Single-Tube Multiplex-PCR Screen for Anti-3.7 and Anti-4.2 α-Globin Gene Triplications

The coexistence of α-globin gene triplication (ααα) is an important modulator of the severity of β-thalassemia trait or β-thalassemia intermedia, exacerbating the phenotypic severity of β-thalassemia by causing more globin chain imbalance (1)(2). Typically, the inheritance of a single β-thalassemia allele is associated with mild anemia and hypochromic microcytic red cells. Compared with simple β-heterozygotes, co-inheritance of triplicated or quadruplicated α-globin genes in β-heterozygotes often leads to more significant anemia, splenomegaly, more pronounced red cell abnormalities, the presence of circulating normoblasts, higher hemoglobin F concentrations, and even the presence of inclusion bodies in erythroblasts (3)(4). Because the α- and β-globin gene clusters are physically unlinked and segregate independently, β-thalassemia carriers who also carry triplicated or quadruplicated α-globin genes have a 25% risk of having a similarly affected offspring, although their partners may be entirely normal. Triplicated α-globin genes appear to be ubiquitous and have been found in most populations (2). They result from misalignment and unequal crossover between the homologous X-, Y-, and Z-box segments of the α-globin gene cluster during meiosis (Fig. 1A ). Generally, two types of triplicated alleles can be generated from an unequal crossover, αααanti3.7 and αααanti4.2. If the crossover occurs between the homologous Z2 and Z1 boxes, also referred to as a “rightward crossover”, this produces a −α3.7 single-gene deletion allele and the reciprocal αααanti3.7 triplicated allele. However, if the crossover occurs between the X2 and X1 boxes (a “leftward crossover”), a −α4.2 single-gene deletion allele and the reciprocal αααanti4.2 triplicated allele are generated (5). Formation and molecular detection of α-globin gene triplications. (A), misalignment and unequal crossover at the α-globin gene cluster generates a single gene deletion and reciprocal gene triplication. Crossovers between misaligned Z boxes give rise to −α3.7 and αααanti3.7 chromosomes (rightward crossover). Crossovers between misaligned X boxes give rise to −α4.2 and αααanti4.2 chromosomes (leftward crossover). (B), determination of α-globin genotype by a combination of seven-deletion multiplex-PCR (top gel) and anti-3.7/4.2 multiplex-PCR (bottom gel). Lane M, 1-kb DNA ladder (Fermentas); lane AL, allelic ladder. (C), expected hybridizing bands of various unequal crossover derivative alleles (left) and genotypes (middle), and actual autoradiogram of three DNA samples (right), after Southern hybridization of BamHI and BglII digests with [32P]dATP-labeled α-globin gene probe. A Sri Lankan study of individuals with severe to moderate β-thalassemia revealed a 2% frequency of α-globin gene triplications (6), whereas a preliminary study in Hong Kong suggests that the frequency of α-globin gene triplication carriers among individuals with β-thalassemia and iron deficiency is ∼3.3% (E.S.K. Ma, A.Y.Y. Chan, L.C. Chan, unpublished observation). The true prevalence of α-globin gene triplications is not well defined in other populations. Because the α- and β-globin gene clusters are unlinked, it is likely that the percentages derived from the study of β-thalassemia cohorts are a reasonable indicator of the general population frequency of α-globin gene triplication carriers. Current PCR methods can detect the anti-3.7 but not the anti-4.2 triplication (7)(8), and Southern blotting is required to detect both the anti-3.7 and anti-4.2 triplication and quadruplication alleles (9). Although highly sensitive and specific, Southern blotting is labor-intensive and time-consuming, and is thus not an ideal screening tool. We have developed and validated a simple, rapid, and reliable single-tube multiplex-PCR assay to screen for the presence of the anti-3.7 and anti-4.2 α-globin gene triplications. Genomic DNA samples carrying either an αααanti3.7 or αααanti4.2 triplication allele, as determined by Southern blot analysis, were used in assay optimization. Final validation of the optimized assay was accomplished by blinded analysis of 31 samples that were either positive (n = 21) or negative (n = 10) for triplicated α-globin genes. Oligonucleotide primers were designed to specifically amplify the unequal crossover region(s) within the anti-3.7 or anti-4.2 chromosomes. Primers AT3.7-F and AT3.7-R were designed to anneal to the unique stretches between the Y1 and Z1 homology boxes and between the Z2 and X1 boxes, respectively, to amplify an ∼1.9-kb fragment only from the hybrid Z1Z2 box of the anti-3.7 allele (Fig. 1A ). Similarly, primers AT4.2-F and AT4.2-R anneal to the unique regions between the Z2 and X1 boxes and between the X2 and Y2 boxes, respectively, to amplify an ∼1.7-kb fragment only from the hybrid X1X2 box of the anti-4.2 allele. Additionally, amplification of a large fragment (∼2.5 kb) of the 3′ untranslated region of the LIS1 gene was included as a control for amplification success, using primers LIS1-2.5F and LIS1-2.5R. Each 50 μL of anti-3.7/4.2 multiplex-PCR reaction contained 200 μM each of the deoxynucleotide triphosphates, 1.5 mM MgCl2, 1× Q-solution (Qiagen), 2 U of HotStarTaq DNA polymerase in supplied reaction buffer (Qiagen), 100 ng of genomic DNA, and the six primers at different concentrations (Table 1 ). Thermal cycling was performed in a T3 instrument (Biometra) under conditions identical to those for the seven-deletion multiplex-PCR assay (10). We analyzed 10 μL of each anti-3.7/4.2 multiplex-PCR product by electrophoresis through a 1% agarose gel in 1× Tris-borate-EDTA buffer at 15 V/cm for 1 h. The seven-deletion multiplex-PCR assay was performed and analyzed exactly as described previously and was used to detect the seven common α-thalassemia deletions (10). Anti-3.7/4.2 multiplex-PCR primers and expected amplicon sizes (top), and results of blinded analysis of 31 samples (bottom). UTR, untranslated region. −−THAI (1.2 kb), −(α)20.5 (1.0 kb), −−MED (0.8 kb), and −−FIL (0.5 kb) amplicons were also screened for but were negative in these samples. Anti-3.7/4.2 multiplex-PCR primers and expected amplicon sizes (top), and results of blinded analysis of 31 samples (bottom). UTR, untranslated region. −−THAI (1.2 kb), −(α)20.5 (1.0 kb), −−MED (0.8 kb), and −−FIL (0.5 kb) amplicons were also screened for but were negative in these samples. To validate the anti-3.7/4.2 multiplex-PCR assay, we performed a double-blind analysis of 21 DNA samples harboring either an anti-3.7 or an anti-4.2 α-globin gene triplication, together with 10 negative control samples (αα/αα). The α-globin genotypes of these samples had been determined previously by Southern blot analysis. The anti-3.7/4.2 multiplex-PCR was assessed in combination with the α-thalassemia seven-deletion multiplex-PCR assay (10) to determine their α-globin genotype. Therefore, for each DNA sample tested, two amplification reactions were performed. The seven-deletion multiplex-PCR assay screens for the seven common α-thalassemia deletions. In this assay, a control α2-globin gene fragment is amplified to indicate heterozygosity for a nondeleted allele when a deletion is also present (Fig. 1B ); it cannot, however, distinguish between a wild-type nondeleted allele (αα) and a triplicated (ααα) allele. We used the anti-3.7/4.2 multiplex-PCR assay to detect the presence of the triplicated allele in these samples (Fig. 1B ). In the blinded analysis, the anti-3.7/4.2 multiplex-PCR assay detected the presence of the correct triplicated allele (αααanti3.7 or αααanti4.2) in all 21 triplication-positive DNA samples (Table 1 ). No anti-3.7 or anti-4.2 junction fragment was detected in any of the 10 negative control DNA samples. In combination with the seven-deletion multiplex-PCR assay, the α-globin genotype of all 31 DNA samples was correctly determined (Table 1 ). The anti-3.7/4.2 multiplex-PCR assay thus serves as a useful rapid screen for the presence of the anti-3.7 and/or the anti-4.2 types of triplication. For the majority of individuals negative for a triplication by this rapid PCR assay (∼96.5–98.0% based on data extrapolation), no further analysis is required. Southern blot analysis is required only in the presumptive 2.0–3.5% of samples that are PCR-positive, to distinguish between the heterozygous and homozygous states (Fig. 1C ). It is likely that the result of our anti-3.7/4.2 multiplex-PCR assay will also be abnormal with the rarer ααααanti3.7 and ααααanti4.2 quadruplicated alleles, which have two copies of the crossover junction, one copy more than their triplicated counterparts. To distinguish between the triplicated and quadruplicated alleles, however, will require Southern blotting because both triplicated and quadruplicated anti-3.7 and anti-4.2 alleles should yield essentially the same ∼1.9-kb and ∼1.7-kb junction fragments, respectively, with this assay. Without an initial anti-3.7/4.2 multiplex-PCR screen, use of Southern analysis alone to detect triplications and quadruplications would require a minimum of two blots, each containing DNA digested by a different restriction enzyme, because no single enzyme digestion can provide full genotype information (9). For example, digestion with BamHI followed by hybridization with an α-globin gene probe enables detection of and differentiation between a triplication and a quadruplication, as well as determination of zygosity, but it cannot distinguish between the anti-3.7 and anti-4.2 types (Fig. 1C ). Conversely, a BglII-digested DNA blot hybridized with the same probe enables the type of triplication or quadruplication to be determined (anti-3.7 or anti-4.2), but it does not provide information on zygosity. An initial anti-3.7/4.2 multiplex-PCR screen enables determination of the type of triplication (and potentially quadruplication) present in the patient (anti-3.7 or anti-4.2), thus eliminating the need for a BglII-digested Southern blot. Therefore, only a BamHI-digested Southern blot hybridized with an α-globin probe is required to detect the presence of the rare quadruplicated and homozygous genotypes. We thank A.S.C. Tan and G.H. Yeo for technical assistance. This work was supported by Grants NMRC/0365/1999 and NMRC/0732/2003 from the National Medical Research Council, Singapore (to S.S.C.).