Bcl11a Gene In Foetal Haemoglobin Production Importance Haemoglobinpathies Biology Essay

Published: Last Edited:

This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.

Inherited Haemoglobin disorders (?-globinopathies) were once considered characteristic of the tropics and subtropics, but with the onset of globalisation & sustained migration they are common worldwide. A conservative estimate suggests 300,000 children are born each year with either a sickle cell disorder or a form of thalassaemia (Modell & Darlison, 2008). Genetic defects in haemoglobin form the most common human genetic disorders worldwide. These haemoglobinopathies cause significant morbidity, mortality and healthcare expenditure (Kauf, et al 2009).

Sickle cell anaemia (SS disease) is a result of a single-base pair mutation of adenine to thymine which produces a substitution of valine for glutamine at the sixth codon of the ?-globin chain (?2?2 6glu-Val). In the heterozygous state (sickle cell trait, Hb AS) only one chromosome carries the mutation. Pathogenesis occurs as a result of the Hb S molecules being insoluble in its deoxygenated state. The molecules polymerise with a reduction in flexibility and become ridged taking up the characteristic sickle appearance. The clinical presentation from the Hb S aggregation leads to haemolytic anaemia and clogging of the small vessels. The long term manifestations are broad and eventually affect nearly every organ, details of which are beyond the scope of the essay.

In ?-Thalassaemia pathogenesis arises from an unbalanced production of ? and ? chains. Either no ?-chains are produced (?0), or ?- chain production is severely reduced (?+). This leads to excess ?-globin accumulation and precipitation in early erythroid cells and cause apoptosis and ineffective erythropoiesis (Bank, 2006). In homozygous ?-Thalassaemia (Cooley's anaemia), children present in the first year of life with failure to thrive and recurrent bacterial infections. Eventually extramedullary haemopoiesis leads to hepatosplenomegaly and bone expansion, giving rise to the characteristic thalassaemic facies. In both ?-globinopathies drug and transfusion therapies are effective in reducing morbidity and mortality with varying success. XXXXXXX

During the last decade research into the human globin genes have led to different strategies for the therapy of ?-globinopathies. Whilst the focus of the essay is on regulation of human foetal haemoglobin, other approaches have shown promise.

"A one-time genetics correction has long been conceived as the ultimate method of cure" [for inherited ?-globinopathies] (Perumbeti & Malik, 2010). Current gene therapy technology has resulted in a permanent correction of other monogenic disorders, such as X-linked severe combined immune deficiency (X-linked SCID) (Hacein-Bey-Abina, S. et al. 2003) and chronic granulomatous disease (CGD) (Ott, M.G. et al. 2006). The delivery of gene modified hematopoietic stem cells (HSC) in immunodeficiency diseases produce a phenotypic correction. Issues arise in the fact that an enormous amount of gene expression is necessary to correction of ?-globinopathies, particularly Cooley's anaemia. Conventional retroviral vectors have proven incapable of efficient delivery of large locus control region (LCR) and the ?-globin gene. But HIV-1 based lentivirus vectors (LV) hold the promise of delivering high expression and transducing a high proportion of HSC (Perumbeti & Malik, 2010).

An overview of haemoglobin switching

Over the course of ontogeny; haemoglobin composition changes with a direct effect on the host physiology. There are two "switches", one of which occurs in all mammals and is known as the primitive to definitive haemoglobin switch. "This involves a switch from haemoglobin subunits produced exclusively in the transiently-produced embryonic primitive wave of erythrocytes to the haemoglobin subunits produced in the earliest definitive wave of erythrocytes arising from the foetal liver" (Mcgrath & Palis, 2008).

The second switch is from the intermediary, foetal haemoglobin to adult haemoglobin; occurring in humans and certain primates alone. The ?-globin gene responsible for the expression of the embryonic haemoglobin (HBE1) is superseded in activity by the ?-globin genes. "During primate evolution, the genes encoding the foetal haemoglobin subunits were duplicated, such as there are two foetal haemoglobin genes in humans, HBG1 (A?) and HB2 (G?) which differ only by a single amino acid" (Sankaran, 2010). "The evolutionary development of the human ?-globin genes and the HbF production in primates is relatively recent; mice have no ?-globin genes" (Bank, 2006). This phenotype has most likely remained due to the increased oxygen affinity of HbF over HbA across the placenta (see appendices Fig. 1). Although it should be noted that mammals with gestation times greater than humans do not always have unique foetal-type haemoglobin, as such the relationship is not as linear as expected. The ?-globin genes coding for foetal haemoglobin (HbF ?2?2) is in turn superseded by the activity of the ?-globin genes HbA (?2?2) and (?2?2), the latter of which expresses ~2% of the total haemoglobin. This switch is typically completed by 6 months of age, at which point the "HbF which comprises <5% of the total Hb at 6 months, continues to fall, reaching the adult level of <1% by 2 years of age" (Thein & Menzel, 2009) As both foetal (HbF, ?2?2) and adult Hb (HbA, ?2?2) contain the ?-globin chain; the switch is effectively the replacement of the ?-globin chain for the ?-globin chain. As such the expression of HBG is replaced by expression of HBB (Fig. 2) (Sankaran et al, 2010).

Regulation of haemoglobin switches

"The human globin genes are one of the most extensively characterised in the human genome" (The American society of haematology, 2006). At the ?-globin locus, gene transcription is influenced by a set of complex and potent interactions, most prominent of which is the ?-globin locus control region (?-LCR) (Tuan et al, 1985, Bender et al, 2000). This is found upstream (5 degrees) of the gene cluster and maintains a high level of globin-gene transcription. "The ?-LCR contain 5 DNAase 1 hypersensitive (HS) sites, HS1-5, that are formed in regions devoid of nucleosomes and that are sites more accessible that other regions of chromatin to interactions with transcription factors and downstream gene sequences" (Bank, 2006). The DNAase 1 hypersensitive (HS) sites are clustered over a 10-kb region located about 50-60kb upstream of the ?-globin gene.

The activity of the ?-globin locus is modulated by several transcription factors, binding site for which are found on the ?-LCR, with GATA-1 and Erythroid Krupple-like factor (EKLF or KLF1) considered as the most prominent. Mice deficient in GATA-1 are unable to produce mature erythroid cells (Blobel et al, 1995). Mice deficient of KLF1 die in utero with a disease resembling severe ? thalassaemia (Perkins et al, 1994). Another transcription factor, NF-E2 has been shown to cause more than 100 fold increase in murine ?-globin transcription (Sawado et al, 2001). The ?-LCR also facilitates the establishment of an "open" active chromatin region with the aid of histone acetyl transferase (HAT) enzymes, and ATP-dependant nucleosomes remodelling complexes. Each gene in the ?-gene cluster is activated sequentially; the temporal sequence corresponds with the order on the chromosome. However, an understanding of the development and control of the haemoglobin switches remained elusive.

HbF variation & distribution

The switch from foetal to adult Hb is neither total nor irreversible. Adults retain the ability to produce some HbF, with a continuing slow decrease over adult life (Rutland et al, 1983). There is a huge variation of HbF amongst adults and this is considered to be genetically controlled (Garner et al, 2000) but with the absence of clear Mendelian inheritance patterns. Whilst persistently high levels of HbF have no clinical consequence in healthy adults high HbF levels reduce morbidity and are associated with milder disease progression and fewer complications in patients with SS Disease and ? thalassaemia (Nagel & Platt, 2001). "The principle that elevated HbF ameliorates the severity of the ?-heamaglobin disorders has been the driving force behind efforts to stimulate haemoglobin production in humans" (Sankaran et al, 2010).

HbF is not uniformly distributed in adults; the cells which contain measureable amounts are termed F Cells (FC) (Boyer et al, 1975). The amount of HbF per FC is also variable, and in normal adults range up to 25% of the total Hb in an erythrocytes. Interestingly the distribution of both the HbF trait and the FC trait reveals itself as a normal (Gaussian) distribution (after log transformation) (Thein & Menzel, 2009). This normal distribution suggests that the variability does not originate from a single genetic locus but from a number genes. The distribution of both traits can vary considerably; by more than 20-fold and is continuous and positively skewed (Economou et al, 1991) with twin studies showing that the levels of HbF and FC are genetically controlled with a heritability of 0.89 (Garner et al, 2000). The remainder of FC variance is accounted for by age (2%) and unknown environmental factors (Thein et al, 2009). This goes some way in explaining the non-mendelian inheritance pattern of the common HbF variability.

"Natural mutations at the ?-gene cluster in patients with increased HbF have shed important light on the mechanisms regulating HbF" (Bank, 2006). Hereditary persistence of foetal haemoglobin (HPFH) is a descriptive term for a range of conditions with a genetically determined persistence of HbF production into adult life, in the absence of any haematological disorder (Bollekens & Forget, 1991). Certain HPFH arise from point mutations in the human ? promoter (both the G? and A? genes); these are characterised as non-deletion-type HPFH (Pancellular HPFH). In most cases these point mutations are found between ' 114 and ' 202 upstream of the human ?-globin gene (Forget, 1995). Homozygous deletions of the ?-globin gene are rare but benign with the adult expressing the ?-globin gene into adult life, with 100% HbF and no anaemia (Forget, 1998). Heterocellular HPFH (hHPFH) was once considered another type of HPFH, but was soon discovered to be individuals from the upper tail of the normal distribution of HbF. This was found to be a distinct condition with clear Mendelian inheritance as alleles of the HBB (Thein & Craig, 1998).

Importantly whilst increased levels of HbF ameliorate the effects of certain ?-globinopathies, the reverse is also true. "Genetic disorders such as ? thalassaemia and SCD are also associated with variable increases in HbF levels" (Thein & Menzel, 2009). SS disease carriers (Hb AS) have HbF within the normal range, but sickle cell homozygotes (HbS) have HbF levels ranging from <1% to 30% (Nigel & Platt, 2001). All ? thalassaemias, heterozygous or homozygous have variable increase in HbF (Weatherall & Clegg, 2001). Interestingly "in the majority of ? thalassaemia and SCD-SS, the mutation itself does not directly result in increased HbF production, a large part of the HbF response is related to the erythropoietic stress and expanded mass secondary to the ineffective erythropoiesis or haemolytic process, and preferential survival of the red cell precursors that contain HbF (i.e FC)" (Thein & Menzel, 2009). As the evidence suggests variation in multiple genes, produces high heritability of HbF. Thus in genetic terms, HbF persistence is considered a "quantitative trait" (QT): multiple genes together with a small environment component determine the value measured in any given individual (Garner et al, 2000).

Whilst an understanding of the ?-LCR, deletions and mutations in HPFH and variability in HbF, have gone some way in assisting in an understanding of the variation in globin gene activity, it is quite clear that temporal regulation with haemoglobin switching is a complex multifactorial process.

HbF QTL (Quantitative Trait Loci)

At one point it was unclear as to whether genetic variability at the ?-globin gene played a role beyond providing rare variants with significantly raised HbF. Could this contribute -in any tangible sense- to the common variation in HbF in the general population? The first suggestion that this could be the case was in homozygous individuals with identical HbS mutations, who had huge variability in HbF levels (Platt et al, 1994). Carriers for the ?S gene on the Arab-Indian haplotype had the highest HbF and a mild clinical course (Labie et al, 1985). Whilst individuals with the ?S gene on the Bantu haplotype; had the lowest HbF levels and the most severe clinical course (Gilman & Huisman, 1985). This discovery had led to the identification of a single nucleotide polymorphism (SNP) at position -158 of the HBG2 promoter (Gilman & Huisman, 1985) that creates a restriction site for the enzyme Xmn1-HBG2 on rs7482144. The term "158 Xmn I site" was termed and this SNP had eventually been found to be associated with substantial increases in HbF in anaemic individuals (Lanclos et al, 1991). It has also been shown to be associated with hHPFH in non-anaemic adults. Initial studies had shown that the -158 Xmn I site accounts for a considerable proportion of the variability of HbF and FC, measured genotype analysis had shown the -158 Xmn I site to account for 13-32% of the total FC variance in a non-anaemic north European population (Garner et al, 2000). As well as having a high impact on the variance of FC it also has a huge frequency; ~30% in most population groups (Thein et al, 2009). Thus "-158 Xmn I site" was considered to be the first Quantitative Trait Loci (QTL) underlying HbF heritability.

Once the first QTL had been discovered, the expectation was that further QTLs were responsible for the huge variation in F cells. Certain studies had suggested that high HbF determinants were being segregated independently of the HBB ?-globin gebe in some families with ?-thalassaemia and SS disease (Thein & Weatherall, 1989). These families were found to have a mild form of the disease as a result of the high levels of HbF. A family of Asian-Indian origin were investigated using genome-wide linkage analysis and identified as having a major locus on chromosome 6q23-q24. Further investigation narrowed 1.5mb region to five genes; (ALDH8A1, HBS1L, MYB, AHI1, PDE7B) (Close et al, 2004). At this point linkage analysis was unable to improve the resolution further, the 1.5mb chromosomal region was inherited in one linkage disequilibrium (LD) block in the family. High resolution association mapping helped in overcoming this stumbling block. A sample of north European ancestry discovered a contiguous section within the 6q23 region. A segment, ~79kb within HBSIL, the intergenic region 5 degree to HBSIL and the MYB oncogene showed very strong association with F cell levels (p-value ~10-75) (Thein at al, 2007). HMIP (HBS1L-MYB intergenic polymorphism) refers to an array of single-nucleotide polymorphisms (SNPs) situated in the interval between the gene for HBS1L and the MYB oncogene. HMIP SNPs exist in 3 linkage disequilibrium (LD) blocks, HMIP-1, HMIP-2, and HMIP-3, and the genotype at each block influences the number of F cells and foetal haemoglobin (HbF) levels. HMIP-2 was shown to be the SNP most functionally active. (Menzel et al, 2007)Further studies had shown in Northern Europeans the 6q23 QTL accounts for ~19% of the population trait variance (Thein et al, 2009).

The discovery of BCL11A

After the first two QTLs had been discovered there was a period of stagnation. Whilst there was general consensus that the first two QTLs could not be solely responsible for the variation in HbF; there was no tool at the appropriate resolution capable of identifying the remaining QTLs. By early 2006 there was marked development in the genetic tools and genotyping platforms, this led to a dramatic leap forward in the applicability of genetic association studies, resulting in genome-wide association studies (GWAS). Research published in 2007 (Menzel et al, 2007) used GWAS to identify a further QTL associated with HbF variability. Crucial in their research was the sampling method; they had elected to use a small population on either extreme end of the phenotypic spectrum; >P95 or <P5. This innovative approach identified a further QTL within a 126-kb segment on chromosome 2p15 (nucleosides 60456396 to 6082798) considered close to BCL11A oncogene (see appendices). Interestingly tests of interactions between OTLs were non-significant, suggesting that the contributions of each QTL were additive. Combined they constitute 44% of the total variance in the twin panel (Menzel et al, 2007) -half of the overall 0.89 heritability- (Garner et al, 2000).

The BCL11A gene is expressed in red blood cell precursors and had been implicated in lymphoid development and lymphoma pathogenesis; acting as a transcriptional repressor (Satterwhite et al, 2001, Liu et al, 2003) but had not been thought to have been involved in HbF variability or haemoglobin switching. The gene encodes a zinc finger transcription factor with multiple isoforms that share a common N terminus but differ in the number of C-terminal zinc fingers. Once BCL11A had been unexpectedly identified as a QTL for HbF variation, linkage and association studies had been used to investigate different SNP association with HbF levels across all 3 QTLs.

The results had shown that the SNP with the strongest association with HbF activity was found on the SNP rs11886868 at intron 2 of the BCL11A gene; (Table 1| Uda et al, 2008) with a p < 6.70 X 10-35 and P < 10-9 in the follow up sample. Further analysis of rs11886868 had shown the genotype was markedly different between individuals with normal HbF levels (?0.8%) and those with higher HbF levels (>0.8). The C allele had shown 2-fold frequency enrichment and the C/C allele had shown 5-fold frequency enrichment between a normal population sample and a HPFH population sample (Uda et al, 2008).

Fig 3| shows the association between the rs11886868 genotype and HbF levels (Uda et al, 2008). Strong evidence now suggested an association between the BCL11A gene and HbF levels (Sankaran et al, 2010).

BCL11A activity & variation

Cis-regulatory elements are well established as a causal element for the interspecies difference in gene expression (Porcu S et al, 1997). The influence of trans-acting environments is less clear. Human cis-elements in mouse trans-acting environments have provided much insight into our understanding of the regulation of haemoglobin switches. These ?-globin locus transgenic mice (?-globin mice) proved to be a valid system for evaluating human temporal globin gene regulation but established differences exist. The onset of ?-globin expression in mice occurs at the embryonic, yolk sac stage of erythropoiesis, whereas in humans this expression occurs during the foetal liver stage. In addition the switch from ?-globin to ?-globin expression occurs during early foetal liver erythropoiesis in mice (Porcu S et al, 1997), whereas it occurs during the time of birth in humans. Furthermore the distribution of positively stained "with immunohistochemistry- ?-globin-positive cells in ?-globin mice was different to that of human foetal liver cells (Sankaran et al, 2009). In ?-globin mice the ?-globin expression was limited to megaloblastic primitive cells. This model has shown that in ?-globin mice; ?-globin is expressed in a similar fashion to the mouse embryonic gene ?y-globin (Hbb-y) (Sankaran et al, 2009). But importantly primary transcript RNS fluorescence in situ hybridisation (PT-FISH) indicates that the level of expression of the human ?-globin parallels that of the mouse embryonic ?-like genes in the mouse trans-acting environment. This finding tells us that whilst fundamental differences exist, ?-globin mice are an appropriate model when looking for changes in HbF activity.

BCL11A has shown to be strongly associated with ?-globin activity (in human erythroid cells (Thein & Menzel, 2009, Uda et al, 2008). It has been shown to have robust activity in bones marrow erythroblasts, substantially lower levels in foetal liver cells, and no activity in primitive erythroblasts. This reinforces the suggestion that the BCL11A gene is a developmental-stage-specific repressor of ?-globin activity (Sankaran et al, 2009). The variation in murine and human globin switching presents an opportunity to better understand how BCL11A mediates its effects. ?-globin mice models show that BCL11A protein (Xl/L isoforms) and RNA transcripts are absent in primitive erythroid cells of mice, whereas shorter variants of BCL11A exist in human embryonic cells (stage-matched cells). Secondly, whilst full length BCL11A forms (Xl/L isoforms) are expressed in similar levels in foetal liver cells and bone marrow cells of mice and human samples, shorter variants could not be identified in mice samples (Sankaran et al, 2009). This variation in BCL11A expression suggests an involvement -at least in part- for the interspecies divergent expression of ?-like globin genes. Studies indicate that the BCL11A locus is developmentally regulated, such that full-length XL and L isoforms are expressed almost exclusively in adult-stage erythroblasts (Sankaran, 2008)

Direct investigation of BCL11A activity using BCL11A knockout mice (bcl11a-/-) further supported this hypothesis. During E14.5 and E18.5 in mice control we expect robust definitive erythropoiesis is taking place in the foetal liver. Whereas in the bcl11a-/- mice, whilst phenotypically and morphologically erythropoiesis appeared normal, globin gene expression was altered. Silencing of the embryonic globin gene failed to occur in E14.5 and E18.5 foetal liver erythroid cells. Expression of ?y was upregulated by 70-fold at E14.5 and at E18.5 ?y globin transcription was increased 2,600-fold (Sankaran et al, 2009). When the ?-locus transgene was introduced into the bcl11a-/- developmental silencing of ?-globin activity was markedly impaired (Sankaran et al, 2009). The distribution of haemoglobin followed the expected trend when comparing bcl11a-/- homozygotes and bcl11a+/- heterozygotes.

Fig 4|Expression of human b-globin locus genes is shown for animals with the various Bcl11a genotypes in the presence of the b-locus YAC transgene (YAC1) at E14.5 (n54, 6 and 4 for the Bcl11afl/1, Bcl11a1/2 and Bcl11a2/2 animals, respectively) and E18.5 (n54, 7 and 4). All c- and b-globin levels for the different genotypes are significantly different (P,131025, two-sided t-test) (Sankaran et al, 2009).

BCL11A Interactions

Whilst there is a strong indication as to the nature of BCL11A activity, the process remains unclear. Reporter assays in K562 cells using promoter constructs showed binding of the BCL11A to a GGCCGG motif at position -56 to -51 of the HBG2 proximal promoter to forma repressor complex (Chen et al, 2009). Co-immunoprecipitation assays show that BCL11A in tandem with GATA-1, may be involved in the same repressive nucleosome-remodelling and histone deacetylase (NuRD) complex (Jawaid 2010). This interaction with NuRD was suggested by the interaction of GATA1 with FOG-1, a co-factor shown to associate the NuRD complex (Sankaran 2008). BCL11A/GATA1 binding were confirmed in locations known to be strongly associated with changed in HbF levels. Amongst the ?-globin gene co-localisation was seen ~3.3 kb downstream of HBG1, deletion of which are known to cause HPFH. Co-localisation was also seen ~2 kb upstream of HBD, in an area where deletions are known to cause ??-Thalassaemia. BCL11A showed no binding to its own gene but strong signals were observed in the GATA1 gene (Jawaid 2010). These findings strongly suggest that BCL11A is the first genetically validated regulator of both developmental control of haemoglobin switching and silencing of ?-globin expression in adults.

Concluding remarks