that the increased migration to Turkey due to the civil war in Syria caused these results.

This study has revealed significant differences in the distribution pattern of HBB variants among Turkish and Syrian patients. We have discovered a novel variant in three Syrian immigrants and reclassified another variant. Furthermore, we indicated that the variant spectrum in the Turkish-born population is not fully known, and immigrants may carry variants that are not previously known.

It is important to know the updated variant profiles and classifications for planning thalassemia molecular diagnosis studies, organizing thalassemia prevention programs effectively and for appropriate genetic counseling. Although this study makes a regional contribution to the literature on these subjects, more comprehensive studies are needed to become a resource for population genetics science.

The authors thank Dr. Muserref Basdemirci for her contribution in data collection.

This study was approved by Karatay University Ethics Committee with the date of March 29, 2022, and decision number 30576. Written informed consent was obtained from all patients for genetic diagnosis and retrospective study. Written informed consent was obtained from foreign patients accompanied by a sworn translator.

The authors have no conflict of interest to declare.

This study was not supported by any sponsor or funder.

Hatice Kocak Eker contributed to design of the work, to collection of laboratory data of the patients, and prepared the tables. Hatice Kocak Eker and Ozgur Balasar carried out the analysis, discussed the results, and wrote the manuscript.

The data that support the findings of this study are available from the corresponding author upon reasonable request.