| Abstract |
The human β-globin gene cluster and the mechanisms which control hemoglobin switching constitute an ideal model for the study of gene function and regulation during ontogeny and erythropoietic differentiation. A unique feature of hemoglobin switching in humans, is the availability of many mutants associated with continued HbF production in the adult life. Studies of the resulting mutants called (δβ)ο thalassemia and hereditary persistence of fetal hemoglobin (HPFH), has offered important insights into the understanding of the control of globin gene expression and hemoglobin switching. The purpose of this Thesis, was to test experimentally the validity of two hypotheses which have been formulated to explain the resulting phenotypes of δβ-thalassemia and HPFH (i.e. the juxtaposition of enhancers and/or the deletion of silencers), through the identification of novel cis-acting regulatory elements located within the cluster, which activate γ-gene expression and are possibly involved in the mechanisms of hemoglobin switching. Our studies resulted in the identification of several novel regulatory elements. We identified two enhancers located 3Ά to the breakpoints of the HPFH-3 and HPFH-6 deletions and four silencers two of which are located 3Ά of the Aγ gene and two 5Ά of the δ gene. The first part of the Thesis focuses on the hypothesis, which assumes that the increased levels of γ-gene expression is due to the juxtaposition of an enhancer element from the 3Ά end of the cluster. We started by studying in vivo the role of the bridging fragment of the HPFH-3 deletion which contains the fetal Aγ-gene with about 6.2 kb of juxtaposed sequence normally residing at the 3Ά end of the cluster. Two mice were found to express Aγ mRNA in both embryonic and fetal liver erythroid cells. Since the Αγ-gene is normally expressed only in embryonic cells of transgenic mice, the data suggest that the juxtaposed sequences partially altered the γ-gene developmental specificity, by shifting the Αγ-gene expression beyond the embryonic stage. We subsequently examined whether these juxtaposed sequences contained a transcriptional enhancer. Transient assays were performed in K562 and HeLa cells which showed that a fragment 0.7 kb located immediately 3Ά of the breakpoint enhanced CAT activity 3-fold. The HPFH-3 enhancer element was further tested on its ability to increase transcription from other globin promoters (ε, δ and β) or non-globin promoters (DHFR) in K562 cells. The data indicated that the HPFH-3 enhancer element is restricted mainly to the fetal globin gene promoter. Sequence analysis of the enhancer element revealed a complex array of repeated sequences which represents part of an LTR belonging to the family of human endogenous retroviruses ERV-9. A second copy of an ERV-9 element with a 87% homology and similar enhancer activity, is located within the 5Ά region of the LCR. The same ERV-9 region is present in the DNA lf gibbon, orangutan and gorilla. To further gain insights for novel regulatory elements flanking the HPFH-3 breakpoint region, we extended the sequence analysis for 5,337 bp downstream of the breakpoint, thus completing the analysis of the 6,261 bp region tested in transient assays. Sequence analysis of the novel 5,337 bp region, has identified several potential binding sites for various transcription factors, 30 open reading frames (ORFs) ranging from 102 to 369 bp, and the presence of a novel transposable element exhibiting structural similarities to retroviruses designated as Mammalian apparent LTR-retrotransposons (MaLRs). We than proceeded with the structural and functional analysis of the HPFH-4 and German (Αδγβ)ο thalassemia deletions. The location and the size of these two deletions is comparable with that of the HPFH-3 deletion. A region of 2,255 bp which was sequenced and is located between the HPFH-4 and HPFH-3 deletion breakpoints, was analyzed by functional assays in K562 and HeLa cells using a luciferase reporter gene. None of the fragments tested showed any detectable enhancer activity. Thus, these data documented that the above three comparable deletions, share a common regulatory mechanism via the HPFH-3 enhancer. To further elucidate the mechanisms leading to fetal gene activation by these mutations, we generated somatic cell hybrids for the German (Αδγβ)ο thalassemia deletion. Two of these hybrids expressed γ-globin, distributed in 2-3% of the cells, unlike the HPFH-1 and HPFH-2 x MEL hybrids previously reported by us, where the distribution of the γ-globin was pancellular. The differences in the pattern of γ-gene expression between the two syndromes, is due to the presence of a novel enhancer which we identified 3Ά of the HPFH-2 breakpoint and to the strength of the individual juxtaposed enhancers (HPFH-1 vs. HPFH-3). We next tested the 3Ά juxtaposed sequences of the Dutch (β)ο and Spanish (δβ)ο thalassemia deletions for the presence of an enhancer element. Functional assays were carried out in K562 and HeLa cells. None of the fragments tested showed any detectable enhancer activity excluding the presence of an enhancer element in the juxtaposed sequences. Thus the activation of the γ-gene in these deletions, must be operating through a mechanism other than that of an imported enhancer. Since the 3Ά juxtaposed sequence of the Spanish (δβ)ο thalassemia extends beyond the available deletions, we proceeded with its sequencing. The analysis of the 3,077-bp fragment indicated that this region is composed of several repeat elements belonging to the LINE 1 and Alu families. We then proceeded with the structural and functional analysis of the comparable Chinese and Thai (Aγδβ)ο thalassemias. The juxtaposed region was analyzed by functional CAT assays in K562 cells. All fragments except one, lacked enhancer activity. The single fragment (HPFH-6 enhancer) located downstream of the Thai deletion, enhanced CAT activity by 5-6 fold. This fragment is deleted in the Chinese thalassemia but it is preserved in the Thai deletion. Based on the resulting phenotypes (δβ thalassemia vs. HPFH) we proposed that the Thai (Aγδβ)ο thalassemia should be renamed HPFH-6. The HPFH-6 enhancer element was further tested on its ability to increase transcription from other globin promoters (ε, δ and β) or non-globin promoters (DHFR) in K562 cells. Results indicated that the HPFH-6 enhancer element is restricted mainly to the fetal globin gene promoter. Sequence analysis of the 1,476 bp HPFH-6 enhancer, revealed the presence of multiple consensus binding sites for several transcription factors. Deletion mutagenesis on the original fragment from both the 5Ά and 3Ά ends and in transient assays in K562 cells, suggested a synergistic role of the individual motifs and the putative binding factors for the maximum effect of the enhancer. Our sequence analysis for a total of 3,157 bp, was extended downstream past the Chinese (Aγδβ)ο thalassemia breakpoint. Our analysis indicated that the 3Ά end breakpoint of the HPFH-6 deletion is located 2,809 bp upstream from the 3Ά end breakpoint of the Chinese (Aγδβ)ο thalassemia. Analysis of the HPFH-6 enhancer sequence indicated that it contains an ORF of 230 amino acids with homology to the olfactory receptor (OR) genes. It is of interest that one more OR gene is located within the HPFH-1 enhancer element. Three more OR genes located in the 5Ά LCR region have also been identified, though their function remains to be determined. By employing RT-PCR analysis, we obtained and sequenced several PCR products from various RNA samples form established hematopoietic cell lines (K562, MEL and HL60). These products were identical with the orthologous genomic sequence of the β-cluster, suggesting that this OR gene (OR-6H) is indeed expressed in these hematopoietic cell lines. However, expression of the OR-6H gene was not detected by Northern analysis possibly due to the low expression of the OR gene. We also screened by PCR a unique human olfactory tissue library. No cDNA clone was obtained, indicating that this OR gene is not expressed in this tissue and might represent a subset of OR genes with a specific repertoire for hematopoietic tissues. In the second part of the Thesis, we examined the second hypothesis for the generation of the δβ-thalassemia and HPFH phenotypes, whereby the main cause for the γ-gene activation is the deletion of silencers located between the Αγ and δ gene region. Extended functional CAT and luciferase analysis of the 13.4 kb region revealed the presence of four silencer elements. These elements were found to exhibit an average activity, which was 30-40% of the control vector. Two elements were located 0.4 and 1.6 kb (Enh and F) from the 3Ά end of the Αγ-gene, respectively, while the other two were located 2.3 and 0.7 kb (O and P) 5Ά of the δ-gene, respectively. Our results are in agreement with previous studies in transgenic mice, where the Aγ-gene was autonomously silenced in the adult stage when the construct contained the Enh and F silencers. We then proceeded with identifying the binding factors of the F-silencer. Our analysis indicated that the F-silencer 364 bp was found to contain several binding sites, with varying binding affinity for the ubiquitous transcription factor YY1. We then characterized the functional role of the three strong binding sites (26, 108 and 311), using site-directed mutagenesis to introduce specific substitutions within the YY1 motif. Following transient assays in K562 cells we observed that substitution at site 108 resulted in a loss of the silencers negative effect, suggesting that this site might be an important parameter for its function. In the last study, we further examined the regulatory role of the O and P silencers on the levels of expression of the γ-genes, using comparable deletion mutants such as the Italian HPFH-5 and the US Black (δβ)ο thalassemia. The bridging fragments of both deletions along with various bridging fragments were analyzed by functional assays in K562 and COS-7 cells. Our results indicated that the bridging fragment of the HPFH-5 deletion, which includes a small region of the truncated O silencer fragment, fused with an almost intact fragment of the 3Ά β enhancer was found to increase CAT activity by 4-5 fold. On the contrary, the comparable bridging fragment of the δβ-thalassemia deletion, containing an almost intact O silencer fragment fused with the intact 3Ά β enhancer could not override the negative effect of the O fragment, which was found to have an average CAT activity even lower of that of the γ-CAT reporter gene. To document further whether the presence of the O silencer was the reason for this diminished activity, we performed a systematic deletion mutagenesis of the O silencer fragment of the δβ-thalassemia bridging fragment, which resulted in the reactivation of the γ-CAT reporter gene by 4-fold. These results mimic closely the in vivo situation regarding the levels of γ-globin gene, and support the conclusion that the presence of the O silencer fragment is responsible for the phenotypic difference of the two syndromes. Based on these results we are in the process of analyzing these novel cis-elements in vivo in terms of their effect on γ-globin gene expression. These results are expected to provide new insights into the mechanisms of hemoglobin switching which will help us understand the regulation of these genes, but also they will further facilitate us to devise new strategies for gene therapy of severe hereditary disorders such as β-thalassemia and sickle cell anemia by incorporating the enhancer elements in constructs of retroviral vectors and/or the inactivation of the silencers by homologous recombination.
|
Μελέτη των cis-ρυθμιστικών στοιχείων του συμπλέγματος των γονιδίων της β-σφαιρίνης του ανθρώπου που συμμετέχουν στην αναπτυξιακή ρύθμιση των εμβρυϊκών γ-γονιδίων
