THALASSEMIA results from an imbalance in the synthesis of a- and ß-globin Chains. To date we know of > 1000 naturally occurring inherited mutations that affect either the structure or synthesis of the a- and ß-like globin chains.  Thalassemia occurs when there is a deficiency in a globin expression and ß thalassemia occurs when ß globin synthesis is downregulated. Many different mutant alleles have been selected to reach very high frequencies in tropical and subtropical regions of the world because heterozygotes are, to some extent, protected from the effects of falciparum malaria. Carriers of a and ß thalassemia are clinically normal but the associated hypochromic microcytic blood picture requires careful diagnostic screening to differentiate it from other conditions with similar hematological pictures (e.g., iron deficiency and sideroblastic anemia). In areas where a and ß thalassemia commonly occur, compound heterozygotes and homozygotes for these mutations may suffer from the clinically severe forms of  thalassemia (HbH disease and Hb Bart’s hydrops fetalis) and ß thalassemia (ß thalassemia major and the less severe ß thalassemia intermedia). A summary of the key clinical syndromes (a) and pathophysiology of a (b) and ß (c) thalassemia. The consequences of reduced a and ß globin synthesis are indicated by dotted lines. Note, a thalassemia affects both the fetus and adult whereas ß thalassemia only becomes apparent after birth. HbF may persist or become re-activated during adult hematopoiesis in ß thalassemia.

Epidemiology : thalassemia is the world’s most common monogenic disorder. In areas where thalassemia is prevalent, the carrier frequency of  thalassemia varies from 1% (e.g., in Southern Spain) to 90% (e.g., the tribal populations of India). Similarly, the carrier frequency of ß thalassemia may vary from 1% (e.g., in Northern Italy) to 50–70% (e.g., ß-E in some regions of southeast Asia). Considerable variation occurs even within a single country. It has been estimated that worldwide there are 270 million carriers of mutant globin alleles which can potentially cause severe forms of thalassemia and hemoglobinopathy (including sickle cell disease), and every year some 300–400,000 severely affected infants are born^ref. > 95% of these affected births occur in Asia, India and the Middle East. However, over the past few decades there has been considerable mobility in the world’s populations so that now in Australasia, Europe and North America a significant proportion of recent immigrants originate from countries in which thalassemia commonly occurs. Therefore, the management of families from ethnic minorities, at risk of producing offspring with severe life-threatening forms of thalassemia, has become a new important aspect of hematologic practice. It has also been noted over the past 50 years that, as the world economy improves and the overall infant mortality rate falls, thalassemia is rapidly emerging as a major economic and health burden in developing countries since infants with thalassemia who would otherwise have died from infection and malnutrition now survive and require lifelong treatment. Therefore, worldwide, thalassemia is an enormous and ever-increasing hematologic problem in both developed and developing countries^ref. Today’s epidemiology of thalassemia is strikingly different from that of the past. Thalassemia is now a heterogeneous group of diseases with varied ethnicities, phenotypes and treatments. In the past, Hb E-ß-thalassemia and Hb H disease were rarely seen in North America and Europe. Now, as a result of changing demographics, these disorders have become more common than classical ß-thalassemia major in many regions^{ref1, ref2, ref3, ref4, ref5}. HbH, HbH-Constant Spring and homozygous a-thalassemia affect at least a million people worldwide^ref. The natural history of these disorders is highly variable and poorly studied. The phenotype, for patients with similar mutations, can range from asymptomatic to transfusions dependency. Ethnic groups originating from regions with malaria have a high frequency of thalassemia mutations. There are almost 300 million carriers of hemoglobin disorders in the world, with the majority living in South East Asia^{ref1, ref2, ref3, ref4}. Worldwide, the Asian, Indian and Middle Eastern regions account for 95% of thalassemia births. The frequency of a-thalassemia reaches 25% in Thailand, and HbE approaches 60% in many regions of Thailand, Laos and Cambodia. Hb Constant Spring is found in 1–10% of the population in these areas. The WHO estimates that Thailand will have over 250,000 symptomatic thalassemia patients diagnosed over the next few decades^{ref1, ref2, ref3, ref4, ref5}. At least 100,000 new cases of HbE-ß-thalassemia are expected. Similar estimates are predicted for India, Sri Lanka and Malaysia^ref. In southern China, with a population of over 350 million, approximately 5% are carriers for a-thalassemia and 4% for ß-thalassemia or HbE^ref. In the last decade, 75% of immigrants to North America came from areas where thalassemia mutations were prevalent. These demographic changes have resulted in thalassemia becoming a more significant health problem in North America. Hb E-ß-thalassemia and a-thalassemia disorders are now the most common thalassemias in California. In response to the high prevalence of -thalassemia and its associated morbidity, California has initiated universal newborn screening for HbH and HbH-Constant Spring^ref. Characterizing the molecular defects underlying  and ß thalassemia has transformed the management of these conditions in many developed countries. Over the past 25 years, throughout Europe, Australasia and North America comprehensive control programs involving education, counselling and pre-natal diagnosis have succeeded in limiting the numbers of new births of affected individuals. One of the best-documented examples of this approach has been in Cyprus where thalassemia was first recognized in the 1940s following the eradication of malaria. With a population size of ~700,000 and a carrier frequency for ß thalassemia of 15–17%, in the early 1970s it was estimated that to treat all severely affected children would require 78,000 units of blood per year, 40% of the population would be donors and the total cost to the health service would exceed the island’s health budget^ref. With judicious application of a control program there are now fewer than 600 severely affected patients with thalassemia in Cyprus and only 2-3 new cases arising each year. A similar highly effective control program was also established by pioneering work in Sardinia^ref. In First World countries the outlook for the relatively small numbers of patients with ß thalassemia major and intermedia has also dramatically improved over the last 25 years.
Pathogenesis : uneffective erythropoiesis : decreased or no synthesis of one or more kinds of globin => precipitation of the remaining kinds => peripheral hemolysis / hemocatheresis. Complicated by parvovirus B₁₉ infection

• a-thalassemias => HbH (b₄)

Epidemiology : the gene frequencies of a-thalassemia exceed those of ß-thalassemia. Expecially in Southeast Asia and southern China^{ref1, ref2, ref3, ref4} : 5÷10 % in Thailand. 7% of the world's population are carriers, and 400,000 affected babies are born each year. The WHO predicts a doubling of these statistics within the next few decades. Once rare in California, it has become the most frequent disorder detected in the state newborn screening program of over 700,000 annual births. In India, 100 million people are carriers of this disorder, and at least 10-thousand babies are born every year with thalassemia. Thousands of affected patients live in the Middle East, Mediterranean region and North America^{ref1, ref2}.
Aetiology : usually deletion type. The clinical severity is strongly influenced by the type of mutation
• a⁰-thalassemia : a₂ and a₁ deletion => -- locus
• heterozygous : only 2 a genes (--/a₂a₁) : a-thalassemia minor
• homozygous : no a gene (--/--) : hydrops foetalis^ref
• 80% Hb Bart (g₄) in utero has an extremely high oxygen affinity and therefore delivers little oxygen to fetal tissues. The severe hypoxia results in cardiac failure, massive ascites and intrauterine death^ref. 5 cases survived beyond the newborn period, all with hypospadias secondary to the in utero edema leading to failure of fusion of urogenital folds or due to defect or deletion of another gene at 16p13.3^ref.
• 20% Hb Portland (z₂g₂)
Symptoms & signs : congenital malformations associated with homozygous a-thalassemia include hypospadias, other genitourinary defects, and limb malformations. Infants surviving to delivery without prenatal intervention are usually hydropic and commonly have neurologic impairment
Laboratory examinations : in pregnancies known to be at risk, chorionic villus sampling (CVS) with molecular analysis identifies homozygous a-thalassemia within the first months of pregnancy. The ethical issues of managing the fetus known to have homozygous a-thalassemia are complex.
Therapy : intrauterine transfusions following early detection of homozygous a-thalassemia have resulted in the birth of several nonhydropic infants, some but not all of whom have no significant neurologic abnormalities or congenital anomalies. Affected infants who survive gestation and the neonatal period subsequently require chronic transfusion therapy or may be appropriate candidates for allogeneic hematopoietic stem cell transplantation^{ref1, ref2}. Occasionally infants with homozygous a-thalassemia are born without hydrops even in the absence of intrauterine transfusions^{ref1, ref2}. Increased levels of Hb Portland, a normally functioning embryonic hemoglobin, may be responsible for this improved clinical course. Obstetric complications and the necessity for long-term transfusion therapy are serious considerations. Increased risk of both maternal and fetal morbidity should be included in counseling families at risk for an affected fetus. The prognosis for patients with the hemoglobinopathies that are newly emerging in North America and elsewhere is strongly dependent on the health care system’s approach to these disorders. Prenatal diagnosis, neonatal screening, comprehensive care and access to new therapies including transfusion therapy, stem cell transplantation and HbF–enhancing agents are important components of care. Simple approaches such as the use of tea to inhibit iron absorption may have profound benefits. Addressing the economic and cultural barriers and increasing the knowledge of physicians about these disorders should improve the overall care of affected patients.
• a⁺-thalassemia : a₂ or a₁ down-regulation,
• mutation
• Hb Constant Spring (142 stop => Gln => 31 extra amino acid residues at the C terminus of the a chain). Hb CS^{ref1, ref2, ref3, ref4, ref5} is the most common nondeletional a-thalassemia mutation associated with Hb H disease. The gene frequency of Hb CS approaches 8% in Southeast Asia. Its mRNA is highly unstable, producing < 1% of the total protein output of a normal gene. Cells containing Hb CS become overhydrated, modifying the expected microcytosis. In comparison with cells from patients with deletional HbH disease, cells containing Hb CS have increased hemoglobin bound to the membrane, increasing the rate of hemolysis^ref (Styles LA. Hemoglobin H-Constant Spring disease: an under recognized, severe form of  thalassemia. J Pediatr Hematol/Oncol. 1997;4:69–74). Hb H/CS disease has significantly more ineffective erythropoiesis and erythroid apoptosis than HbH disease^ref. The laboratory and clinical course of Hb H/CS disease is more severe than Hb H disease^{ref1, ref2, ref3, ref4}.
• hemoglobin G : any of various abnormal hemoglobins with an amino acid substitution on the a chain
• the most common one is hemoglobin G Philadelphia, which causes a-thalassemia.
• hemoglobin I : an abnormal hemoglobin resulting from an amino acid substitution in the a chain, causing a-thalassemia.
• deletion of a₁ or a₂ genes => a₂-  or a₁- locus
• heterozygous : only 3 a genes (-a₁/a₂a₁ or a₂-/a₂a₁) : silent carrier (normal MCV)
• homozygous : only 2 a genes (-a₁/-a₁ or a₂-/-a₁ or a₂-/a₂-) : a-thalassemia minor
Epidemiology : on the north coast of Papua New Guinea malaria transmission is intense and a⁺-thalassemia affects > 90% of the population (homozygotes comprise approximately 55% and heterozygotes 37% of the population)^ref
Symptoms & signs : most patients have moderately severe splenomegaly. Rarely, Hb H/CS disease and other nondeletional HbH disorders have caused fatal hydrops foetalis syndrome^ref.
Laboratory examinations : the average hemoglobin is 2 g/dL less than in deletional HbH disease. The mean MCV is a near normal 72 fL compared to 59 fL for deletional Hb H disease, since cells containing Hb CS become overhydrated, modifying the expected microcytosis.
Therapy : over 50% require splenectomy, which often results in improved hemoglobin levels but is associated with a high rate of portal vein thrombosis. 90% of patients with Hb H/CS disease have been intermittently transfused, and up to 40% have required repeated transfusions, particularly in early infancy and in later adulthood. Iron overload occurs in 75% of patients by adulthood
• a⁰-a⁺-thalassemia : 1 only a gene (a₂-/-- or -a₁/--) => hemoglobin H disease (b₄)

Aetiology :
• deletions on chromosome 16 (75%) cause a milder form of the disorder
• 2 deletions + a point mutation or insertion in the a globin gene (25%). Non-deletional HbH is often severe and likely to require RBC transfusions^{ref1, ref2, ref3} (Styles LA. Hemoglobin H-Constant Spring disease: an under recognized, severe form of  thalassemia. J Pediatr Hematol/Oncol. 1997;4:69–74)
In both groups, however, there is marked phenotypic variability.
Symptoms & signs : chronic hemolytic anemia associated with splenomegaly
Laboratory examinations : red blood cell hypochromia, anisocytosis, and poikilocytosis are accompanied by inclusion bodies detectable by supravital staining
Pathogenesis : HbH is unstable and forms intracellular precipitates that result in early cell death. Hemolysis rather than ineffective erythropoiesis is the primary cause of anemia in HbH disease.
Symptoms & signs : HbH disease may be a mild disorder, but recent studies suggest its clinical course is often more severe than previously recognized. Pregnancy is often associated with an increased severity of anemia as well as pre-eclampsia. Iron overload and iron-induced heart failure have occasionally occurred in patients not receiving transfusions
Laboratory examinations : the diagnosis of Hb H may be difficult. Hb H and Hb Barts are fast moving hemoglobins that may appear on electrophoresis. However, they are unstable and often go undetected. Patients with Hb H disease have > 20% Hb Barts at birth. In the California newborn screening for Hb H disease, 90% of cases were diagnosed by the finding of elevated levels of Hb Barts on high performance liquid chromatography (HPLC). DNA testing was required to distinguish Hb H Constant Spring (CS) from Hb H disease^ref. After the newborn period, Hb Barts is no longer seen on electrophoresis, but Hb H may be detected. Red cells from patients with Hb H form inclusions when stained with brilliant cresyl blue or methylene blue, but these tests are nonspecific. DNA testing is often helpful in identifying both deletional and nondeletional mutations. The diagnosis of deletional Hb H disease is often made only after the detection of complications such as cholelithiasis, exacerbations of the anemia induced by infection, or the findings of splenomegaly and growth failure^ref (Styles LA. Hemoglobin H-Constant Spring disease: an under recognized, severe form of  thalassemia. J Pediatr Hematol/Oncol. 1997;4:69–74). The mean hemoglobin in deletional Hb H is quite variable, but averages 9.5 g/dL. 29 to 50% of patients with deletional Hb H require intermittent transfusion therapy, but the need for chronic transfusion therapy is uncommon.
Therapy : many patients require intermittent transfusions
• b-thalassemias

Epidemiology : expecially in Mediterranean basin (Southern Italy (Sardinia) and Po delta (Ferrara))
Aetiology : usually nondeletion type mutations
Pathogenesis : aspecific EPO stimulation leads to bone marrow hyperplasia
• heterozygous (thalassemia minor) : increased levels of HbA₂ (5-7%), which is a result of increased binding of a-globin to d-globin, decreased levels of HbA (93-95%), and normal levels of HbF (< 2%)
• homozygous (thalassemia maior or Cooley disease) : increase in [HbF]
Symptoms & signs : mongoloid facies, "brush skull" at Rx

• b⁰-thalassemia(no expression of b-globin)
• homozygous : b-thalassemia major or Cooley disease
• associated with defects causing increased synthesis of g chains : intermediate b-thalassemia (Hb 6-9 g/dl)
• heterozygous : b-thalassemia minor or minima (Hb > 9 g/dl)
• b⁺-thalassemia(lower-than-normal expression of b-globin)
• black type
• homozygous : intermediate b-thalassemia (Hb 6-9 g/dl; 30÷50% HbA, 50÷70% HbF)
• heterozygous : b-thalassemia minor or minima (Hb > 9 g/dl; HbA₂ increase)
• Mediterranean type
• homozygous : b-thalassemia major or Cooley disease (10÷20% HbA)
• heterozygous : b-thalassemia minor or minima or Rietti-Greppi-Micheli syndrome (Hb > 9 g/dl)
• b-thalassemia with normal HbA₂
• type 1
• homozygous : intermediate b-thalassemia (Hb 6-9 g/dl)
• heterozygous : b-thalassemia minima (Hb > 9 g/dl; normal HbA₂)
• type 2
• homozygous : unknown
• heterozygous : thalassemia minor or Rietti-Greppi-Micheli syndrome (Hb > 9 g/dl; normal HbA₂)
From a world health point of view, ß thalassemia is the most important disorder of globin synthesis to consider, although some patients with HbH disease also have transfusion-dependent anemias^ref. At present, approximately 200 different molecular defects which downregulate ß globin expression have been characterized (Huisman THJ, Carver MFH, Baysal E. A Syllabus of Thalassemia Mutations. Augusta, GA, USA: The Sickle Cell Anemia Foundation; 1997). Homozygotes and compound heterozygotes for these alleles have variable clinical phenotypes ranging from severe transfusion-dependent anemia to mild asymptomatic hypochromic microcytic anemia. In all forms of ß thalassemia there is a reduction in the synthesis of ß globin chains which results in underproduction of HbA. However, the most deleterious effect on erythropoiesis occurs as a result of the precipitation of excess, unpaired  globin chains in erythroblasts (causing dyserythropoiesis) and mature red cells (causing hemolysis)^ref. To a large extent, the clinical and hematologic outcome of the interaction of any 2 ß globin alleles depends on the degree to which ß globin expression is downregulated; this in turn can be predicted from precise knowledge of the molecular defects. However, this oversimplifies the problem. It is well established, and commonly observed, that patients with identical ß globin genotypes may have radically different clinical phenotypes, ranging from severe transfusion-dependent ß thalassemia major to mild forms of thalassemia intermedia. This clinical heterogeneity has been the subject of intense study over the past 20 years. Although still not fully understood, 2 important genetic modifiers of phenotype have emerged, both of which influence the levels of free  chains in erythroid cells (see above). Firstly, inheritance of cis- or trans-acting mutations which increase the ability of adult erythroblasts to synthesize  globin chains, which can combine with excess  chains to form HbF, can significantly ameliorate the clinical and hematologic features of ß thalassemia. Similarly, many studies have shown that co-inheritance of  thalassemia can also reduce the pool of free  chains and thereby influence the clinical severity of ß thalassemia^{ref1, ref2, ref3}. The primary importance of free  chains in causing the phenotype of ß thalassemia has been underscored by many reports of patients who inherit just one abnormal ß globin allele (ßt/ß) but co-inherit 5 (/), six (/) or more  globin genes. Rather than having the phenotype of ß thalassemia trait, such patients often have ß thalassemia intermedia^{ref1, ref2}. These key observations on natural variants which increase  globin synthesis or decrease  globin synthesis, and act as important modifiers of ß thalassemia, highlight 2 pathways that could be manipulated with therapeutic agents to redress the globin chain imbalance underlying ß thalassemia. Although not all patients with ß thalassemia would necessarily benefit, it seems likely that by both increasing  and decreasing  globin synthesis the spectrum of ß thalassemia could be shifted and the clinical outcome improved for substantial numbers of patients. To date nearly all scientific investigation and therapeutic intervention has been directed toward increasing  globin synthesis and relatively little attention has been paid to understanding how one might reduce  globin expression. Clearly this requires a detailed knowledge of how the  globin genes are normally regulated. Understanding in detail how the a globin genes are regulated addresses this important clinical issue and, in a complementary way, contributes to a more general question in human molecular and cellular biology. In the post-genome era there is a major interest in understanding how, during development and differentiation, specific programs of expression emerge from the ~25,000 genes that are potentially available. While global approaches are being applied to this problem, a complementary approach is to understand how a single gene, such as a globin, is activated or repressed as a pluripotent stem cell undergoes lineage specification and differentiation, for example during hematopoiesis. Defining the chromosomal segment containing all of the cis-acting sequences required for fully regulated expression of the  globin genes
The human a globin cluster lies very close (~150 kb) to the telomere of the short arm of chromosome 16, in a GC-rich, Alu dense, gene-dense segment of DNA^ref. In fact the a cluster is surrounded by widely expressed genes and its upstream regulatory elements (including HS-40) lie in the introns of an adjacent housekeeping gene (C16orf35). To determine how much of the chromosome spanning the a cluster might be required to fully regulate expression we have taken an evolutionary approach^ref. It seemed likely that the minimal segment of the chromosome required for full regulation of a globin expression should have been maintained throughout evolution. Such a segment might be defined by comparing multiple species and identifying the smallest region of conserved synteny. This approach has identified an 135–155 kb segment of DNA including the  genes and all of its known regulatory elements. This region of conserved synteny has been maintained throughout over 500 million years of evolution^ref.

The telomere is shown as a black oval. Genes transcribed toward the telomere are located below the line and genes transcribed away from the telomere are above the line. Globin genes are annotated and HS-40 shown and annotated as a box above the line. The extent of the  globin PAC is shown^ref. And the extent of the domain of acetylation is indicated. A box representing the region of conserved synteny is shown and highly conserved non-coding sequences within this region are shown as small black boxes. Common deletions causing a thalassemia remove one (-a), both (--) of the normal a genes (aa) (Weatherall DJ, Clegg JB. The Thalassaemia Syndromes (ed 4). Oxford: Blackwell Science; 2001; Stamatoyannopoulos G, Nienhuis AW, Leder P, Majerus PW. The Molecular Basis of Blood Diseases. Philadelphia: W.B. Saunders Company; 1987; Steinberg MH, Forget BG, Higgs DR, Nagel RL. Disorders of Hemoglobin. Cambridge: Cambridge University Press; 2001) The deletions spanning the upstream regulatory elements^ref are shown as black lines and the shortest region of overlap (SRO) between these deletions is shown.
To identify critical cis-acting elements within this region we have similarly taken an evolutionary approach. Comparing the 135–155 kb segment of conserved synteny in 23 diverse species and using multiple sequence alignments we have identified several small (~100 bp) non-coding regions of DNA that potentially represent conserved cis-acting regulatory sequences. These correspond to specific regions of chromatin which are hypersensitive to DNase 1 in erythroid cells (so-called DNAse 1 hypersensitive sites, DHS). From studies on many mammalian genes, such elements are known to correspond to protein/DNA complexes associated with regulatory elements (e.g., enhancers and promoters)^ref. Close inspection of these non-coding sequences shows that they contain highly conserved binding sites for previously described transcription factors (e.g., GATA-1, NF-E2, SCL) that are known to play a key role in erythropoiesis^ref. In vitro and in vivo protein footprinting, together with more recent in vivo chromatin immunoprecipitation assays demonstrate that these small conserved regions are highly specific binding sites for hematopoietic protein complexes^ref. The potential role of each of the conserved non-coding cis-acting elements, on their own and in combination, has been tested when stably integrated into erythroid cell lines and transgenic mice^{ref1, ref2, ref3, ref4, ref5}. In the human cluster the only active elements in these experimental assays are the a-like globin promoters and HS-40, a regulatory element lying 40 kb upstream of the  globin genes. The individual roles of the other conserved elements remain to be determined. The most helpful guide to the potential role of these elements in vivo comes from the analysis of patients with  thalassemia who have deletions of these elements^ref. a Thalassemia most frequently results from deletion of the structural genes but deletions affecting the  globin promoter have recently been described highlighting the critical importance of this region in vivo. A second group of rare deletions remove ~50–150 kb of the upstream region lying between the  genes and the 16p telomere, where all the remaining conserved non-coding sequences are found. All of these deletions remove three of the conserved elements (HS-48, HS-40 and HS-33) but leave the  genes themselves intact. Although small constructs containing various combinations of these conserved upstream elements linked to the -like genes direct tissue- and developmental stage-specific expression, none of them directs the high levels of expression seen from the endogenous  globin genes in their natural chromosomal environment. However, more recent experiments^{ref1, ref2} suggest that consistently higher levels of regulated expression may be seen in transgenic mice containing large fragments of DNA (~110–150 kb) including most if not all of the region of conserved synteny. In fact, such transgenes are able to fully rescue mice with no endogenous  globin expression^ref (Sharpe JA. Regulation of the human  globin gene cluster in transgenic mice. Weatherall Institute of Molecular Medicine. Oxford: The Open University; 2001).42,43 This suggests that the critical cis-elements controlling globin gene expression may depend on their being within the correct chromatin environment to be fully functional, and this may only be recreated in the context of the large region of conserved synteny contained, for example, in a P1 artificial chromosome (PAC). The potential components of this chromatin environment are considered in the next section.
Although gene expression may ultimately depend on the binding of transcriptional activators, which increase the rate of initiation or elongation of transcription, regulation of this process is greatly influenced by the way in which genes and their regulatory elements are packaged into chromatin. Chromatin packaging is complex and is associated with and/or influenced by a variety of interdependent nuclear processes including the sublocalization of the genes in the nucleus, the timing and replication in the cell cycle, histone modifications and DNA methylation. In general, silent chromatin co-localizes with heterochromatin and replicates late in the cell cycle; the associated chromatin is not acetylated and DNA is methylated. The opposite is generally true for active, accessible chromatin. Many hundreds of proteins have been shown to play a role in establishing and maintaining changes in chromatin accessibility, but 2 classes of proteins are particularly important. One group covalently modifies chromatin, for example by acetylating (e.g., p300/CBP) or de-acetylating (e.g., HDAC1) histones: many other functionally important chromatin modifications (methylation, phosphorylation, ATP-ribosylation, ubiquitination) have also been described (Turner BM. Chromatin and Gene Regulation. Oxford: Blackwell Science; 2001). A second group consists of multiprotein complexes which act as molecular ‘motors’ which can remove or change the position of nucleosomes in chromatin thereby altering the accessibility of key cis-acting regulatory elements^ref. Many of these modifications and chromatin-mediated effects can be stably inherited from one cell to its progeny and they are often referred to as epigenetic modifications^ref. From studies spanning the last 20 years we now have an outline of the epigenetic changes that occur in the  and ß globin clusters during erythropoiesis. As multipotent hemopoietic progenitors become committed to erythropoiesis both the a and ß clusters bind erythroid-enriched transcription factors at their key cis-elements (e.g., a-HS-40 and the ß-LCR) and this is associated with the development of histone modifications which occur with activation (e.g., acetylation) of chromatin and an increase in accessibility to DNA binding proteins, as judged by nuclease sensitivity. By contrast, in non-erythroid cells, the chromatin associated with the  and ß clusters appears very different. Whereas the chromatin associated with the ß cluster has the characteristic features of silent chromatin, the  cluster retains many of the features of active chromatin^ref. Teleologically, this makes sense because the  globin cluster is surrounded by housekeeping genes that must be expressed in both erythroid and non-erythroid cells^ref. The important conclusion from these and other observations, accumulated over many years using indirect assays of chromatin structure and function, is that despite their common evolutionary origin, during the last 500 million years the human a and ß globin clusters have translocated to, and evolved in, quite different chromosomal environments. This in turn suggested that trans-acting mutations involving either transcription factors or proteins which influence the different epigenetic modifications associated with these two clusters might have quite different effects on a and ß globin expression and thereby give rise to the phenotype of a or ß thalassemia. It seemed possible that the identification of such mutations might point to new pathways through which to alter the relative expression of a and ß globin in patients with thalassemia. Differences in the structure and function of the a and ß globin gene clusters :

	a cluster 16p13.3 telomeric	ß cluster 11p15.5 interstitial
GC content	54%	39.5%
CpG islands	common	none
gene density	high	low
Alu family repeats	25%	5%
LINE repeats	rare	present
chromatin	open	closed => open
replication timing	early	late => early
matrix attachment sites	none detected	common
effect of ATRX mutations	present	absent
regulatory element	enhanced	LCR
effect of deletion of major regulatory element on long range chromatin structure	no	yes
predominant mutations	deletions	point mutations
evolution of intergenic regions	rapid	slow
expression in transient assays	enhancer independent	enhancer dependent
expression in hybrids	early	late

Together these findings suggest that the chromosomal regions that host the a and ß globin gene clusters are structurally different in many ways, and that these differences may affect transcription, repair, recombination and replication^ref. Although the vast majority of mutations which cause thalassemia are linked to the a and ß clusters, recently several rare conditions have been described in which a and ß thalassemia have been caused by mutations in trans-acting factors. This was first illustrated in families with ß thalassemia rather than a thalassemia.
• in one family, a mutation affecting the key erythroid transcription factor GATA-1 has been shown to cause thrombocytopenia and ß thalassemia^ref. Subsequently, other families with similar mutations in GATA-1 causing ß thalassemia have been recognized. It has been suggested that differences in the types of GATA-1 binding sites present in the a and ß clusters may explain why this trans-acting mutation has different effects on expression of the 2 clusters^ref.
• the second example of a trans-acting mutation causing ß thalassemia has been characterized in a group of 19 patients with a rare autosomal recessive disorder called trichothiodystrophy (TTD) characterized by dry photosensitive skin, brittle hair, short stature and variable degrees of mental retardation. These individuals inherit mutations in one protein (XPD helicase) of the multicomponent, general transcription factor TFIIH^ref (Viprakasit V. The regulation of a and ß globin gene expression. Weatherall Institute of Molecular Medicine. Oxford: Oxford University; 2002). To date, all of these patients have the hematologic features of ß thalassemia trait. The simplest explanation for this is that the  and ß promoters have different requirements for the activity of TFIIH in initiating transcription during erythropoiesis but the details of this are currently unknown.
• there are several other families with ß thalassemia in whom no mutations have been detected in the genes or their regulatory elements indicating that other trans-acting mutations remain to be found^{ref1, ref2}.
• in addition, it has been shown that there are genetic modifiers of a globin expression which are unlinked to the ß globin cluster. These modifiers have been linked to chromosomes 6, 8 and X but the identity of these genes and their function are unknown^{ref1, ref2} (Garner C, Silver N, Best S, et al. A quantitative trait locus on chromosome 8q influences the switch from fetal to adult hemoglobin. Blood. 2004).
As for ß thalassemia, nearly all mutations causing a thalassemia are cis-acting defects involving the a genes, their promoters or the upstream regulatory elements. Nevertheless over the past 10 years 2 conditions (a-thalassemia/mental retardation syndrome, nondeletional type, X-linked (ATR-X) and a thalassemia myelodysplasia syndrome (ATMDS)) have been characterized in which mutations in a trans-acting factor called ATRX cause the phenotype of  thalassemia. Some years ago we reported the unusual association of a thalassemia (HbH disease) with multiple developmental abnormalities^ref in patients of North European origin. It is now clear that this association may occur as a result of 2 quite distinct mechanisms; one resulting from large deletions involving the telomeric region of chromosome 16 (a thalassemia/mental retardation syndrome, deletion-type on chromosome 16, ATR-16 syndrome); the other, caused by mutations in an unlinked X-encoded protein called ATRX (a-thalassemia/mental retardation syndrome, nondeletional type, X-linked (ATR-X)). The latter condition was immediately recognized as providing a new and potentially important insight into the trans-regulation of globin gene expression. Over 100 families with ATR-X syndrome have now been characterized and the core clinical features include severe psychomotor delay, a characteristic abnormal facial appearance, urogenital abnormalities and a variable degree of a thalassemia^{ref1, ref2}. Although mutations in ATRX nearly always downregulate a globin expression, at present its normal role in regulating gene expression is not clear. ATRX is a large (280 kD) protein that is widely expressed throughout development. It contains 2 highly conserved subdomains. At the N-terminus is an extended PHD-like domain which is most closely related to similar domains found in the de novo methyltransferases (DNMT3a and b). At the C-terminus is a helicase/ATPase domain which provisionally classifies ATRX as a member of the SWI2/Snf2 family of molecular ‘motors’ which use the hydrolysis of ATP as a source of energy^{ref1, ref2}. These proteins are frequently present in multicomponent complexes that remodel chromatin and thereby influence the wide range of the epigenetic nuclear processes described above (e.g., DNA replication, DNA repair, recombination, transcription, DNA methylation)^ref (Turner BM. Chromatin and Gene Regulation. Oxford: Blackwell Science; 2001). The recent isolation of endogenous ATRX has confirmed that it is an ATPase and, like other members of this family, has translocase activity: that is to say, in vitro, the protein can move along double-stranded DNA in the presence of ATP^ref. The functional importance of the PHD-like and helicase domains has been demonstrated by the fact that nearly all inherited ATRX mutations fall within these 2 regions of the protein^ref. Within cells, ATRX is localized in the nucleus and is concentrated in 3 distinct subcompartments: associated with heterochromatin^ref, at rDNA arrays^ref and in PML bodies^ref. ATRX may also be present at more diffuse euchromatic loci, including the a globin complex, and there appears to be some interchange of ATRX between these compartments. Proteins that have been shown to interact with ATRX (e.g., heterochromatin protein 1^{ref1, ref2, ref3} and Daxx^ref) are consistent with these sublocalizations and may also shuttle between the same compartments. Although we have some insight into the function of heterochromatin (a repressive environment), PML bodies (containing proteins involved in regulation of gene expression, apoptosis and the cell cycle) and rDNA (the synthesis of rRNA for protein synthesis), the connection to ATRX has not yet been established. Perhaps the clearest example of a biological function for ATRX concerns its role in DNA methylation. It was of particular interest when it was shown that CpG methylation at heterochromatic loci is perturbed in patients with inherited mutations in ATRX suggesting that ATRX is involved in the establishment and/or maintenance of DNA methylation, an important epigenetic modifier^ref. Despite these advances in understanding of the localization and function of ATRX protein, it is frustrating that none of these insights has revealed how ATRX influences  globin gene expression. The fact that ATRX mutations affect  but apparently not ß globin expression may be informative. These 2 clusters are embedded in quite different chromosomal environments, which supports the view that ATRX influences gene expression via one or more of the epigenetic features that distinguish these different regions. Current research priorities are to determine whether ATRX is detectable at the a globin locus or its upstream regulatory sites and when during erythropoiesis ATRX mutations have their greatest effect. On their own, these observations demonstrating a consistent but relatively modest effect of inherited ATRX mutations on a globin expression are tantalizing but, to some extent, leave open the question of whether ATRX plays a central or peripheral role in the epigenetic regulation of globin gene expression. Showing that acquired mutations in ATRX virtually abolish  globin expression has considerably strengthened and extended the potential importance of the ATRX multiprotein complex in regulating globin gene expression. Although abnormal patterns of hemoglobin synthesis are nearly always inherited, occasionally individuals with previously normal hematology may develop aberrant hemoglobin synthesis as an acquired abnormality, usually within the context of hematologic malignancy^ref. In 1960, 2 groups described a series of previously normal patients suffering from clonal hematopoietic disorders who developed an unusual acquired form of thalassemia during their illness^ref (Bergren WR, Sturgeon PH. Hemoglobin H: some additional findings. New York: Proceedings of the 7th International Congress in Hematology; 1960:488). This syndrome was characterized by a marked hypochromic and microcytic anemia and the presence of HbH (ß₄ tetramers) demonstrable by gel electrophoresis and supravital staining of peripheral red blood cells. Similar patients were soon found to have reduced a/ß globin chain synthesis ratios, demonstrating unequivocally that a thalassemia could occur as an acquired abnormality in the context of hematologic malignancy^{ref1, ref2}. In the light of these reports, a registry of patients with acquired  thalassemia was established in the early 1980s and currently includes 67 verified cases. Acquired  thalassemia is not limited to the geographical regions in which the inherited forms of  thalassemia are common (Mediterranean region, Africa, Middle East, Asia). In fact, most patients have been of Northern European descent. A second, striking finding is that most patients with the acquired form of a thalassemia reported to date have been male. In addition most have been elderly, with a median age at diagnosis of 68 years. As more cases have been evaluated, it has become clear that acquired  thalassemia most commonly develops in the context of myelodysplastic syndrome (MDS) and this condition is now referred to as the a thalassemia myelodysplasia syndrome (ATMDS). Most patients with MDS have an unexplained macrocytic or unremarkable normocytic anemia. Based on hematologic data from the registry, we can now say that ATMDS will be most frequently found in patients with MDS who have the typical abnormal red cell morphology, an MCV < 80 fL and an MCH < 26 pg. To confirm the diagnosis it is still essential to perform an independent assessment for HbH inclusions when a diagnosis of ATMDS is suspected on the criteria set out above. Since many clinical facilities no longer offer supravital staining, laboratory provides this service free of charge for suspected ATMDS patients, if sent 1 mL of fresh blood by courier. Distribution of red cell indices in a thalassemia myelodysplasia syndrome (ATMDS) compared with a more general population with MDS :

Scattergram of MCV and MCH distribution in ATMDS patients (open circles) versus MDS (shaded circles). Eleven of 47 (23%) of ATMDS patients for whom an MCH is available have had an MCH of 27 pg or above, while 15 of 55 (27%) have had an MCV above 80 fL (these indices are denoted by dashed lines). In contrast, almost all MDS patients with an MCH < 27 pg and MCV < 80 fL have evidence of acquired thalassemia. Several studies in the past have tried to determine at what level  globin expression is downregulated in ATMDS (e.g., transcription, mRNA processing translation). The /ß globin mRNA ratio has only been studied in a few ATMDS cases and was reduced to < 0.5 in all patients analyzed^{ref1, ref2}. The a/ß globin chain synthesis ratio is strikingly abnormal (often < 0.2), which is less than one would expect to see in patients with only a single functional  globin gene (--/-a), suggesting that in many ATMDS cases, all four  globin genes are downregulated. The  globin gene cluster itself has been carefully analyzed by Southern blot hybridization and direct sequencing and with one exception was found to be entirely normal, supporting the hypothesis that ATMDS results from a trans-acting mutation. A major clue toward identifying the molecular defect in ATMDS came from a cDNA microarray study of peripheral blood neutrophil RNA from a man newly diagnosed with ATMDS compared with pooled granulocyte RNA from several normal individuals. This demonstrated that the ATRX gene at Xq13.3 was one of the genes with the lowest expression in the ATMDS patient compared to controls, a result that was confirmed by RT-PCR. An ATRX point mutation (G>A) was then found in the canonical splice donor site of intron 1 in this patient, which likely resulted in nonsense-mediated decay of the truncated ATRX transcript^ref. In light of this encouraging finding, archival material including unfractionated blood and marrow from ATMDS registry patients was screened by denaturing high-performance liquid chromatography, which is sensitive to low-level mosaicism, and by sequencing of the 10 kb ATRX mRNA. This revealed 11 additional patients with point mutations and/or splicing abnormalities in ATRX. The mutations clustered in and near the PHD-like and helicase domains which are commonly mutated in the inherited ATR-X syndrome^ref. Schematic representation of ATRX mutations described in boys with ATR-X syndromand in  thalassemia myelodyspasia syndrome (ATMDS) :

The ATRX gene (top, with introns not to scale) is aligned with ATRX protein (bottom) to allow comparison of mutation site with functionally important protein domains such as the ADD (i.e., ATRX, DNMT3, DNMT3L) zinc finger domain (ADD includes a C2-C2 type of zinc finger and the closely located PHD motif), helicase domains, P-box and Q-box. Filled circles represent mutations predicted to cause protein truncation (i.e., frame shifts and nonsense mutations) and null mutations, while unfilled circles represent amino acid changes, including in-frame insertions and deletions. Recurrent mutations in ATRX syndrome are indicated by a number inside the circle, representing the number of families in which the mutation has been identified. These findings have demonstrated that both inherited and acquired mutations in ATRX cause  thalassemia, suggesting that ATRX normally plays a central role in the regulation of  globin gene expression. The widespread developmental abnormalities in ATR-X syndrome also suggest that ATRX plays a more general role in gene regulation. However, at present the mechanism by which ATRX controls gene expression is not clear. Nevertheless, many different lines of evidence suggest that ATRX is a chromatin associated protein which may act as a transcriptional co-factor and/or influence the epigenetic control of gene expression. The discovery of ATRX mutations in ATMDS has also highlighted several other unanswered questions. Solutions to these puzzles may provide important insights into how ATRX normally regulates gene expression and may also uncover some common molecular defects underlying MDS that have more general importance in the area of molecular hematology. It is curious that the hematologic phenotype in ATMDS patients is, in general, much more severe than that seen in boys who have congenital ATR-X syndrome^ref. For example, a boy with a germline 576G>C; L192F ATRX mutation had 0.1% HbH-containing erythrocytes, while an ATMDS patient who had acquired an identical mutation (restricted to myeloid cells) had 50% HbH cells. As described above, patients with ATMDS usually have very low a/ß globin synthesis ratios (< 0.2 in 52% of patients) and large amounts of HbH (median 30%) compared to boys with inherited ATR-X syndrome, who generally have only mildly reduced /ß synthesis ratios with low and sometimes undetectable amounts of HbH. Since the spectrum of ATRX mutations is similar in these congenital and acquired forms of HbH disease, this suggests that the different hematologic effects of ATRX mutations in erythroid cells depends on either when the mutation occurs in development or, more likely, the cellular context in which the mutation occurs. Could there be a relatively common pattern of genetic or epigenetic changes in the erythroblasts of patients with MDS that drastically enhances the deleterious effects of ATRX mutations on a globin expression? The myeloid cells of patients with MDS often have many different genetic mutations, the full repertoire of which is only just beginning to be defined, and recent reports also demonstrate widespread MDS-associated epigenetic abnormalities reflected in the patterns of DNA methylation and histone acetylation^{ref1, ref2, ref3}. Current, incomplete observations suggest that a globin expression per se is not altered in patients with MDS. Therefore, it seems more likely that, in the context of MDS, there are genetic or epigenetic mutations that directly or indirectly exacerbate the deleterious effects of an ATRX mutation. Given the presumed rarity of ATRX mutations, these interacting mutations might occur quite frequently in all types of MDS.
Despite many advances over the past 50 years, the thalassemia syndromes still represent a major worldwide health problem and are still amongst the most common monogenic diseases affecting man. Comprehensive characterization of over ~ 200 molecular defects causing ß thalassemia and ~100 causing a thalassemia have enabled many countries in Australasia, Europe and North America to establish comprehensive screening programs to prevent new cases arising. Although these programs address the recent immigration of individuals from areas where thalassemia commonly occurs, they reach < 1% of the total population at risk. For new cases of severe thalassemia that arise in developed countries current therapy is effective but expensive and burdensome both for the patient and the health care system. New approaches to the management of these common diseases in both affluent and developing nations are urgently required. One approach is to learn from and mimic the naturally occurring genetic interactions which significantly ameliorate the clinical phenotype of ß thalassemia. The best examples of this are those conditions which increase the expression of fetal (g) globin chains or decrease the synthesis of a globin chains. The epigenetic program regulating a globin expression is quite different from that associated with the ß globin cluster.
Prevention : preimplantation genetic diagnosis (PGD)
Therapy^{ref1, ref2} :
• packed red cell transfusion (PRCT) & iron-chelating drugs^{ref1, ref2} : in the last 30 years, conventional treatment of ß-thalassemia major, based primarily on regular blood transfusions and iron chelation therapy with desferrioxamine (DFO), has markedly improved the prognosis of the disease. Adequate administration of parenteral DFO reduces or prevents iron accumulation and iron-mediated organ damage, resulting in a consistent decrease of morbidity and mortality (Borgna-Pignatti C, Rugolotto S, De Stefano P, et al. Survival and disease complications in thalassemia major [abstract]. Cooley’s Anemia-Seventh Symposium. 1998;850:227). However, for several reasons chelation with DFO may be inadequate. In emerging Middle Eastern and Far Eastern countries, where thalassemia is a substantial public health problem, the high costs of the drug and of the supplies for its administration make it unavailable for most patients. Worldwide, it has been estimated that DFO is prescribed for only 25,000 of 72,000 patients with thalassemia major who are regularly transfused (Beutler E, Hoffbrand AV, Cook JD. Iron deficiency and overload. Hematology (Am Soc Hematol Educ Program). 2003:46–52). Despite wide availability of DFO in Western countries, some patients are unable to comply sufficiently with the prescribed treatment because the effective chelation regimen is unpleasant and cumbersome, requiring prolonged daily subcutaneous or intravenous administration. Moreover, some patients may develop side effects of variable severity that affect compliance or sometimes result in cessation of treatment. In a recent report, 105 of 328 patients from North America who had ever received chelation therapy with DFO reported complications requiring modifications of the dose or route of administration of the chelator and 20 reported stopping DFO^ref.3 The inability to cope with the rigorous and demanding long-term use of DFO may result in death from iron overload. The unavailability of DFO for most patients with thalassemia major and the failure of prescribed therapy to prevent complications in other patients have led to a search for alternative iron chelators. Over the last 25 years > 1000 molecules of synthetic, microbiological or plant origin have been tested in vitro and in vivo and several interesting compounds have been identified. One of them, deferiprone (DFP), has been commercially available for 9 years in India and 5 years in Europe, while others are under intensive clinical investigation. Iron accumulates at different rates in various organs, and these organs show a different susceptibility to the damage induced by reactive iron species such as nontransferrin bound iron (NTBI) and the intracellular labile iron pool (LIP). More work is needed to understand the relative accessibility of iron chelators to those different pools. Methods for measurement of these iron species are becoming available and more sensitive, and they could improve the understanding of the mechanism of action of the various chelators and facilitate the design of specific chelation strategies^ref. Based on these findings, targeted chelators that remove iron from specific organs could prove useful. Therefore, besides the overall reduction of iron, effective chelation should aim for a decrease of iron in specific organs, especially the heart, since cardiac disease is the most important life-limiting consequence of iron overload in thalassemia (Borgna-Pignatti C, Rugolotto S, De Stefano P, et al. Survival and disease complications in thalassemia major [abstract]. Cooley’s Anemia-Seventh Symposium. 1998;850:227). During the last 25 years, investigators have made great strides in developing new iron chelators for the treatment of iron overload in thalassemia. Many candidate drugs have been screened, but only a few have had the physicochemical and biological properties suitable for potential clinical application. Some of these are now under intensive investigation. Patients may ultimately benefit from having a choice between several chelators, including orally active drugs. New strategies of chelation, such as combination therapy and organ-targeted chelation, may soon have a considerable impact on the therapeutic outcome and quality of life of patients with thalassemia.

Complications : iron overload : although patients with severe thalassemia have iron overload from repeated blood transfusions, increased iron stores in patients with subclinical thalassemia are due to inappropriately increased gastrointestinal absorption. Athough the pathophysiological process is still under investigation, one explanation involves the molecule hepcidin, a small peptide that controls iron absorption from the gastrointestinal tract. Hepcidin, which increases in proportion to total body iron stores and limits gastrointestinal iron absorption, is inappropriately low in patients with thalassemia^ref (Kearney SL, Nemeth E, Neufeld EJ, et al. Urinary hepcidin in congenital chronic anemias. Pediatr Blood Cancer). Because iron overload is the main cause of complications and death in patients with thalassemia, further assessments of iron stores in the tissue of this patient were conducted to determine whether she should consider iron-chelation therapy.
• clinical presentation : the cardiac complications of thalassemia major were first described prior to the introduction of chelation therapy. Engle described an increase in heart size after 10 years of age with first-degree heart block in a third of the patients. Episodes of pericarditis occurred in half of the patients after 11 years of age. Congestive cardiac failure developed at a mean age of 16 years. As heart failure progressed, atrioventricular (AV) block worsened and either complete heart block or right or left bundle branch block occurred in a third of patients. Atrial arrhythmias were noted in half and repetitive ventricular tachycardia was present in a minority. The duration of life after the onset of heart failure was < 3 months in over half of the patients. Currently, cardiac complications are reported to cause 71% of deaths in patients with thalassemia major. Therefore the heart is the target lethal organ in thalassemia. Once heart failure develops, the outlook is usually poor. The cardiomyopathy may be reversible if iron chelation treatment is intensified in time, but the diagnosis is often delayed by the unpredictability of cardiac iron deposition and the late development of symptoms and echocardiographic abnormalities. The early diagnosis of iron-induced cardiomyopathy with established clinical techniques such as echocardiography and stress radionuclide angiography has had limited success. In particular, overt dysfunction on echocardiography presents late in the disease process and the progression from mildly abnormal echocardiographic parameters to fulminant cardiac failure is often rapid and relentless.
• pathophysiology : iron deposition is reported to be greatest in the ventricular walls with less in the atria and conduction system. There is also greater iron in the epicardial layer. The degree of cardiac dysfunction is considered to depend on the quantity of iron deposited in individual myocardial fibers and the number of fibers affected. In patients with only mild cardiac dysfunction, iron deposits are usually limited to the perinuclear areas, with only a few fibers involved, whereas in patients with significant cardiac dysfunction, iron deposits occupy large areas of the myocardial fibers. The relatively mild degree of fibrosis in most autopsy studies suggests that even when advanced, the cardiomyopathy is potentially reversible. The iron is stored in intracellular lysosomes in the relatively non-toxic forms of ferritin or hemosiderin but above a certain concentration, NTBI is released and the cell begins to fail. The toxicity of iron depends on intracellular iron concentration and its oxidative state. Reduction of ferric to ferrous iron promotes hydroxyl radical formation via the Haber-Weiss reaction.
• iron chelation therapy and the reversibility of iron-induced cardiomyopathy : the iron chelator DFO was first introduced > 30 years ago. DFO reduces hepatic iron and the progression of hepatic fibrosis. Subsequent studies showed improved survival in thalassemia patients on long-term chelation therapy. However, many patients still die of iron overload, and this continued early mortality may be due to several factors: late presentation of cardiac disease, difficulties in assessing myocardial iron, and difficulties in compliance to long-term treatment with daily parenteral iron chelation regimens. The reversibility of iron-induced cardiomyopathy has been demonstrated in patients with hereditary hemochromatosis treated with recurrent venesection and in patients with thalassemia major who are treated with chelation therapy. In vitro, DFO removes iron from cardiac myocytes, reverses lipid peroxidation, and reverses the iron-induced abnormalities of cellular contractility and rhythmicity. However, the clinical reversibility of iron-induced cardiac failure has not been universally accepted, partly due to the high mortality of patients presenting with advanced cardiac failure despite chelation treatment, and more recently by alternative explanations for the heart failure such as myocarditis.
• assessment of tissue iron stores : there is variability in iron deposition between and within different organs. Serum ferritin is the most commonly used indirect estimate of body iron stores, but this reflects only 1% of the total iron storage pool. In addition, interpretation of ferritin values may be complicated by a variety of conditions. Reliance on serum ferritin alone can lead to an inaccurate assessment of body iron stores in individual patients. The hepatic iron concentration by liver biopsy is considered the most accurate and sensitive method for determining the body iron burden. However, this is an invasive technique with a low but recognized complication rate; the result is affected by hepatic fibrosis (especially in small biopsies), which is common in thalassemia as a result of increased liver iron and HCV infection; and iron has been shown to be unevenly distributed in the thalassemic liver even in the noncirrhotic stages. Endomyocardial biopsy can evaluate iron deposition in the heart in thalassemia, but is not a routine part of patient management due to the invasiveness of the technique and the heterogeneous deposition of iron in the heart. In addition, iron has been reported to be absent from the right ventricular subendocardium in some patients with cardiac iron overload. The relation between myocardial iron and other measures such as serum ferritin and liver iron has been unknown until recently, and therefore assessing cardiac risk from these other measures has at best been uncertain. There are noninvasive techniques that can assess tissue iron. Both magnetic resonance (MR) and magnetic susceptometry are significantly affected by iron, and both show changes in iron overload. However, to date only MR can be applied to a moving organ such as the heart. Fortunately there has been substantial progress in the speed and image quality of cardiovascular MR in recent years, and therefore we investigated its use in assessing myocardial iron in thalassemia.
• choice of MR technique for measuring myocardial iron :
• signal intensity ratios : there are reports of the use of signal intensity ratios for the assessment of tissue iron levels, particularly the liver, but rather less so in the heart^ref. These measurements are semi-quantitative because while they rely on alterations in relaxation time caused by iron, they effectively relate the relaxation of one tissue of interest to another such as skeletal muscle, which it is assumed has a low propensity for iron uptake and therefore acts as a reference tissue. Attempts to derive absolute myocardial iron concentration using this approach have relied on tissue calibration curves from liver and the use of an empiric correction factor related to hearts with normal iron levels^ref. Reproducing the results between scanners is very difficult when imaging parameters vary even slightly, and this makes it difficult to compare results between centers. The multiple assumptions and problems associated with signal intensity ratio measurements limit their value and suggest that direct measurement of myocardial relaxation would have significant potential advantages.
• T2 : on first exploring the optimal way to determine myocardial relaxation, we attempted to measure myocardial T2, a measure which in animals has been shown to reflect iron levels^ref. This requires the use of spin echo cardiac magnetic resonance (CMR) techniques, but in our experience with the techniques routinely available in the late 1990s for human in vivo use, these were not sufficiently reliable or reproducible. For example, there was marked motion artifact in noncompliant patients. Some groups have pursued myocardial T2 measurements^ref. Although there are now more advanced T2 imaging techniques, we have focused on assessing myocardial T2* measurement^ref. T2* is measured from gradient echo images, which are fast and easy to acquire, are robust to motion for CMR, and in principle are more sensitive to iron levels than T2. In addition, adjustment of the echo time within the sequence to build up a signal decay curve at multiple echo times that were short enough for moderate or severe iron overload was significantly easier using gradient echo techniques.
• T2* : the principle of using T2* relaxation measurements to determine tissue iron is simple. As the echo time of the images is increased, the signal intensity of all tissues decreases and this rate of decay is strongly enhanced in the presence of iron deposition. Thus, increased myocardial iron reduces myocardial T2*. The key issue in evaluating myocardial T2* was to use echo times that allowed reproducible measurements. Initial experiments showed that echo times from approximately 4 to 20 ms separated by 2 ms increments allowed the best assessment of T2* in mild to severe iron overload. Patients with higher T2* values would have decreased measurement reproducibility, but we had already established that such patients had normal myocardial iron and therefore would not require intensive follow-up to monitor response to a change in treatment. In our first studies, where we were unable to control the sequence repeat time, we chose a low flip angle to minimize T1 effects that can cause errors. The scan resolution was fixed to allow typical breath-hold times per acquisition of 8–10 seconds, which is suitable for children as well as adults. In our early studies, one breath-hold acquisition was required per single image, which had a single echo time. We then acquired up to 9 images, each at a lengthening echo time. A fixed gating delay of 0 ms after the R-wave was chosen in order to obtain myocardial images in a consistent position in the cardiac cycle irrespective of the heart rate. The resulting image is therefore acquired in early systole. An end-systolic image would have offered the benefits of decreased wall motion and thicker ventricular myocardium in which to draw the region of interest, but the delay time would require alteration to end-systole for each patient and also between breath-holds if there was any variation in the patient’s heart rate. A short axis mid-ventricular slice was chosen to minimize artifacts from blood flow. The region of interest was drawn in the ventricular septum, encompassing all layers of the myocardium from the right ventricle through to the left ventricle (LV) endocardium. The region was drawn away from the cardiac veins and the lungs, which cause susceptibility artifacts. More recently, we have developed a single breath hold technique (15 seconds) that acquires all the images at lengthening echo times^ref. This has advantages of speed of acquisition and excellent registration, such that the heart position is constant between images of different echo times, which improves the ease of quantitative analysis. To calculate the myocardial T2*, the signal intensity of the myocardial region of interest was measured in each of the nine images using in-house software (CMRtools, Cardiovascular Imaging Solutions, London, UK). The value was plotted against the echo time for each image. A trendline was fitted to the resulting exponential decay curve, with an equation of the form y = Ke^–TE/T2* where K represents a constant, TE represents the echo time and y represents the image signal intensity.
• normal ranges for T2* relaxation times : to determine normal ranges for T2* values in the heart we scanned 15 healthy volunteers. The normal value for the myocardium was 52 ± 16 ms. There is limited literature with which to compare these results. Previous reports of myocardial T2* values using a variety of echo times include values of 33 ± 6.5 ms, 48 ± 9 ms and 38 ± 6 ms. Importantly, however, the lower limit of normal (defined as the mean minus 2 standard deviations) with our technique was 20 ms, which is very similar to that of the other reports. Therefore, the lower boundary limit we have adopted for myocardial T2* is 20 ms, and values below this represent iron overload. No other clinical scenarios other than iron overload have been shown to cause myocardial T2* below this value.
• calibration of the T2* technique : to investigate the relationship between tissue iron and T2* relaxation, we prospectively studied 30 patients undergoing liver biopsy. The biopsy iron concentrations were compared with the liver T2* measurements. We found a significant, curvilinear, inverse correlation between liver T2* and the liver iron content for all samples (r = 0.81, Figure 2). As predicted from the known variability of biopsy iron determination in the presence of hepatic fibrosis, there was a better correlation with the non-fibrotic liver samples (r = 0.93, P < 0.0001) than the fibrotic samples (r = 0.68, P < 0.0001). Therefore, we subsequently employed nonfibrotic samples to generate predictions of liver iron content from the measured T2* values. Log transformation for the nonparametric data showed a linear relation between liver iron and liver T2*, confirming that tissue T2* can be used as a quantitative measure of tissue iron. These data confirm that tissue T2* is sensitive to tissue iron levels, but they do not provide calibration for myocardium.
• reproducibility of T2* measurements : in order for the T2* sequence to be suitable for the clinical assessment of tissue iron, good reproducibility of the technique is required. Therefore, patients with thalassemia major were scanned on two occasions to assess the interstudy reproducibility of the T2* technique. The coefficient of variation (SD of differences divided by the mean) was found to be 3.3% for the liver and 5.0% for the heart^ref. Reproducibility between scanners was also shown to be very good at 9.4%^ref. To assess intercenter reproducibility, patient scans were performed in 2 different countries and the patients rescanned in a center in London (Westwood MA. Cardiovascular magnetic resonance in myocardial siderosis. MD Thesis. Imperial College, London. 2004). The myocardial reproducibility in each country was 2.4% and 3.5%. Finally, myocardial reproducibility of the single breath-hold technique was 2.3% for abnormal T2* values^ref. All these reproducibility values were considered appropriate for clinical application. The improved reproducibility of these gradient-echo T2* techniques over the fast-spin-echo and spin-echo techniques results principally from improved quality of the original images from which T2* is measured.
• myocardial T2*, cardiac function and normality : LV function was shown to fail in association with myocardial T2* values below the normal range (< 20 ms). As myocardial iron increased, the LV dilated, ejection fraction fell and mass increased^ref. Importantly, no patient was observed to have heart failure who did not have a low myocardial T2* unless there was an obvious alternative cause, such as congenital heart disease. Another notable finding in this cohort study was that in patients with normal myocardial T2* the ejection fraction was usually in the high range for normal subjects. Because of the anemia in these patients, it is possible that to compare their cardiac volumes and other parameters against normal ranges for non-anemic subjects could lead to a significant underestimation of the number of thalassemia patients recognized as having the early stages of LV impairment. We therefore studied a further cohort of thalassemia patients and sought to determine the normal ranges for LV function in those patients in whom myocardial iron loading could be confidently excluded by the presence of a myocardial T2* which was greater than 20 ms (Westwood MA. Cardiovascular magnetic resonance in myocardial siderosis. MD Thesis. Imperial College, London. 2004). This study showed a significant difference for normal LV parameters between the published normal ranges and "normal function" in thalassemia. The latter group had increased LV end-diastolic volume, ejection fraction and mass with reduced end-systolic volume in comparison with normal subjects. These findings confirm that great care must be taken to interpret LV function parameters in thalassemia if significant misdiagnosis of the presence of early cardiomyopathy is not to occur. Further analysis of other parameters of cardiac function has also been related to myocardial T2*. Diastolic function is significantly affected by T2* in the abnormal range of < 20 ms, and there is a progressive increase in the ratio of early to late peak filling rate (Westwood MA. Cardiovascular magnetic resonance in myocardial siderosis. MD Thesis. Imperial College, London. 2004). However, the early detection of cardiomyopathy using diastolic function is still relatively poor when compared with the more direct measure of myocardial T2*. There is also a strong inverse relation between right ventricular function and myocardial T2*, indicating that myocardial iron is responsible for bi-ventricular dysfunction (Anderson LJ. Assessment of cardiac iron overload in thalassaemia. MD Thesis. Imperial College, London. 2002)
• validation of the T2* technique : some have questioned whether the myocardial T2* technique has been adequately validated. In brief, a technique can be validated in relation to established clinical parameters known to reflect a disease process^{ref1, ref2, ref3, ref4}. As described above, there are important relationships between T2* in the range below normal and the remodeling of heart failure, which is the classic response of the heart in cardiomyopathy. There is no doubt, therefore, that since nothing is known to reduce T2* to abnormal levels other than iron loading, iron affects T2*, and T2* is related, in turn, to cardiomyopathy. Therefore, myocardial T2* is validated clinically against cardiac function. What has not yet been achieved is calibration of myocardial T2* against absolute myocardial iron levels. In discussion of whether absolute calibration is vital, the question is whether it is more important to know the relationship between myocardial T2* and cardiac function or between myocardial T2* and directly measured myocardial iron levels. The former is vital, while the latter has additional scientific merit. To date, the myocardial T2* technique can be considered validated but not yet calibrated.
• relation of cardiac iron to liver iron and serum ferritin : no significant correlation could be found between liver and heart T2* in a large cohort (r = 0.15, P = 0.11).5 This lack of correlation has been confirmed by Wood et al^ref. Similarly, no significant correlation was found between heart T2* and serum ferritin level at the time of the scan (r = 0.10, P = 0.32). To confirm that this finding was not due to spurious individual ferritin readings, the mean ferritin for 12 months prior to the scan was also compared to heart T2*, and once again there was no significant correlation (r = 0.09, P = 0.35). Wood found a weak correlation between myocardial T2* and ferritin^ref. This strongly suggests that assessment of myocardial iron from the liver iron is unreliable, and that from ferritin is not ideal.
• relation of cardiac function to liver iron and serum ferritin : in a cohort study of 205 patients, there was no significant difference in left ventricular ejection fraction in patients with liver iron above 15 mg/g (left ventricular ejection fraction [LVEF] 66.1 ± 6.3%) or below 15 mg/g (LVEF 66.2 ± 9.0%, P = 0.96) or with serum ferritin levels above 2500 mg/L (LVEF 65.1 ± 10.1) or below 2500 mg/L (LVEF 66.6 ± 8.2, P = 0.22).9 No significant correlation could be demonstrated between liver iron and LVEF (r = 0.13, P = 0.07) or serum ferritin and LVEF (r = 0.08, P = 0.27) in the cohort of 205 patients. These findings strongly suggest that assessment of the risk of heart failure from either the liver iron or serum ferritin alone is unreliable
• reversibility of cardiac iron in heart failure : a study was performed of 7 thalassemia patients presenting with heart failure who were converted to intravenous DFO treatment, with follow-up over 12 months^ref. One patient died. The survivors showed improved LV function and improved myocardial T2*. This rare prospective study in this condition clearly demonstrates that aggressive iron chelation with DFO is effective in patients with heart failure, and that myocardial T2* improves in concert with parameters of LV function. Of great interest, however, was the clearance rate of iron from the heart, which was substantially slower than from the liver (5.0 ± 3.3% vs 39 ± 23% per month, P = 0.02). Excess myocardial iron persisted in all patients at 1 year, despite liver iron normalization in 4 patients. This suggests that monitoring the duration of treatment by liver iron alone would be misleading in guiding the risk from myocardial iron loading, as the liver would be clear of iron loading very much earlier (by tens of months) than the heart.
• cross-sectional study of cardiac iron between deferiprone and desferrioxamine : a cross-sectional study was performed comparing myocardial T2* and LV function between 2 cohorts of thalassemia patients: one treated with standard DFO therapy, and one treated with DFP^ref. The median myocardial T2* was significantly higher in the DFP group, and their ejection fraction was also significantly better, suggesting less cardiac iron and better ventricular function than in patients treated with DFO. The liver T2* was higher in the DFO group, suggesting more rapid removal of iron from this organ than with DFP. These data require prospective confirmation in randomized controlled trials (these are now underway and will be reported in late 2004 or 2005), but the current data suggest that oral DFP may have preferential access to the heart, in which case it has a potentially unique and valuable role in the treatment of iron overload.
Myocardial T2* measurements have started to unravel some of the complexities of why patients with thalassemia have such a high mortality rate from cardiomyopathy despite clinical management guided by surrogate measures. Myocardial CMR demonstrates the relationship between T2* and LV function. Calibration with myocardial iron is awaited for scientific completeness. Because myocardial T2* is transferable to other scanners and is reproducible, there is considerable potential for widespread application of this technique among the approximately 22,000 MR scanners worldwide.
• allogeneic HSCT in pediatric age is curative and, in the largest reported series, has been carried out in over 800 patients. The proportion of matched sibling donors is a limiting factor and the procedure is associated with an overall mortality of around 20%, the success depending largely on the clinical well being of the patient prior to transplantation^ref. Gaucher-like cells after HSCT^ref. A prompt graft-versus-thalassemia effect upon withdrawal of cyclosporine A in a child who received allogeneic peripheral blood stem cell transplantation^ref
• gene therapy : hemoglobin disorders were among the first diseases for which the gene therapy was envisioned. However, the complexities of the globin gene expression and regulation made this a formidable task. Making gene therapy a therapeutic option in ß-thalassemia requires:
• efficient, safe and high-level gene transfer into target hematopoietic stem cells (HSCs)
• regulated erythroid lineage-specific expression
• therapeutic levels of ß-globin gene expression
A variety of non-viral and viral vector systems have been tried to achieve this goal. While viral components in some vector systems, like the adenovirus and adeno-associated virus, can evoke an immune response to the viral proteins from the host, non-viral methods of gene delivery cause a relatively low immune response and can deliver relatively larger sizes of DNA than can be accommodated by viral vectors. But they are limited by a low rate of gene transfer, a low rate of permanent integration into the host genome and poor long-term expression. In contrast, natural viruses have evolved over millions of years to equip themselves with the ability to enter, transfer and express viral genetic information in the host cells. Retroviruses reverse-transcribe their RNA genome to DNA and integrate permanently into the host genome. Hence, they have been widely exploited as vehicles for gene therapy for diseases requiring life-long correction in the progeny of HSCs. The secondary murine bone marrow transplant model has been used to assess gene transfer into the HSC and the longevity of gene expression. In this model, murine bone marrow stem/progenitor cells are subjected to gene transfer followed by transplant into lethally irradiated syngeneic/congenic mice. Three to six months later, marrow from these primary recipients is transferred into another set of lethally irradiated mice (secondary transplant). Mouse models of disease are very valuable in this regard, such as the ß-thalassemia intermedia mice. Additionally, xenotransplantation of transduced human bone marrow progenitors into highly immune-deficient mice, such as the NOD-SCID and the ß₂-microglobulin null NOD-SCID mice, have been used to assess the therapeutic efficiency of the transferred gene into the progeny of highly primitive human hematopoietic progenitor cells. Although, the short life-span of these immune deficient mice restricts expression in human HSC progeny for 3-5 months, the information obtained from long-term in vivo analysis of human cells is ultimately used to assess therapeutic efficacy and is the most relevant and critical preclinical requisite to gene transfer into humans. Insights into elements required for high-level regulated expression of ß-globin : the ß-globin gene is tailored for a remarkably high-level, erythroid-specific expression. The work of numerous investigators on globin regulation and globin gene therapy has provided important insights into the elements essential for expression in globin genes in vectors:
• high level of globin gene expression requires distal locus control elements: Expression of human ß-globin gene driven by an internal ß-globin promoter within a retroviral vector is very low^ref. A high-level regulated globin gene expression requires presence of a series of distal control DNAse hypersensitive sites (HS) in the 21 kb of DNA upstream from the ß-globin locus, the LCR, that plays a role in maintaining the ß-globin gene locus in an open configuration to allow for high level expression of the globin genes^{ref1, ref2, ref3}
• the LCR is too large to be incorporated into oncoretroviral vectors; therefore, small LCR fragments, consisting of 200–300 bp fragments from the HS, containing only the core sequences have been used. However, inclusion of the core LCR elements into Moloney leukemia virus (MLV)-based oncoretroviral vectors leads to low levels of expression that is prone to position effects^{ref1, ref2, ref3}. Inclusion of larger LCR elements in transgenic mice results in position-independent, copy number-dependent, high-level transgene expression^ref. However, due to the size constraints of MLV vectors, individual HS site cores or a few combinations of these have been tested in MLV vectors with variable, position-dependent expression^{ref1, ref2, ref3}
• oncoretroviral long terminal repeats (LTRs) transcriptionally interfere with LCR elements within MLV vectors, resulting in unstable proviral transmission and/or poor transgene expression^{ref1, ref2}. Transcriptional interference from the LTR^ref can be overcome by self-inactivating (SIN) vectors. In SIN vectors, the LTR promoter/enhancers are deleted upon integration of the provirus. However, SIN MLV vectors have a loss in titers resulting from the LTR deletion. Replacement of the viral LTR with enhancers from other erythroid genes (the GATA-1^ref, or HS-40^{ref1, ref2}, the distal control element from the -globin locus) has met with reasonable success
• ß-globin is an intron-dependent gene that destabilizes oncoretroviral vectors, if placed in antisense orientation: Retroviral vectors are limited to incorporating cDNA, since during production of the viral RNA genome, vector plasmids containing introns would be spliced. This becomes a serious handicap in expressing intron-dependent genes, such as ß-globin. The ß-globin cDNA is very poorly expressed, as ß-globin introns contain elements that enhance transcription, allow proper 3' end processing and the nuclear export of transcripts^{ref1, ref2}. The ß-globin gene is, therefore, inserted in reverse orientation relative to the open reading frame of the viral LTR transcript. This results in vector instability, as several splice sites and polyadenylation signals are introduced into the open reading frame of the vector transcript
• measures taken to overcome instability of globin-containing retroviral vectors:
• deletion of a 372-bp fragment of the intron-2 of the ß-globin gene^{ref1, ref2}, which contains most of the ‘instability elements,’ helps stabilize the viral genomic RNA
• Leboulch et al have mutated numerous deleterious sequences in an MLV vector containing the core LCR elements HS2, 3 and 4 driving expression of ß-globin, showed stable transmission in the majority of mice, but low and variable expression^ref
• improve expression from the ß-globin cDNA, by including viral elements that would replace the functions of the globin introns and show that the woodchuck hepatitis virus post-transcriptional regulatory elements improves expression by 5- to 10-fold (Jiang G, Perelman N, Xu D et al. The woodchuck hepatitis virus post-transcriptional regulatory element greatly increases titers from retroviral vectors and improves expression of the beta globin cDNA. Blood. 1999;94:176a), although levels of ß-globin expression were still subtherapeutic. In summary, efforts to impart stability to the ß-globin mRNA in MLV vectors by modifying various ‘instability elements’ and adding RNA export element vastly improved upon previously reported results, but were still suboptimal to be of therapeutic benefit.
Revival in gene therapy for hemoglobinopathies with lentiviral vectors : LV vectors, derived from the HIV-1 virus^ref have several favorable features for gene therapy:
• can enter a fully intact nuclear membrane and integrate into non-dividing HSC chromatin without prestimulation with cytokines
• are able to package full-length, unspliced RNA due to the presence of a strong RNA export element, the rev response element (RRE). In this regard, May et al have shown that ß-globin gene cassettes that were unstable in MLV vectors were stably transmitted in LV vectors^ref
• can hold a much larger cargo (~9–10 kb) than oncoretroviral MLV vectors, allowing for the larger LCR elements to be incorporated into these vectors. LV vectors, therefore, were able to overcome the hurdles encountered with MLV vectors.
Generation of lentiviral vectors : a 3-plasmid system was initially used to create the first generation LV vectors^ref. It included a packaging plasmid containing the viral gag and pol and accessory genes. A second plasmid encoded an alternative envelope protein, the G glycoprotein of vesicular stomatitis virus (VSV-G), to allow for concentration by ultracentrifugation and broaden host cell infectivity. A third plasmid encoding the expression cassette for the transgene contained all the cis-acting sequences required for encapsidation, reverse transcription and integration. The transgene also contained the CMV internal promoter, intact 5' and 3' LTR, the packaging sequence from the gag gene and RRE. Subsequent generations of LV vectors have attempted to reduce HIV elements and split the packaging gene into different plasmids. Second generation LV vectors improved on biosafety by deleting accessory HIV-1 genes. The third generation vectors deleted HIV regulatory genes, tat and rev, from the packaging construct and expressed the rev protein from a fourth plasmid. Most importantly, a SIN-design of the vector backbone was achieved by deleting the promoter enhancer in the U3 region of the 3' LTR without significant loss in titers or infectivity; during reverse transcription of the viral RNA, this U3 deletion is transferred to the 5' LTR, resulting in transcriptional inactivation of the LTR in integrated proviruses^{ref1, ref2}. Another advantage of SIN-LV vectors was that the level of expression from internal promoters was higher and lineage-specificity was enhanced probably because of the lack of interference from the LTR transcriptional sequences^ref.
Erythroid-specific expression from SIN lentiviral vectors : all current clinical gene therapy trials with MLV vectors have utilized the viral promoter and enhancer to drive expression of the therapeutic gene. This results in unregulated ubiquitous expression of the target gene, which is unnatural and can even be dangerous, as was seen in the X-linked severe combined immune deficiency trial. Moreover, ß-globin gene expression is restricted in the erythroid progeny of HSC. Therefore, it would be important to design vectors that would express in a regulated manner in the desired erythroid lineage. Several SIN-LV vectors have been tested that contain heterologous enhancer/promoters to drive erythroid-specific expression of cDNAs from GFP and human ferrochelatase (FC) or a hybrid human ß/g-globin gene^{ref1, ref2}. Vectors driven by the human ankyrin-1 gene promoter and carrying the HS-40 element from the g-globin locus as an enhancer were found to be erythroid-specific. The cDNA was expressed at high levels and transduced cells showed sustained long-term expression^{ref1, ref2}. The levels of expression of FC from SIN-LV vectors were sufficient to correct the phenotype in a murine model of erythropoietic protoporphyria in secondary mice, suggesting that heterologous promoter enhancers in these vectors are capable of driving expression at clinically relevant levels^ref. Both the GFP-expressing and FC-expressing mice showed an apparent lack of vector silencing. This was unlike the silencing of proviral gene expression, from methylation of the viral LTR, typically observed with MLV vectors. There was no methylation of the human ankyrin-1 promoter in the SIN-LV vectors in secondary mice, suggesting that lentiviral vectors, unlike oncoretroviral vectors, resisted silencing. However, the expression of the vector was highly variable in different mice, or clones derived from cell lines or HSC, since it was dependent on the structure and influences of the surrounding chromatin, where the vectors integrated. This variability in expression of vectors resulting from integration into different chromosomal sites is termed position effects. Therefore, the SIN-LV vectors resisted gene silencing tha