Proper Developmental Control of Human Globin Genes Reproduced by Transgenic Mice Containing a 160-kb BAC Carrying the Human β-Globin Locus

Rhodopsin Genomic Loci DNA Nanoparticles Improve Expression and Rescue of Retinal Degeneration in a Model for Retinitis Pigmentosa

The use of gene therapy may allow replacement of the defective gene. Minigenes, such as cDNAs, are often used. However, these may not express normal physiological genetic profiles due to lack of crucial endogenous regulatory elements. We constructed DNA nanoparticles (NPs) that contain either the mouse or human full-length rhodopsin genomic locus, including endogenous promoters, all introns, and flanking regulatory sequences of the 15-16 kb genomic rhodopsin DNA inserts. We transduced the NPs into primary retinal cell cultures from the rhodopsin knockout (RKO) mouse in vitro and into the RKO mouse in vivo and compared the effects on different functions to plasmid cDNA NP counterparts that were driven by ubiquitous promoters. Our results demonstrate that genomic DNA vectors resulted in long-term high levels of physiological transgene expression over a period of 5 months. In contrast, the cDNA counterparts exhibited low levels of expression with sensitivity to the endoplasmic reticulum (ER) stress mechanism using the same transgene copy number both in vitro and in vivo. This study demonstrates for the first time the transducing of the rhodopsin genomic locus using compacted DNA NPs.

Nanoparticle-mediated rhodopsin cDNA but not intron-containing DNA delivery causes transgene silencing in a rhodopsin knockout model

Previously, we compared the efficacy of nanoparticle (NP)-mediated intron-containing rhodopsin (sgRho) vs. intronless cDNA in ameliorating retinal disease phenotypes in a rhodopsin knockout (RKO) mouse model of retinitis pigmentosa. We showed that NP-mediated sgRho delivery achieved long-term expression and phenotypic improvement in RKO mice, but not NP housing cDNA. However, the protein level of the NP-sgRho construct was only 5-10% of wild-type at 8 mo postinjection. To have a better understanding of the reduced levels of long-term expression of the vectors, in the present study, we evaluated the epigenetic changes of subretinal delivering NP-cDNA vs. NP-sgRho in the RKO mouse eyes. Following the administration, DNA methylation and histone status of specific regions (bacteria plasmid backbone, promoter, rhodopsin gene, and scaffold/matrix attachment region) of the vectors were evaluated at various time points. We documented that epigenetic transgene silencing occurred in vector-mediated gene transfer, which were caused by the plasmid backbone and the cDNA of the transgene, but not the intron-containing transgene. No toxicity or inflammation was found in the treated eyes. Our results suggest that cDNA of the rhodopsin transgene and bacteria backbone interfered with the host defense mechanism of DNA methylation-mediated transgene silencing through heterochromatin-associated modifications.

Sustained expression from DNA vectors

DNA vectors have the potential to become powerful medical tools for treatment of human disease. The human body has, however, developed a range of defensive strategies to detect and silence foreign or misplaced DNA, which is more typically encountered during infection or chromosomal damage. A clinically relevant human gene therapy vector must overcome or avoid these protections whilst delivering sustained levels of therapeutic gene product without compromising the vitality of the recipient host. Many non-viral DNA vectors trigger these defense mechanisms and are subsequently destroyed or rendered silent. Thus, without modification or considered design, the clinical utility of a typical DNA vector is fundamentally limited due to the transient nature of its transgene expression. The development of safe and persistently expressing DNA vectors is a crucial prerequisite for its successful clinical application and subsequently remains, therefore, one of the main strategic tasks of non-viral gene therapy research. In this chapter we will describe our current understanding of the mechanisms that can destroy or silence DNA vectors and discuss strategies, which have been utilized to improve their sustenance and the level and duration of their transgene expression.

Expression of the HSF4 DNA binding domain-EGFP hybrid gene recreates early childhood lamellar cataract in transgenic mice

PURPOSE

The clinical management of cataracts in infancy involves surgical removal of the lens to ensure transmission of light to the retina, which is essential for normal neural development of the infant. This surgery, however, entails a lifelong follow-up and impaired vision. To our knowledge, no animal models recapitulate human lamellar opacities, the most prevalent form of early childhood cataracts. We present data on the recreation of the human lamellar cataract phenotype in transgenic mice.

METHODS

Mutations in the DNA binding domain (DBD) of the heat shock transcription factor 4 (HSF4) are known to be associated with early childhood autosomal dominant lamellar cataract. We used bacterial artificial chromosome (BAC) transgenesis to express a hybrid gene: Hsf4 (DBD)-enhanced green fluorescent protein (EGFP), by recombineering EGFP sequences into the DBD of the Hsf4 gene, to interfere with the DNA binding properties of Hsf4.

RESULTS

We recapitulated the human lamellar cataract, in its temporal as well as spatial presentation, within the transgenic mouse lens. This phenotype was reproduced faithfully using four different BACs, indicating that EGFP can be used to target transcription factor function in transgenic mice. Molecular and cell biological examination of early postnatal transgenic lens reveals impairment of secondary fiber cell differentiation.

CONCLUSIONS

Recreation of the human lamellar cataract phenotype in mice allows investigation of this human pathology at a level not possible previously and points to the relevance of fiber cell heterogeneity dictated by fiber cell-specific gene activity in the biogenesis of the lamellar cataract.

Impact of iron overload and potential benefit from iron chelation in low-risk myelodysplastic syndrome

Myelodysplastic syndromes (MDSs) are a group of heterogeneous clonal bone marrow disorders characterized by ineffective hematopoiesis, peripheral blood cytopenias, and potential for malignant transformation. Lower/intermediate-risk MDSs are associated with longer survival and high red blood cell (RBC) transfusion requirements resulting in secondary iron overload. Recent data suggest that markers of iron overload portend a relatively poor prognosis, and retrospective analysis demonstrates that iron chelation therapy is associated with prolonged survival in transfusion-dependent MDS patients. New data provide concrete evidence of iron's adverse effects on erythroid precursors in vitro and in vivo. Renewed interest in the iron field was heralded by the discovery of hepcidin, the main serum peptide hormone negative regulator of body iron. Evidence from β-thalassemia suggests that regulation of hepcidin by erythropoiesis dominates regulation by iron. Because iron overload develops in some MDS patients who do not require RBC transfusions, the suppressive effect of ineffective erythropoiesis on hepcidin may also play a role in iron overload. We anticipate that additional novel tools for measuring iron overload and a molecular-mechanism-driven description of MDS subtypes will provide a deeper understanding of how iron metabolism and erythropoiesis intersect in MDSs and improve clinical management of this patient population.

Measurement of Epstein-Barr virus DNA load using a novel quantification standard containing two EBV DNA targets and SYBR Green I dye

Background

Reactivation of Epstein-Barr virus (EBV) infection may cause serious, life-threatening complications in immunocompromised individuals. EBV DNA is often detected in EBV-associated disease states, with viral load believed to be a reflection of virus activity. Two separate real-time quantitative polymerase chain reaction (QPCR) assays using SYBR Green I dye and a single quantification standard containing two EBV genes, Epstein-Barr nuclear antigen-1 (EBNA-1) and BamHI fragment H rightward open reading frame-1 (BHRF-1), were developed to detect and measure absolute EBV DNA load in patients with various EBV-associated diseases. EBV DNA loads and viral capsid antigen (VCA) IgG antibody titres were also quantified on a population sample.

Results

EBV DNA was measurable in ethylenediaminetetraacetic acid (EDTA) whole blood, peripheral blood mononuclear cells (PBMCs), plasma and cerebrospinal fluid (CSF) samples. EBV DNA loads were detectable from 8.0 × 10² to 1.3 × 10⁸ copies/ml in post-transplant lymphoproliferative disease (n = 5), 1.5 × 10³ to 2.0 × 10⁵ copies/ml in infectious mononucleosis (n = 7), 7.5 × 10⁴ to 1.1 × 10⁵ copies/ml in EBV-associated haemophagocytic syndrome (n = 1), 2.0 × 10² to 5.6 × 10³ copies/ml in HIV-infected patients (n = 12), and 2.0 × 10² to 9.1 × 10⁴ copies/ml in the population sample (n = 218). EBNA-1 and BHRF-1 DNA were detected in 11.0% and 21.6% of the population sample respectively. There was a modest correlation between VCA IgG antibody titre and BHRF-1 DNA load (rho = 0.13, p = 0.05) but not EBNA-1 DNA load (rho = 0.11, p = 0.11).

Conclusion

Two sensitive and specific real-time PCR assays using SYBR Green I dye and a single quantification standard containing two EBV DNA targets, were developed for the detection and measurement of EBV DNA load in a variety of clinical samples. These assays have application in the investigation of EBV-related illnesses in immunocompromised individuals.

SATB1 regulates beta-like globin genes through matrix related nuclear relocation of the cluster

The nuclear location and relocation of genes play crucial regulatory roles in gene expression. SATB1, a MAR-binding protein, has been found to regulate beta-like globin genes through chromatin remodeling. In this study, we generated K562 cells over-expressing wild-type or nuclear matrix targeting sequences (NMTS)-deficient SATB1 and found that like wild-type SATB1, NMTS-deficient SATB1 induces out loop of beta-globin cluster from its chromosome territory (CT), while it is unable to associate the cluster with the nuclear matrix as wild-type SATB1 does and had no regulatory functions to the beta-globin cluster. Besides, our data showed that the transacting factor occupancies and chromatin modifications at beta-globin cluster were differentially affected by wild-type and NMTS-deficient SATB1. These results indicate that SATB1 regulates beta-like globin genes at the nuclear level interlaced with chromatin and DNA level, and emphasize the nuclear matrix binding activity of SATB1 to its regulatory function.

Advances in high-capacity extrachromosomal vector technology: episomal maintenance, vector delivery, and transgene expression

Recent developments in extrachromosomal vector technology have offered new ways of designing safer, physiologically regulated vectors for gene therapy. Extrachromosomal, or episomal, persistence in the nucleus of transduced cells offers a safer alternative to integrating vectors which have become the subject of safety concerns following serious adverse events in recent clinical trials. Extrachromosomal vectors do not cause physical disruption in the host genome, making these vectors safe and suitable tools for several gene therapy targets, including stem cells. Moreover, the high insert capacity of extrachromosomal vectors allows expression of a therapeutic transgene from the context of its genomic DNA sequence, providing an elegant way to express normal splice variants and achieve physiologically regulated levels of expression. Here, we describe past and recent advances in the development of several different extrachromosomal systems, discuss their retention mechanisms, and evaluate their use as expression vectors to deliver and express genomic DNA loci. We also discuss a variety of delivery systems, viral and nonviral, which have been used to deliver episomal vectors to target cells in vitro and in vivo. Finally, we explore the potential for the delivery and expression of extrachromosomal transgenes in stem cells. The long-term persistence of extrachromosomal vectors combined with the potential for stem cell proliferation and differentiation into a wide range of cell types offers an exciting prospect for therapeutic interventions.

CFTR expression from a BAC carrying the complete human gene and associated regulatory elements

The use of genomic DNA rather than cDNA or mini-gene constructs in gene therapy might be advantageous as these contain intronic and long-range control elements vital for accurate expression. For gene therapy of cystic fibrosis though, no bacterial artificial chromosome (BAC), containing the whole CFTR gene is available. We have used Red homologous recombination to add a to a previously described vector to construct a new BAC vector with a 250.3-kb insert containing the whole coding region of the CFTR gene along with 40.1 kb of DNA 5′ to the gene and 25 kb 3′ to the gene. This includes all the known control elements of the gene. We evaluated expression by RT-PCR in CMT-93 cells and showed that the gene is expressed both from integrated copies of the BAC and also from episomes carrying the oriP/EBNA-1 element. Sequencing of the human CFTR mRNA from one clone showed that the BAC is functional and can generate correctly spliced mRNA in the mouse background. The BAC described here is the only CFTR genomic construct available on a convenient vector that can be readily used for gene expression studies or in vivo studies to test its potential application in gene therapy for cystic fibrosis.

Cluster specific regulation pattern of upstream regulatory elements in human alpha- and beta-globin gene clusters

Located in different chromatin contexts and with different developmental switching mode, human alpha- and beta-globin gene clusters are co-regulated temporally and quantitatively to keep balanced expression. Here, by exchanging their key upstream regulatory elements (UREs) in cluster level, and investigating the expression level of exogenous globin genes in the bacterial artificial chromosome (BAC) mediated transgenic mice, we explored the similarities and differences in the regulatory effects between alpha-upstream regulatory element (alpha-URE) and beta-locus control region (beta-LCR). The results showed that, after exchange, the developmental switching modes of human alpha- and beta-like globin genes had changed, with lost expression of epsilon- and alpha1-genes. Their expression levels also decreased. Our study suggests that the regulation of alpha-URE and beta-LCR on the expression level and developmental switching mode of downstream globin genes is cluster specific.

Gene order in human alpha-globin locus is required for their temporal specific expressions

The human alpha-globin cluster represents a unique model of transcriptional regulation and provides challenges to the current understanding of interactions between distal and proximal regulatory elements. Although the gene proximal regions are believed to possess almost all the necessary elements for temporal and spatial specificity of gene transcription, it is still not clear whether the relative distance of embryonic zeta- and fetal/adult alpha-genes to their distal regulatory element alpha-URE plays any role in transcriptional switching. To investigate the role of gene order in regulating temporal expression, we inverted the entire structure gene region of human alpha-globin locus in a BAC clone bringing alpha-genes closest to alpha-URE and zeta-gene the farthest away. Expression analysis of the reverted locus in transgenic mice showed that alpha-globin genes, now relocated closer to alpha-URE, maintained their expression levels through all developmental stages. However, the zeta-globin gene suffered a total loss at both embryonic and fetal/adult stages. It indicates that proximal location of zeta-globin gene to alpha-URE is necessary for its normal embryonic expression and necessary to prevent embryonic expression of the alpha-globin gene. We proved that, in the human alpha-globin gene cluster, the normal order of structural genes relative to alpha-URE plays a crucial role in the regulation of developmental switching.

Site-specific transfer of an intact β-globin gene cluster through a new targeting vector

The ideal gene-therapy vector for treating genetic disorders should deliver intact therapeutic genes and their essential regulatory elements into the specific "safe genomic site" and realize long-term, self-regulatory expression. For beta-thalassemia gene therapy, viral vectors have been broadly used, but the accompanying insertional mutation and immunogenicity remain problematic. Hence, we aimed to develop new non-viral vectors that are efficient and safe in treating diseases. As previous studies have demonstrated that physiological expression of beta-globin genes requires both a 5' locus control region and 3' specific elements, we constructed a new human chromosome-derived targeting vector to transfer the intact beta-globin gene cluster into K562 cells. The whole beta-globin gene cluster was precisely integrated into the target site and expressed in a self-regulatory pattern. The results proved that the human chromosome-derived vector was specifically targeted to the human genome and this could provide a novel platform for further gene therapy research.

A general method to modify BACs to generate large recombinant DNA fragments

Bacterial artificial chromosome (BAC) has the capacity to clone DNA fragments in excess of 300 kb. It also has the considerable advantages of stable propagation and ease of purification. These features make BAC suitable in genetic research, such as library construction, transgenic mice production, and gene targeting constructs. Homologous recombination in Escherichia coli, a process named recombineering, has made the modification of BACs easy and reliable. We report here a modified recombineering method that can efficiently mediate the fusion of large DNA fragments from two or more different BACs. With the introduction of kanamycin-resistant gene and proposed rare-cutting restriction endonuclease (RCRE) sites into two BACs, a 82.6-kb DNA fragment containing the inverted human alpha-globin genes (theta, alpha1, alpha2, and zeta) from BAC191K2 and the locus control region (LCR) of human beta-globin gene locus (from the BAC186D7) was reconstructed. This approach for combining different BAC DNA fragments should facilitate many kinds of genomic experiments.

Humanized beta-thalassemia mouse model containing the common IVSI-110 splicing mutation

Splicing mutations are common causes of beta-thalassemia. Some splicing mutations permit normal splicing as well as aberrant splicing, which can give a reduced level of normal beta-globin synthesis causing mild disease (thalassemia intermedia). For other mutations, normal splicing is reduced to low levels, and patients are transfusion-dependent when homozygous for the disease. The development of therapies for beta-thalassemia will require suitable mouse models for preclinical studies. In this study, we report the generation of a humanized mouse model carrying the common IVSI-110 splicing mutation on a BAC including the human beta-globin ((hu)beta-globin) locus. We examined heterozygous murine beta-globin knock-out mice ((mu)beta(th-3/+)) carrying either the IVSI-110 or the normal (hu)beta-globin locus. Our results show a 90% decrease in (hu)beta-globin chain synthesis in the IVSI-110 mouse model compared with the mouse model carrying the normal (hu)beta-globin locus. This notable difference is attributed to aberrant splicing. The humanized IVSI-110 mouse model accurately recapitulates the splicing defect found in comparable beta-thalassemia patients. This mouse model is available as a platform for testing strategies for the restoration of normal splicing.

The binding of the ubiquitous transcription factor Sp1 at the locus control region represses the expression of beta-like globin genes

To investigate the function of transcription factor Sp1 in beta-like globin gene activation, we analyzed the recruitment of Sp1, fetal Krüppel-like factor 2 (FKLF2), and related factors at the human beta-globin locus in a human fetal liver and mouse erythroleukemia hybrid cell (A181gamma cell) that contains a single copy of human chromosome 11. Sp1 binds at the GT boxes of the cis-elements throughout the beta-locus, but it is phosphorylated and lost over DNase I hypersensitive site (HS)2, HS3, HS4, and the human beta-globin gene promoter after A181gamma cell differentiation. The binding of FKLF2 at HS2 and HS3 was unchanged. Histone deacetylase 1, which could be recruited by Sp1, is also lost over HS2 and HS3 after differentiation, resulting in the acetylation of histones 3 and 4 across the human beta-globin locus. We previously detected in vivo GT footprints over the beta-globin locus after A181gamma differentiation. Here, we report that after differentiation, the p300/CREB-binding protein-associated factor is recruited by FKLF2 to the locus control region to acetylate histones 3 and 4 at the human beta-globin gene locus. Our results suggest that Sp1 is an inhibitor of beta-like globin gene transcription during erythroid terminal differentiation. Its phosphorylation and release allow the erythroid-specific FKLF2 or erythroid Krüppel-like factor to interact with other erythroid-specific transcription factors to initiate the transcription of beta-like globin genes.

Adult stage gamma-globin silencing is mediated by a promoter direct repeat element

The human beta-like globin genes (5'-epsilon-Ggamma-Agamma-delta-beta-3') are temporally expressed in sequential order from the 5' to 3' end of the locus, but the nonadult epsilon- and gamma-globin genes are autonomously silenced in adult erythroid cells. Two cis elements have been proposed to regulate definitive erythroid gamma-globin repression: the DR (direct repeat) and CCTTG elements. Since these two elements partially overlap, and since a well-characterized HPFH point mutation maps to an overlapping nucleotide, it is not clear if both or only one of the two participate in gamma-globin silencing. To evaluate the contribution of these hypothetical silencers to gamma-globin regulation, we generated point mutations that individually disrupted either the single DR or all four CCTTG elements. These two were separately incorporated into human beta-globin yeast artificial chromosomes, which were then used to generate gamma-globin mutant transgenic mice. While DR element mutation led to a dramatic increase in Agamma-globin expression only during definitive erythropoiesis, the CCTTG mutation did not affect adult stage transcription. These results demonstrate that the DR sequence element autonomously mediates definitive stage-specific gamma-globin gene silencing.

Transgene copy number-dependent rescue of murine beta-globin knockout mice carrying a 183 kb human beta-globin BAC genomic fragment

We report the generation and characterisation of the first transgenic mice exclusively expressing normal human beta-globin ((hu)beta-globin) from a 183 kb genomic fragment. Four independent lines were generated, each containing 2-6 copies of the (hu)beta-globin locus at a single integration site. Steady state levels of (hu)beta-globin protein were dependent on transgene copy number, but independent of the site of integration. Hemizygosity for the transgene on a heterozygous knockout background ((hu)beta(+/0), (mu)beta(th-3/+)) complemented fully the hematological abnormalities associated with the heterozygous knockout mutation in all four lines. Importantly, the rescue of the embryonic lethal phenotype that is characteristic of homozygosity for the knockout mutation was also demonstrated in two transgenic lines that were homozygous for two copies of the (hu)beta-globin locus, and in one transgenic line, which was hemizygous for six copies of the (hu)beta-globin locus. Our results illustrate the importance of transgene copy number determination and of the hemizygosity/homozygosity status in phenotypic complementation studies of transgenic mice containing large heterologous transgenes. Transgenic mouse colonies with 100% (hu)beta-globin production from the intact (hu)beta-globin locus have been established and will be invaluable in comparative and gene therapy studies with mouse models containing specific beta-thalassemia mutations in the (hu)beta-globin locus.

No existence of translocus balancer to coordinate the expression and regulation of human hemoglobin genes in transgenic mice study

All mammals use hemoglobin (Hb) to transport oxygen. Each Hb molecule is a tetramer of two pairs of unlike globin polypeptide chains. Equal amount of subunit globin chains derived from the corresponding alpha- and beta-like genes can always result during development though the two separate gene clusters are located on two different chromosomes and spatially transcribed within different nuclear domains. Disturbance of this balance will result in degradation or precipitation of the excessive globin chains, which is the character of various thalassemic syndromes. In previous studies, we had established two kinds of bacterial artificial chromosome (BAC) mediated transgenic mouse models, which contain respectively the entire human alpha- and beta-globin cluster. Here, we investigated the regulatory relationship between the two clusters by interbreeding these two kinds of transgenic mice. The levels of human alpha- and beta-mRNA in the various hybrid lines reflect the levels in the original transgenic lines that contain either the alpha- or beta-globin cluster alone. The results suggested that there is no apparent cross talk or regulatory interaction between the two human globin clusters in transgenic mice.

Functional analysis of bacterial artificial chromosomes in mammalian cells: mouse Cdc6 is associated with the mitotic spindle apparatus

Bacterial artificial chromosomes (BACs) provide a well-characterized resource for studying the functional organization of genes and other large chromosomal domains. To facilitate functional studies in cell cultures, we have developed a simple approach for generating stable cell lines with variable copy numbers of any BAC. Here we describe hamster cell lines with BAC transgenes that express mouse Cdc6 at levels that correlate with BAC copy number; show that mouse Cdc6 is regulated normally during the cell cycle, binds chromatin, and is degraded during apoptosis; and report a novel fraction of Cdc6 that associates with the spindle apparatus during mitosis. With RNA interference to assess genetic complementation by BAC alleles, this system will facilitate functional studies on large chromosomal domains at variable copy number in mammalian cell models.

Genomic domains and regulatory elements operating at the domain level

The sequencing of the complete genomes of several organisms, including humans, has so far not contributed much to our understanding of the mechanisms regulating gene expression in the course of realization of developmental programs. In this so-called "postgenomic" era, we still do not understand how (if at all) the long-range organization of the genome is related to its function. The domain hypothesis of the eukaryotic genome organization postulates that the genome is subdivided into a number of semiindependent functional units (domains) that may include one or several functionally related genes, with these domains having well-defined borders, and operate under the control of special (domain-level) regulatory systems. This hypothesis was extensively discussed in the literature over the past 15 years. Yet it is still unclear whether the hypothesis is valid or not. There is evidence both supporting and questioning this hypothesis. The most conclusive data supporting the domain hypothesis come from studies of avian and mammalian beta-globin domains. In this review we will critically discuss the present state of the studies on these and other genomic domains, paying special attention to the domain-level regulatory systems known as locus control regions (LCRs). Based on this discussion, we will try to reevaluate the domain hypothesis of the organization of the eukaryotic genome.

Evaluation of optimal expression cassette in retrovirus vector for beta-thalassemia gene therapy

Trials of retroviral vector-mediated human beta-globin gene transfer were hampered by low titers, unstable vector transmission, and low-level expression of transferred gene. With the goal of optimizing the retrovirally encoded human beta-globin gene expression cassette for gene therapy of beta-thalassemia, we generated 3 series of vector constructs (a total of 12 constructs) and investigated the effects of the proximal promoter, 3' - enhancer, and derivatives from the beta-locus control region or alpha-major regulatory element on virus titer, vector transmission stability, and gene expression. The virus titers for 9 of the 12 vector constructs ranged between 2.8 x 10(4) cfu/mL and 1.0 x 10(6) cfu/mL. We found that proviral DNA was intact in most G418- resistant murine erythroleukemia (MEL) cell clones for 5 vector constructs, while obvious genetic instability was observed for 4 other vector constructs. MEL cells harboring the intact provirus were induced to differentiate, and human beta-globin gene expression was analyzed with RNase protection assay. The percentage of human beta-globin transcript relative to endogenous murine alpha-globin transcript were 101.8 +/- 64.3% (n = 10), 40.1 +/- 28.7% (n = 4), 31.1 +/- 31.9% (n = 12), 52.4 +/- 11.2% (n = 12), and 53.6 +/- 8.6% (n = 12) for the 5 constructs, respectively, demonstrating the development of optimized retroviral vectors for beta-globin gene therapy with murine erythroid cell lines as a model. Unexpectedly, we also documented that the point mutation 8700(C-->T) in DNase I hypersensitive site 2 (HS2) core fragment might contribute to low-level expression of the human beta-globin gene, based on a comparison of results from transfected and transduced MEL cells and sequence analysis of proviral DNA.

Further understanding of the beta-globin locus regulation at the molecular level: looping or linking models?

The human beta-globin locus is a classic model of the eukaryotic multigene family with tissue- and temporally specific expression. Over the past few years, great advances have been achieved in studies of beta-globin locus regulation. The dominant role of the beta-globin locus control region (LCR) in chromatin opening and developmental switching has been challenged, and elements beyond the LCR have been studied in depth. More recently, the fields of research have been expanded to intergenic transcription, nuclear localization and histone modification. Several models have been proposed to elucidate the regulation mechanism; among them, the looping and linking models are the most prevalent. Different models are the summarization of the observations made at different times and a persuasive model must be based on a systematic understanding of the numerous observations. The objective of this review is to provide an overview of progress in the area of beta-globin regulation and then to discuss models for it.

The regulatory network controlling beta-globin gene switching

The human globin gene cluster, which represents a prototypical eukaryotic multigene locus, has been investigated for more than two decades and is classic model for coordinate control of tissue-specific gene expression. It is well known that globin gene expression is restricted to specific tissues and that globin genes are sequentially switched on during development. What intricate regulatory mechanisms account for tissue-specific transcriptional control of globin gene expression? Previous studies have focused on the interactions of trans-acting factors and cis-acting elements including the locus control region (LCR), which is considered a potent enhancer in globin gene switching. More recent studies have not only focused on the local DNA regulatory elements but also on remodelling of chromatin and transcription at the globin gene cluster within the native genomic context. Moreover, several studies have presented extensive data that address whether the LCR is required to open the chromatin. Although there is increased insight into the regulation of the beta-globin gene switching, many aspects relating to the developmental activation of distinct globin genes remain elusive.

Screening regulatory sequences from bacterial artificial chromosome DNA of alpha- and beta-globin gene clusters

In the forthcoming postgenomic era, identification of regulatory DNA sequences is becoming increasingly important for characterizing DNA-binding proteins and for elucidating the regulatory mechanisms of gene expression. Presently, there lack efficient methods to broadly screen and identify DNA regulatory elements on a large scale. We established herein an efficient strategy to screen regulatory sequences from bacterial artificial chromosome (BAC) DNAs containing human alpha- and beta-globin gene clusters based on polymerase chain reaction and electrophoretic mobility shift assay (EMSA) techniques without purified transcription factors. Twenty-three subclones derived from alpha-BAC DNA by bulk EMSA selection retained the ability to bind nuclear proteins of K562 cells when retested by EMSA. In 19 clones sequenced, 14 are identical to those registered in GenBank and five have one base difference. All of the 24 randomly picked beta-BAC clones showed specific binding with nuclear proteins of K562 cells. In 11 clones sequenced, eight are identical to those registered in GenBank and three have one base difference. This approach could be particularly powerful if combined with other systematic methods for identifying cis-regulatory DNA elements.

Development of sensitive fluorescent assays for embryonic and fetal hemoglobin inducers using the human beta -globin locus in erythropoietic cells

Reactivation of fetal hemoglobin genes has been proposed as a potential therapeutic procedure in patients with beta-thalassemia, sickle cell disease, or other beta-hemoglobinopathies. In vitro model systems based on small plasmid globin gene constructs have previously been used in human and mouse erythroleukemic cell lines to study the molecular mechanisms regulating the expression of the fetal human globin genes and their reactivation by a variety of pharmacologic agents. These studies have led to great insights in globin gene regulation and the identification of a number of potential inducers of fetal hemoglobin. In this study we describe the development of enhanced green fluorescence protein (EGFP) reporter systems based on bacterial artificial chromosomes (EBACs) to monitor the activity of the epsilon-, (G)gamma-, (A)gamma-, delta-, and beta-globin genes in the beta-globin locus. Additionally, we demonstrate that transfection of erythroleukemia cells with our EBACs is greatly enhanced by expression of EBNA1, which also facilitates episomal maintenance of our constructs in human cells. Our studies in human cells have shown physiologically relevant differences in the expression of each of the globin genes and also demonstrate that hemin is a potent inducer of EGFP expression from EGFP-modified epsilon-, (G)gamma-, and (A)gamma-globin constructs. In contrast, the EGFP-modified delta- and beta-globin constructs consistently produced much lower levels of EGFP expression on hemin induction, mirroring the in vivo ontogeny. The EGFP-modified beta-globin eukaryotic BAC (EBAC) vector system can thus be used in erythroleukemia cells to evaluate induction of the epsilon- and gamma-globin genes from the intact human beta-globin locus.

The expression of human alpha -like globin genes in transgenic mice mediated by bacterial artificial chromosome

After screening a bacterial artificial chromosome of human genomic DNA library with human HS-40, zeta-, alpha-, and theta-globin probes, a 110-kb clone bearing the whole human alpha-globin gene cluster was obtained and rare restriction endonuclease mapping was performed. The bacterial artificial chromosome DNA was isolated, and transgenic mice were generated. Three founders were detected from 35 newborn mice. The copy numbers were 1, 2, and 2, and the expression of human alpha-globin genes in various tissues at different developmental stages in the transgenic mice was assayed. The human alpha-globin mRNA can be detected in bone marrow, kidney, liver, brain, but not in muscle, testis, or thymus. The human zeta-globin genes were switched off, and the alpha-globin genes were switched at day 11.5 in mouse embryo, indicating that developmental stage-specific expression of the alpha-like globin genes was properly regulated. The human alpha-globin mRNA ranged between 17-68% of the endogenous mouse alpha-globin, suggesting that the expression of human alpha-globin genes is integration site-dependent in transgenic mice. The ratio of human alpha(2)- and alpha(1)-globin gene expression in adult transgenic mouse is about 2.5:1 similar to the expression in human.