Antonio Cao, * Renzo Galanello, M. Cristina Rosatelli
Clinica e Biologia dell'Et� Evolutiva, Universita Studi Cagliari, Cagliari, Sardinia, Italy
Address: *Ist. Clinica e Biologia dell'Eta Evolutiva, Universit� Studi Cagliari,
Via Jenner s/n, 09121 Cagliari, Sardinia, Italy
Tel. 0039-070-503341
Fax 0039-070-503696
Email: acao@mcweb.unica.it

Download as a PDF here

SUMMARY

The most important aspects of carrier detection procedures, genetic counselling, population screening and fetal diagnosis of the thalassaemias and sickle cell anaemia are reviewed. Carrier detection can be made retrospectively, i.e. following the birth of an affected child, or prospectively. Most carrier detection and genetic counselling in populations at risk for a-thalassaemia and sickle cell anaemia is retrospective. However, some prospective carrier screening programmes for sickle cell anaemia are ongoing in Cuba and Guadeloupe and very limited screening for a-thalassaemia is in progress in some South East Asian populations. As regards b-thalassaemia, several programmes, based on carrier screening and counselling of couples at marriage, preconception, or early pregnancy, have been operating with several populations at risk in the Mediterranean. These programmes have been very effective, as is proved by the fact that he target population has improved its knowledge of thalassaemia and its prevention, and by the marked decline that has been observed in the incidence of thalassaemia major. Carrier detection is carried out by haematological methods, followed by mutation detection by DNA analysis. Prenatal diagnosis is accomplished by mutation analysis on PCR-amplified DNA from chorionic villi. Future prospects include automation of the process of mutation detection, simplification of preconception and preimplantation diagnosis, and fetal diagnosis by analysis of fetal cells in the maternal circulation.

Key words: b-thalassaemia; a-thalassaemia, sickle cell anaemia; carrier detection, population screening, prenatal diagnosis

INTRODUCTION

The inherited haemoglobinopathies are large groups of autosomal recessive disorders that include the thalassaemias and sickle cell anaemia. The thalassaemias are caused by the defective (+) or absent (0) production of one of the globin chains of the haemoglobin tetramer. According to the type of globin chain involved, we distinguish a-, b- and d-thalassaemias (Weatherall and Clegg, 1981). The sickle disorders most commonly result from homozygosity for the S mutation, an A�T substitution at codon 6 of the b-globin gene leading to the replacement of valine for glutamic acid.

The incidence of b-thalassaemias is high in the Mediterranean area, the Middle East, the Indian Subcontinent, and the Far East. The frequency of the a-thalassaemias is particularly high in the Far East, but the condition is not rare in the Mediterranean area, the Middle East, and the Indian Subcontinent. Sickle cell anaemia shows the highest incidence in tropical Africa. However, due to the population flow and the slave trade, these diseases have spread widely nowadays and occur also with relatively high frequency in northern Europe, North and South America, and in the Caribbean. The best available approximate estimate indicates that about 250 million people, 4.5% of the world population, are carriers of a defective globin gene. Each year about 300,000 affected homozygotes are born, approximately equally divided between sickle cell disorders and thalassaemia syndromes (WHO Scientific Group, 1996).

At present a limited proportion of affected homozygotes may be cured definitively by bone marrow transplantation from human leukocyte antigen (HLA) identical siblings. However, the majority of patients can only count on supportive management at present.

We herein review carrier detection and fetal diagnosis of inherited haemoglobinopathies, which are fundamental issues in the clinical management of these disorders nowadays.

CARRIER DETECTION

Heterozygous b-thalassaemia, both the � and the b+ type, is characterised by a high red blood cell count, microcytosis, hypochromia, increased haemoglobin A2 (HbA2) levels, and an unbalanced a-globin non-a-globin chain synthesis. However, several environmental or genetic factors may modify this haematological phenotype, causing carrier identification difficulties (Table 1).

Although iron deficiency may decrease the typical high HbA₂ levels of heterozygous b-thalassaemia, in our experience they remain within the b-thalassaemia carrier range except in the case of severe anaemia (Galanello et al, 1981). At any rate, iron studies may serve to rule out associated iron deficiency.

In many carrier detection procedures, the preliminary selection of individuals at risk of being heterozygous for a form of thalassaemia is based on the determination of mean corpuscular volume (MCV) and mean corpuscular haemoglobin (MCH) values. It is worth noting, however, that since double heterozygotes for b-thalassaemia and a-thalassaemia may have normal MCV and MCH values, they could be missed by this approach (Melis et al, 1983; Rosatelli et al, 1984). In carriers of b�-thalassaemia, the a-thalassaemia determinants that can give this effect are deletion of two of the four a₂-globin structural genes and presence of non-deletion mutations affecting the major a₂-globin gene.

Elevation of HbA₂ is the most important feature in the identification of heterozygous b-thalassaemia (Weatherall and Clegg, 1981). Nevertheless, a number of heterozygotes for b-thalassaemia may have normal or borderline HbA₂ levels (Galanello et al, 1994) (Table 2). The first groups of these atypical carriers are heterozygotes for a few mild b+-thalassaemia mutations, i.e. mutations associated with a consistent residual output of b-globin chains from the affected b locus. A typical example in this category is the case of heterozygotes for the b+ IVS I nt 6 T�C mutation. A normal HbA₂ level is also a characteristic feature of the b- and d-thalassaemia double heterozygotes, who still maintain low MCV and MCH values. These double heterozygotes should be differentiated from the a-thalassaemia carrier by globin chain synthesis analysis and/or gene analysis. d-globin gene analysis may be carried out by the same methods described for b-thalassaemia, using complementary primers or probes to the prevalent d-globin gene mutation in each population at risk. gdb- and db-thalassaemias also have normal HbA₂. However, db-thalassaemia is easily defined by the presence of thalassaemia-like haematological features and by a marked increase in HbF, which is heterogeneously distributed.

Another major problem in carrier screening is the identification of silent b-thalassaemia and the triple a-globin gene arrangement, both of which may lead to intermediate forms of thalassaemia by interaction with typical heterozygous b-thalassaemia. Silent b-thalassaemias are characterised by normal MCV and MCH values and normal HbA₂ and HbF, and are defined only by a slight imbalance in the a-globin/non a-globin synthesis (Gonzales-Redondo et al, 1989; Galanello et al, 1994). However, on examining the haematological features of these carriers, it should at times be possible to find some borderline HbA₂ or MCV and MCH values, which may signal the presence of atypical b-thalassaemia, thus calling for further studies. The most common silent b-thalassaemia is the b+ -101 G�T mutation, the other types being very rare (Gonzales-Redondo et al, 1989). Nevertheless, according to globin chain synthesis analysis, the phenotype resulting from the triple a-globin gene arrangement is at times completely silent. It is worth noting, however, that compound heterozygotes for silent or typical b-thalassaemia and double heterozygotes for typical b-thalassaemia and the triple a-globin gene arrangement result in attenuated forms of thalassaemia (Galanello et al, 1983; Thein et al, 1984; Kulozik et al, 1987).

An extreme though rare instance of the thalassaemia gene combination that may result in pitfalls of carrier diagnosis is the presence of a-, d- and b-thalassaemias together. This case may lead to a completely silent phenotype (Galanello et al, 1988).

The sickle cell trait is easily identified by haemoglobin electrophoresis or high-pressure liquid chromatography (HPLC). The phenotype of the sickle cell trait may be modified by co-inherited a-thalassaemia, which leads to reduced HbS levels in varying degrees, depending on the number of affected a-globin genes. However, this does not lead to carrier identification problems (Higgs et al., 1982). Based on this analysis of carrier identification, we suggest to follow the flow chart outlined in Figure 1 (Cao and Rosatelli, 1993).

The first group of tests includes MCV and MCH determination and haemoglobin chromatography by HPLC. HPLC may lead to the detection of the most common, clinically relevant haemoglobin variants, such as HbS, HbC, HbD Punjab, HbO Arab and HbE, all of which may result in a sickle disorder in homozygosity or compound heterozygosity.

HPLC may also be used to quantitate HbA₂ and HbF (Galanello et al, 1995). It should be noted that HPLC can also detect Hb Knossos, a mild b-thalassaemia allele, which is not defined by commonly used electrophoretic procedures in haemoglobin analysis. With this flow chart, the only cases that could be missed are the silent b-thalassaemia and the triple a-globin gene arrangement. In the presence of low MCH and MCV levels and high HbA₂ levels, a diagnosis of heterozygous b-thalassaemia is made. A phenotype characterised by microcytosis, hypochromia, normal-borderline HbA₂ and normal HbF may result from iron deficiency, a-thalassaemia, gdb- thalassaemia, b+d-thalassaemia or mild b-thalassaemia. After excluding iron deficiency by erythrocyte ZnPP determination and evaluation of transferrin saturation, the different thalassaemia determinants leading to this phenotype are discriminated by globin chain synthesis analysis and eventually by a-, d- and b-globin gene analysis. In the presence of normal MCV and borderline HbA₂ levels, we suspect the presence of a silent mutation (for instance b+ -101 C�T, b+ -92 C, or b+ IVS II nt 844 CCG) or the triple -globin gene arrangement, and proceed directly to a- and b-globin gene analysis, since in many of these cases the a/b ratio could be normal. Definition of the type of thalassaemia in these carriers is recommended only in the event of mating with a person with a typical, high HbA₂-b-thalassaemia, or with a carrier of an undetermined type of thalassaemia. In those rare cases that show normal or low MCH and MCV, normal or reduced HbA₂ levels and high HbF, we suspect the presence of db-thalassaemia, which should be differentiated from HPFH. We distinguish between b-thalassaemia and HPFH by analysing the red blood cell distribution of HbF, which is heterogeneous in db-thalassaemia and homogeneous in HPFH, by globin chain synthesis analysis (normal in HPFH and unbalanced in db-thalassaemia) and/or b cluster gene analysis.

In populations with a relatively low incidence of both b- and a-thalassaemia, screening by MCV and MCH or osmotic fragility could be acceptable, because in this condition the number of false negatives resulting from double heterozygosity for a- and b-thalassaemia may be very low.

MOLECULAR DIAGNOSIS

In couples at risk, identified by the above carrier detection procedure, the specific mutation is defined using one of the several polymerase chain reaction (PCR) based methods.

b-Thalassemia

b -Thalassemia is very heterozygous at the molecular level. To date, at least 150 different molecular defects have defined (Huisman and Carver, 1998). The majority of mutations affecting the b -globin gene are point mutations or oligonucleotide additions or deletions. Very rarely b -thalassemia results from the globin gene deletion mechanism. In spite of this marked heterogenity a limited number of molecular defects are prevalent in every population at risk (Table 3). This may be very useful in practice, because the most appropriate probes or primes can be selected according to the carrier's ethnic origin. Mutation detection is carried out on PCR-amplified-b -globin genes. The most commonly used screening procedures for known mutations today are reverse oligonucleotide hybridisation (RDB) with oligonucleotide probes or primer specific amplification (ARMS).

Reverse dot-blot hybridisation (RDB)

Reverse dot-blot hybridisation uses membrane-bound allele-specific oligonucleotide probes that hybridise to the complementary sequence of the PCR product prepared using patient DNA as the starting template (SAIKI et al, 1989) ( Figure 2). In this format, multiple pairs of normal and mutant allele-specific oligonucleotides can be placed on a small strip of membrane. Hybridisation with PCR-amplified b -globin gene DNA will detect any of the mutations screened in a single procedure. Up to 20-30 mutations have indeed been screened in one single step.

Primer-specific amplification (ARMS)

With this method, the target DNA fragment is amplified using a common primer and either of two primers: a primer complementary to the mutation to be detected (b -thalassemia primer), or a primer complementary to the normal DNA at the same position (normal primer). Another b -globin gene fragment is simultaneously co-amplified to control the amplification step of the procedure (Newton et al, 1989). Normal DNA is amplified only by the normal primer DNA from homozygotes only by the b -thalassemia primer and DNA from heterozygotes by both primers.

 Other known mutation-detection procedures

Other methodologies which could be used for mutation detection in b -thalassemia carriers are oligonucleotide ligation assay (Nickerson et al, 1990), restriction enzyme digestion (Pirastu et al, 1989), denaturing gradient gel electrophoresis (Myer et al, 1985, Cai and Kan, 1990; Rosatelli et al, 1992), and primer specific restriction map modification (Gasparini et al 1992).

Unknown mutations

For parents in whom the definition of the b -thalassemia mutation is not made by one the procedures above described, characterisation of b -thalassemia is obtained by denaturing gradient gel electrophoresis, chemical mismatch cleavage analysis (Orita et al, 1989) followed by direct sequencing (Sanger et al, 1977) on amplified single-strand DNA (Gyllesten and Erlich, 1988).

The most widely used among these methods is DGGE, a gel system that separates DNA fragments as a function of melting temperature. The b -globin gene is amplified by using five to seven pairs of primers, one pair of each with an added GC clamp. Normal DNA and DNA from homozygous b -thalassemia produce a single band with a typical migration pattern, that depend on the globin gene sequences contained in the amplified fragments (Rosatelli et al, 1992a). DNA from heterozygous b -thalassemia results in the formation of four bands, two of which are homoduplexes of the normal and mutated allele, while the other two are heteroduplexes resulting from annealing the strands of the normal allele to those of the mutated allele ( Figure 3).

After localisation by DGGE, the mutation is defined by direct sequencing of the DNA contained in the fragment. Direct b -globin gene sequencing may be carried out manually or automatically. Alternatively, the unknown mutation may be detected by a fully automated integrated system for DNA fragment analysis (AHPLC).

If a mutation is not detected by DGGE analysis, we search for the presence of small deletions by polyacrylamide gel electrophoresis of the PCR-amplified products prepared for ARMS or RDB analysis. This may lead to the detection of small deletions of the b -globin gene, suspected from very high HbA ₂ levels. Larger deletions of the cluster may be identified with restriction fragment length polymorphism analysis by Southern blot or PCR-based procedures.

In a very limited number of cases, direct sequencing from position ?600 to 60 bp downstream from the b -globin gene failed to detect a mutation causing b -thalassemia (Murru et al, 1990, 1992; Rosatelli et al, 1992b). In these cases, the molecular defect may reside either in the Locus Control Region at the b -globin gene cluster, or in one of the genes outside the b -globin gene region encoding for DNA-binding protein that regulates the function of the b -globin gene (for a review see Orkin , 1990 and Townes and Behriger 1990).

Prediction of a mild phenotype

Homozygosity or compound heterozygosity for b-thalassaemia usually results in the clinical phenotype of transfusion-dependent thalassaemia major. However, a substantial proportion of these homozygotes develop milder forms that range in severity from the asymptomatic carrier state to thalassaemia major (thalassaemia intermedia).

The main pathophysiological determinant of the severity of the thalassaemia syndrome is the extent of a/nona chain imbalance. In other words any factor capable of reducing this a/nona imbalance may have an ameliorating effect on the clinical picture. The most clinically relevant mechanism that consistently results in thalassaemia intermedia is co-inheritance of homozygosity or compound heterozygosity for a silent or mild b-thalassaemia allele, namely a b-thalassaemia defect associated with a consistent residual output of b chains from the affected b-globin locus (Gonzales ? Redondo et al, 1989, Rosatelli et al, 1994, 1995). By contrast, compound heterozygotes for a mild or silent and a severe mutation may result in a spectrum of phenotypes including severe and mild forms.

Other mechanisms capable of ameliorating the phenotype of homozygous b-thalassaemia are the co-inheritance of a-thalassaemia (Wainscoat et al, 1983; Galanello et al, 1989) or genetic determinants capable of sustaining the continuous production of g-chains in adult life, thereby reducing the extent of the a/non-a chain unbalance (Rochette et al, 1994).

However, neither mechanism can be used to predict a mild phenotype prospectively, because neither results in a consistent effect. It is worth noting that thalassaemia intermedia may also result from heterozygosity for hyper-unstable Hb or a compound heterozygous state for typical heterozygous b-thalassaemia and the triple a-globin arrangement.

Nevertheless, in many cases the determinantfor the mild phenotype has not been defined so far.

a-Thalassaemias

Deletion a� or a+-thalassaemias are detected by PCR using two primers flanking the deletion breakpoints that amplify a DNA segment only in presence of the specific deletions (reviewed in Kattamis et al, 1996). As a control, DNA from a normal chromosome is simultaneously amplified using one of the primers flanking the breakpoint and a primer homologous to a DNA region deleted by the mutation. In addition, non-deletion a-thalassaemias are detected by restriction enzyme analysis on selectively amplified a₁- and a₂-globin genes or, when restriction is not applicable, by dot blot analysis with allele-specific oligonucleotide probes. Besides defining the molecular defect in carriers of a-thalassaemia, a-globin gene analysis can also discriminate between heterozygous a-thalassaemia and double heterozygosity for d- and b-thalassaemia or gdb-thalassaemia. Definition of the a-globin gene arrangement could also be useful to predict the clinical phenotype of homozygous b-thalassaemia.

Sickle cell anaemia

Sickle cell anaemia most commonly results from homozygosity for the HbS mutation. Nevertheless, it may also be caused by compound heterozygosity for the HbS mutation and HbC, b+, or b �-thalassemia, or other rare haemoglobin variants such as HbO Arab. Molecular diagnosis for carriers of these conditions may be carried out with the same procedure described for b-thalassaemia, and especially by dot blot analysis with allele-specific probes or primer-specific amplification.

The severity of sickle cell anaemia may be modulated by a number of co-inherited modifying genes, a-thalassaemia and HPHF being the most noteworthy (Steinberg, 1996). The only determinant consistently associated with a milder phenotype is presence of high HbF resulting from the group of heterogeneous conditions discussed previously under genetic counselling for couples at risk for b-thalassaemia.

In order to predict the clinical phenotype appropriately, during genetic counselling to couples at risk, these modifying factors should be defined by appropriate procedures (see section on b-thalassaemia).

Prenatal diagnosis

Prenatal diagnosis of both a- and b-thalassaemia was accomplished for the first time in the 1970s with globin chain synthesis analysis of fetal blood obtained by fetoscopy or placental aspiration (Kan et al, 1975). Molecular definition of the thalassaemias, the development of procedures for their detection by DNA analysis, and the introduction of chorionic villus sampling in the last decade have dramatically improved prenatal detection of these disorders. For a short period the diagnosis of thalassaemia was obtained either indirectly by polymorphism analysis (Kan et al, 1980) or directly by oligonucleotide hybridisation on electrophoretically separated DNA fragments (Pirastu et al, 1983). Today thalassaemias are detected directly by the analysis of amplified DNA from fetal trophoblast or amniotic fluid cells.

Fetal DNA sampling

Fetal DNA for analysis can be obtained from amniocytes or from chorionic villi. At present the most widely used procedure is chorionic villi sampling, mainly because of the clear advantage of being carried out during the first trimester of pregnancy, generally at the 10th-12th week of gestation (Hogge et al, 1986; Cao et al, 1987; Brambati et al, 1988). The risk of fetal mortality associated with this method is of the order of 1%. Chorionic villi may be obtained transcervically or transabdominally. We prefer the transabdominal route for several reasons, mainly because it has a low infection rate, a lower incidence of amniotic fluid leakage, because it is a simple procedure, and also because it is largely preferred by pregnant women.

Fetal DNA analysis

Fetal DNA is analysed using the methods described earlier for the detection of known mutations during carrier definition. To limit the possibility of misdiagnosis, we analyse chorionic villous DNA with two different procedures: i.e. RDB hybridisation and primer-specific amplification (ARMS).

Misdiagnosis may occur for several reasons: failure to amplify the target DNA fragment, mispaternity, maternal contamination, and sample exchange. Misdiagnosis for failure of DNA amplification is obviously limited by the double approach described above. To avoid misdiagnosis resulting from mispaternity or maternal contamination, we carry out DNA polymorphism analysis parallelly with mutation analysis. In addition to this, the presence or effect of maternal contamination could also be limited by careful dissection of the maternal decidua from the fetal trophoblast under the inverted microscope, and by the fact that a minimal amount (about 3 mg) of chorionic villi are requested to reduce the chances of co-amplifying the DNA from the maternal decidua.

The advent of DNA amplification has made it possible to analyse the genotype of a single cell. This has paved the way for pre-implantation or even pre-conceptional diagnosis (Monk and Holding, 1990; Handyside et al, 1992).

Pre-implantation may be carried out by a biopsy of the blastula, obtained by washing the uterine cavity after in vivo fertilisation, or by analysis of a single blastomere from an eight-cell embryo after in vitro fertilisation. Pre-conception diagnosis is based on the analysis of the first polar body of unfertilised eggs, and may lead to distinguish between unfertilised eggs that carry the defective gene and those without the defect. By fertilising in vitro only the eggs without the defect and replacing them in the mother, a successful pregnancy with a normal fetus can be obtained. Of course, the genotype of the fetus will be checked further by chorionic villus biopsy.

Successful pregnancies following the transfer of human embryos in which a single gene defect has been excluded, have been reported.

Pre-implantation and pre-conception gamete diagnoses are very useful for couples against pregnancy termination for ethical reasons, and especially for those who have already had therapeutic abortions due to genetic risks. At present, however, its use in routine monitoring of pregnancies at risk is precluded by the technical demand for these procedures, the difficulty organising the service, and the high costs.

POPULATION SCREENING AND COUNSELLING FOR HAEMOGLOBIN DISORDERS

Couples at risk for haemoglobin disorders may be identified retrospectively, i.e. following the birth of an affected child, or prospectively by analysing childless spouses. The benefit from prospective identification is obviously greater, because it gives parents the opportunity of planning a family without disease, and it alleviates the health burden to society. Prospective identification of couples at risk is carried out by population screening. To date programmes aimed at prospective identification and counselling of couples at risk for inherited haemoglobinopathies have been carried out significantly only for b-thalassaemia. These programmes are ongoing in several areas at risk in the Mediterranean basin, such as Cyprus, Greece, several regions of continental Italy, and Sardinia (Angastiniotis et al, 1995; Cao et al, 1996; Loukopoulos, 1996).

All these programmes are characterised by intensive education and involvement of the population, screening of prospective couples and non-directive counselling. They have been very successful, because the population was well informed about thalassaemia and the methodology for its prevention, and there were no consistent adverse effects on those found to be carriers. Furthermore, in populations in the Mediterranean area where screening and counselling have been introduced, a marked decline has been observed in the incidence of thalassaemia major, the homozygous state of b-thalassaemia (Angastiniotis et al, 1995; Cao et al, 1996; Loukopoulos, 1996).

Very limited prospective screening for a-thalassaemia is carried out in a few areas of South East Asia (Hong-Kong, Southern China, Thailand, Taiwan), and for sickle cell anaemia in the Caribbean, the United Kingdom and the USA.

FUTURE PROSPECTS

Technically, in carrier screening and prenatal diagnosis we can realistically predict further simplification and full automation of the procedures for the detection of the b-thalassaemia mutation. Primer-specific amplification and reverse oligonucleotide hybridisation, for instance, could easily become fully automated. An oligonucleotide microchip assay has been proposed recently for the large-scale detection of mutations in genetic diseases, including b-thalassaemia. Given the alternative features of high throughput, automation and modest cost, the DNA chip has the potential to become a valuable method in future applications of mutation detection in medicine (Yershov et al, 1996).

The foreseeable progressive reduction in the cost of DNA analysis may lead to the use of mutation detection as a future screening method, thus skipping all carrier detection steps based on haematological analysis.

Simplification of pre-implantation and pre-conception gamete diagnosis may lead to a more extensive use of the procedure in the future, especially by couples against pregnancy termination. However, the most relevant advance would be fetal diagnosis by analysis of fetal cells in maternal circulation.

So far, many methods have been proposed for analysis, but none has given reliable results (Bianchi et al, 1990; Ganshirt-Ahlert et al, 1993).

Point mutations responsible for b-thalassaemia or sickle cell anaemia have recently been identified successfully on fetal cells with a procedure based on the density gradient separation of mononuclear cells from maternal blood, the enrichment of fetal cells by magnetically activated cell sorting using the anti-transferrin receptor antibody, the identification of fetal cells by immunostaining with anti-fetal or embryonic haemoglobin antibodies, the isolation of nucleated red cells by micro-dissection under light microscopy, and non-radioactive PCR analysis (Cheung et al, 1996). The simplification and partial automation of this procedure may lead to the introduction of prenatal diagnosis by analysis of fetal cells in the maternal circulation in clinical practice.

Nevertheless, the most important challenge for the future is the organisation of similar genetic preventive programmes to those ongoing in the Mediterranean area, in parts of the world where b-thalassaemia is prevalent, namely the Middle East, the Indian subcontinent, and the Far East. Nonetheless, the resources for population education and the present state of technical development seem to preclude the realisation of such a programme.

Acknowledgements

We thank Rita Loi for editorial assistance.

References

• Angastiniotis M, Modell B, Boulinzhenkov V (1995) Prevention and control of haemoglobinopathies.Bulletin of the World Health Organization73: 375-86.
• Bianchi DW, Flint Af, Pizzimenti MF et al (1990) Isolation of fetal DNA from nucleated erythrocytes in maternal blood.Proceeding of the National Academy of Sciences (USA)87: 3279-83
• Brambati B, Lanzani A, Oldrini A (1988) Transabdominal chorionic villus sampling, clinical experience of 1159 cases.Prenatal Diagnosis8: 609-13.
• Cai SP, Kan YW (1990) Identification of the multiple b -thalassemia mutations by denaturing gradient gel electrophoresis.Journal of Clinical Investigation85: 550-3.
• Cao A, Rosatelli C, Galanello R: Control of b -thalassemia by carrier screening, genetic counselling and prenatal diagnosis: the Sardinian experience. In 1996 Variation of the human genome. Wiley, Chichester (Ciba Foundation Symposium, 197) 137-55
• Cao A, Cossu P, Monni G, Rosatelli C (1987) Chorionic villus sampling and acceptance rate of prenatal diagnosis.Prenatal Diagnosis7: 531-3.
• Cao A, Galanello R, Rosatelli C (1998) Prenatal diagnosis and screening of the haemoglobinopathies. Bailli�re's Clinical Haematology. Sickle Cell Disease and Thalassemia.vol.11No. 1, Bailli�re Tindall. 215-38.
• Cheung M-C, Goldberg JD, Kan YW (1996) Prenatal diagnosis of sickle cell anemia and thalassemia by analysis of fetal cells in maternal blood. Nature Genetics14: 264-8.
• Cotton RGH, Rodrigues NR, Campbell RD (1988) Reactivity of cytosine and thymine in single-base-pair mismatches with hydroxylamine and osmium tetroxide and its application to the study of mutations.Proceeding of the National Academy of Sciences (USA)85: 4397-401.
• Galanello R, Ruggeri R, Addis M, Paglietti E, Cao A (1981) Hemoglobin A2 in iron deficiency b -thalassaemia heterozygotes.Hemoglobin5: 613-8.
• Galanello R, Ruggeri R, Paglietti E et al (1983) A family with segregating triplicated a -globin loci and b -thalassemia.Blood62: 1035-40.
• Galanello R, Paglietti ME, Addis M et al (1988) Pitfalls in genetic counselling for b -thalassemia: an individual with 4 different thalassemia mutations.Clinical Genetics33: 151-5.
• Galanello R, Barella S, Ideo A et al (1994) Genotype of subjects with borderline Hb A2: Implication for b -thalassemia carrier screening. American Journal of Hematology46: 79-81.
• Galanello R, Barella S, Gasperini D et al (1995) Evaluation of a new automatic HPLC analyzer for thalassemia and hemoglobin variants screening.Journal of Automatic Chemistry17: 73-6.
• Ganshirt-Ahlert D, Borjesson-Stoll R, Burschyk M et al (1993) Detection of fetal trisomies 21 and 18 from maternal blood using triple gradient and magnetic cell sorting. American Journal of Reproduction Immunology30: 194-201.
• Gasparini P, Bonizzato A, Dognini M and Pignatti PF (1992) Restriction site generating-polymerase chain reaction (RG-PCR) for the probeless detection of hidden genetic variation: application to the study of some common cystic fibrosis mutations.Molecular and Cellular Probes6: 1-7.
• Gonzales-Redondo JM, Stoming TA, Kutlar A, et al (1989) A C� T substitution at nt ?101 in a conserved DNA sequence of the promoter region of the b -globin gene is associated with "silent" b -thalassemia.Blood73: 1705-11
• Gyllensten UB, Erlich HA (1988) Generation of single stranded DNA by the polymerase chain reaction and its application to direct sequencing of the HLA-DQ locus.Proceeding of the National Academy of Sciences (USA)85: 7652-6.
• Handyside AM, Lesko JG, Tarin JJ et al (1992) Birth of a normal girl after in vitro fertilization and preimplantation diagnostic testing for cystic fibrosis.New England Journal of Medicine327: 905-9.
• Higgs DR, Aldridge BE, Lamb J et al (1982) The interaction of alpha-thalassemia and homozygous sickle-cell disease.New England Journal of Medicine306: 1441-6.
• Hogge WA, Schonberg SA, Golbus MS (1986) Chorionic villus sampling: experience of the first 1000 cases.American Journal of Obstetrics and Gynaecology154: 1249-52.
• Huisman THJ and Carver MFH (1998) The b - and d -thalassemia repository (8^th edn).Hemoglobin19: 213-36.
• Kan YW, Golbus MS, Klein P, Dozy AM (1975) Successful application of prenatal diagnosis in a pregnancy at risk for homozygous b -thalassemia.New England Journal of Medicine292: 1096-9.
• Kan YW, Lee KY, Furbetta M et al (1980) Polymorphism of DNA sequence in the b -globin gene region: application to prenatal diagnosis of b -thalassemia in Sardinia.New England Journal of Medicine302: 185-8.
• Kattamis AC, Camaschella C, Sivera P et al (1996). Human a -thalassemia syndromes: detection of nolecular defects.American Journal of Hematology53: 81-91.
• Kulozik AE, Thein SL, Wainscoat JS et al (1987) Thalassemia intermedia: interaction of the triple a -globin gene arrangement and heterozygous b -thalassaemia.British Journal of Haematology66: 109-12.
• Loukopoulos D (1996) Current status of thalassemia and the sickle cell syndromes in Greece.Seminars in Hematology33:76-86
• Melis MA, Pirastu M, Galanello R et al (1983) Phenotypic effect of heterozygous a and b �-thalassaemia interaction.Blood62: 226-9.
• Monk M, Holding C (1990) Amplification of a b -haemoglobin sequence in individual human oocytes and polar bodies.Lancet325: 985-8.
• Murru S, Loudianos G, Vaccargiu S et al (1990) A b -thalassemia carrier with normal sequences within the b -globin gene.[letter].Blood76: 64-5.
• Murru S, Loudianos G, Porcu S, et al, (1992) A b -thalassaemia phenotype not linked to the b -globin cluster in an Italian family.British Journal of Haematology81: 283-7.
• Myers RM, Fisher SG, Lerman LS , Maniatis T (1985) Nearly all single base substitution in DNA fragments joined to a GC-clamp can be detected by denaturing gradient gel electrophoresis.Nucleid Acids Research13: 3131-5.
• Newton CR, Graham A, Hepteinstall LE et al (1989) Analysis of any point mutation in DNA. The amplification refractory mutation system (ARMS).Nucleic Acid Research17: 2503-16.
• Nickerson DA, Kaiser R, Loggin S et al (1990) Automated DNA diagnostic using an Elisa-based oligonucleotide ligation Assay.Proceeding of the National Academy of Sciences (USA)87: 8923-7.
• Orita M, Iwahana H, Kanazawa H et al (1989) Detection of polymorphisms of human DNA by gel electrophoresis as single-strand conformation polymorphism.Proceedings of the National Academy of Science (USA)86: 2766-70.
• Orkin SE (1990) Globin gene regulation and switching: circa 1990.Cell63: 665-72.
• Pirastu M, Kan YW, Cao A et al (1983) Prenatal diagnosis of b -thalassemia. Detection of a single nucleotide mutation in DNA.New England Journal of Medicine309: 284-7.
• Pirastu M, Ristaldi MS, and Cao A (1989) Prenatal diagnosis of b -thalassemia based on restriction endonuclease analysis of amplified fetal DNA. Journal of Medical Genetics26: 363-7.
• Rochette J, Craig JE, Thein SL (1994) Fetal hemoglobin levels in adults.BloodReviews8: 213-224.
• Rosatelli C, Falchi AM, Scalas MT et al (1984) Hemat
• ological phenotype of the double heterozygous state for a and b thalassemia.Hemoglobin8: 25-35.
• Rosatelli MC, Dozy A, Faa V et al (1992a) Molecular characterization of b -thalassemia in the Sardinian population.Americal Journal of Human Genetics50: 422-6.
• Rosatelli C, Tuveri T, Scalas MT et al. (1992b) Molecular screening and fetal diagnosis of b -thalassemia in the Italian population.Human Genetics83:590-2.
• Rosatelli MC, Pischedda A., Meloni A., Saba L., Pomo A., Travi M., Fattore S., Cao A. (1994) Homozygous b -thalassemia resulting in the b -thalassemia carrer state phenotype.British Journal of Haematology, 88: 562-565.
• Rosatelli MC, Faa V, Meloni A, Fiorenza F, Galanello R, Gasperini D, Amendola G, Cao A, (1995) A promoter mutation, C � T position ? 92, leading to silent b -thalassemia.British Journal of Haematology, 90: 483-485.
• Saiki RK, Walsh PS, Levenson CH, Erlich HA (1989) Genetic analysis of amplified DNA with immobilized sequence-specific oligonucleotide probes.Proceedings of the National Academy of Sciences (USA)86: 6230-4.
• Sanger F, Micklen S, Coulson AR (1977) DNA sequencing with chain terminating inhibitors.Proceedings of the National Academy of Sciences (USA)74: 5463-7.
• Steinberg MG (1996) Modulation of the phenotypic diversity of sickle cell anemia.Hemoglobin20: 1-19.
• Thein SL, Sl-Hakim I and Hoffbrand AB (1984) Thalassaemia intermedia: a new molecular basis.British Journal of Haematology56: 333-7.
• Townes TM, Behringer RR ( 1990) Human globin locus activation region (LAR): role in temporal control.Trends in Genetics6: 219-23.
• Weatherall SJ, Clegg JB (1981) The Thalassaemia Syndromes. 3rd edn. Blackwell Scientific Publication, Oxford.
• WHO Scientific Group (1996) Control of hereditary diseases..WHO Technical Report Series 865.
• Wainscoat JS, Kanavakis E. Wood WG et al (1983) Thalassemia intermedia interaction of a and b thalassemia.British Journal of Haematology53: 411- 416
• Yershov G, Barsky V, Belgovskiy A, et al. (1996) DNA analysis and diagnostics on oligonucleotide microchips. Proceedings of the National Academy of Sciences (USA)93: 4913-8.
  
  
  
  Figures

  Figure 1: flow chart for b -thalassaemia carrier screening

  Figure 2: Screening for the most common b -thalassaemia mutations by reverse dot blot analysis.

  Top: the whole b -globin gene is enzymatically amplified using primer A and B.

  Bottom: A b -thalassaemia heterozygote for b IVS I-110 mutation.

  W = wild type oligonucleotide hybridization

  M = mutated oligonucleotide hybridization

  Figure 3: Denaturating gradient (42% � 72%) gel electrophoresis for detecting heterozygotes for b -thalassemia. The b -globin genotype of each subject is indicated on top.