Identification and characterization of transcriptional arrest sites in exon 1 of the human adenosine deaminase gene

Control of Transcriptional Elongation by RNA Polymerase II: A Retrospective

The origins of our current understanding of control of transcription elongation lie in pioneering experiments that mapped RNA polymerase II on viral and cellular genes. These studies first uncovered the surprising excess of polymerase molecules that we now know to be situated at the at the 5' ends of most genes in multicellular organisms. The pileup of pol II near transcription start sites reflects a ubiquitous bottle-neck that limits elongation right at the start of the transcription elongation. Subsequent seminal work identified conserved protein factors that positively and negatively control the flux of polymerase through this bottle-neck, and make a major contribution to control of gene expression.

The tumor-selective over-expression of the human Hsp70 gene is attributed to the aberrant controls at both initiation and elongation levels of transcription

The tumor selective over-expression of the human Hsp70 gene has been well documented in human tumors, linked to the poor prognosis, being refractory to chemo- and radio-therapies as well as the advanced stage of tumorous lesions in particular. However, both the nature and details of aberrations in the control of the Hsp70 expression in tumor remain enigmatic. By comparing various upstream segments of the Hsp70 gene for each's ability to drive the luciferase reporter genes in the context of the tumor cell lines varying in their p53 status and an immortal normal liver cell line, we demonstrated in a great detail the defects in the control mechanisms at the both initiation and elongation levels of transcription being instrumental to the tumor selective profile of its expression. Our data should not only offer new insights into our understanding of the tumor specific over-expression of the human Hsp70 gene, but also paved the way for the rational utilization of the tumor selective mechanism with the Hsp70 at the central stage for targeting the therapeutic gene expression to human tumors.

Regulation of c-fos expression by RNA polymerase elongation competence

The molecular mechanisms underlying transcription elongation and their role in gene regulation are poorly characterized in eukaryotes. A number of genes, however, have been proposed to be regulated at the level of transcription elongation, including c-myc, c-fos and c-myb. Here, we analyze the control of transcription elongation at the mouse c-fos gene at the nucleotide level in intact cells. We find that RNA polymerases are engaged in the promoter-proximal part of the gene in the absence of gene activation signals and mRNA synthesis. Importantly, we determine that the engaged RNA polymerases originate from a continuous initiation of transcription which, in the absence of gene activation signals, terminate close to the promoter. We also observe that the c-fos gene presents an active chromatin conformation, with the promoter and upstream regulatory sequences constitutively occupied by proteins, accounting for the continuous initiation of RNA polymerase complexes. We propose that activation of c-fos gene expression results primarily from the assembly of elongation-competent RNA polymerases that can transcribe the complete gene. Our results suggest that the engaged RNA polymerases found downstream of a number of other eukaryotic promoters may be associated with transcription termination of elongation-incompetent polymerases in the absence of activating signals.

Genomic organization, expression of the human CBFA1 gene, and evidence for an alternative splicing event affecting protein function

The Cbfa1 gene, which encodes the transcription factor Osf2/Cbfa1 required for osteoblast differentiation in mouse and human, is mutated in cleidocranial dysplasia, a skeletal dysplasia. We describe here the isolation of the full-length human OSF2/CBFA1 cDNAs, the genomic organization of the entire CBFA1 gene, its expression, and the existence of an alternative splicing event. Nucleotide sequence analysis of the human and mouse OSF2/CBFA1 cDNAs showed a 98% homology in the coding sequence and 96% in the 5' untranslated (UTR) sequence. Analysis of CBFA1 genomic clones revealed that the 5' UTR sequence of the human OSF2/CBFA1 cDNA lies 75 kb upstream of the originally described 5' end of the gene. The existence of two OSF2/CBFA1 cDNAs is due to an alternative splicing event around exon 8 that affects the transcriptional activity of the protein. Northern blot analysis demonstrates that the expression of the human OSF2/CBFA1 gene is restricted to osteoblastic cells.

Basic mechanisms of transcript elongation and its regulation

Ternary complexes of DNA-dependent RNA polymerase with its DNA template and nascent transcript are central intermediates in transcription. In recent years, several unusual biochemical reactions have been discovered that affect the progression of RNA polymerase in ternary complexes through various transcription units. These reactions can be signaled intrinsically, by nucleic acid sequences and the RNA polymerase, or extrinsically, by protein or other regulatory factors. These factors can affect any of these processes, including promoter proximal and promoter distal pausing in both prokaryotes and eukaryotes, and therefore play a central role in regulation of gene expression. In eukaryotic systems, at least two of these factors appear to be related to cellular transformation and human cancers. New models for the structure of ternary complexes, and for the mechanism by which they move along DNA, provide plausible explanations for novel biochemical reactions that have been observed. These models predict that RNA polymerase moves along DNA without the constant possibility of dissociation and consequent termination. A further prediction of these models is that the polymerase can move in a discontinuous or inchworm-like manner. Many direct predictions of these models have been confirmed. However, one feature of RNA chain elongation not predicted by the model is that the DNA sequence can determine whether the enzyme moves discontinuously or monotonically. In at least two cases, the encounter between the RNA polymerase and a DNA block to elongation appears to specifically induce a discontinuous mode of synthesis. These findings provide important new insights into the RNA chain elongation process and offer the prospect of understanding many significant biological regulatory systems at the molecular level.

Identification of a murine homolog of the human adenosine deaminase thymic enhancer

We have identified a 236-bp first intron segment of the mouse adenosine deaminase gene (ADA) that shares 71.1% identity with the human ADA thymic enhancer. This segment has the same natural orientation as the human enhancer and a relative location within the first intron very analogous to that of the human enhancer. Four highly conserved regions were defined within this segment, including a 72-bp region having 83.6% identity with a segment containing the critical human enhancer core. Several consensus binding sequences were also conserved within these regions. Transient transfection assays in human and murine T-cell lines revealed that a 1.8-kb murine genomic fragment harboring the 236-bp segment functions as a weak activator of both the human and mouse ADA promoters. In contrast, a 2.3-kb human enhancer fragment exhibited high-level activation in conjunction with either the human or mouse ADA promoter in both the MOLT 4 (human) and S49 (murine) T-cell lines. Interestingly, the murine ADA promoter is significantly stronger than the human promoter in driving cat expression in transient transfection assays in all the T-cell lines tested.

Disruption of the adenosine deaminase (ADA) gene using a dicistronic promoterless construct: production of an ADA-deficient homozygote ES cell line

In man, deficiency of ADA activity is associated with an autosomal recessive form of severe combined immunodeficiency (SCID), a disease with profound defects both cellular and humoral immunity. Current treatments of ADA deficient patients include bone marrow transplantation, enzyme replacement and somatic gene therapy. The mechanism of the selective immune cell pathogenesis in ADA-SCIDS is, however, still poorly understood. Thus, the generation of an ADA deficient mouse model will be of considerable benefit to understand better the pathophysiology of the disorder and to improve the gene therapy treatments. We have disrupted the adenosine deaminase (ADA) gene in embryonic stem cells using a new efficient promoter trap gene-targeting approach. To this end, a dicistronic targeting construct containing a promoterless IRES beta geo cassette was used. This cassette allows, via the internal ribosomal entry site (IRES), the direct cap-independent translation of the beta geo reporter gene which encodes a protein with both beta-galactosidase and neomycin activities. After indentification of targeted clones by Southern blot, successful inactivation of the ADA gene was first confirmed by producing, from our heterozygote clones, an homozygote cell line. This line shows no ADA activity as judged by zymogram analysis. Second, we have been able to detect in the targeted clones, a specific beta galactosidase activity using a sensitive fluorogenic assay. The targeted ES cell clones are currently being injected into blastocysts to create an ADA deficient mouse model.

cry IA(b) transcript formation in tobacco is inefficient

Chimaeric PCaMV35Scry genes direct in tobacco mesophyll protoplasts mRNA levels of less than one transcript per cell. We provide evidence that this low cytoplasmic cry IA(b) mRNA level is not due to a rapid turnover but rather results from a marginal import flow of cry messenger into the cytoplasm. Run-on assays indicate that the frequency of transcription initiation is not limiting. However, the cry precursor mRNA carries at least three regions that are recognized as introns. The absence of high cytoplasmic levels of spliced cry mRNAs suggests that these mRNAs are unstable and/or not efficiently made. Point mutations in the 5' splice site of the most distal intron allows high accumulation levels of the full-length mRNA. This implies that the inefficient formation of full-size mRNA is a major cause of the low expression level of chimaeric cry IA(b) genes in tobacco.

The matrix Gla protein gene is a marker of the chondrogenesis cell lineage during mouse development

Matrix Gla protein (MGP) is, along with osteocalcin, a skeletal member of the family of extracellular mineral-binding Gla proteins. Although the precise function of these proteins remains obscure, circumstantial evidence suggests that they play a role in endochondral ossification. As a first step toward understanding MGP function we have performed a preliminary characterization of its promoter element and studied the developmental pattern of expression of this gene. DNA transfection experiments indicate that the mouse MGP promoter functions better in cells expressing the MGP gene than in cells that do not express the gene. During mouse development, MGP gene expression is detectable as early as day 10.5 of embryonic development (E10.5), before any skeletal structures are identifiable. In situ hybridization analysis shows that MGP mRNA is initially present at the mesenchymal epithelial interphase in lung and limb buds. As development proceeds, MGP gene is predominantly expressed in cells of the chondrocytic lineage in areas that will undergo endochondral ossification as well as in areas that will remain cartilaginous, such as the trachea and bronchi. In growth plate cartilage, MGP mRNA is present in resting, proliferative, and late hypertrophic chondrocytes. Surprisingly, MGP mRNA is absent from the early hypertrophic chondrocytes and from the osteoblasts. Finally, the MGP gene is expressed at a lower level in kidney medulla and uterus smooth muscle but not in brain, spleen, or heart during development. This study demonstrates that during development MGP gene expression occurs early and is predominant at the epithelial mesenchymal interfaces, principally of lung and limb buds, and in cells of the chondrocytic lineage. This finding raises the intriguing possibility that MGP may play distinct roles during embryogenesis and in the adult organism.

Hereditary overexpression of adenosine deaminase in erythrocytes: evidence for a cis-acting mutation

Overexpression of adenosine deaminase (ADA) in red blood cells is inherited as an autosomal dominant trait and causes hemolytic anemia. The increased ADA activity in erythrocytes is due to an increase in steady-state levels of ADA mRNA of normal sequence. Increased ADA mRNA may be due to a cis-acting mutation which results in increased transcription or a loss of down-regulation during erythroid differentiation. Alternatively, it is possible that the mutation is in a trans-acting factor which interacts with normal ADA transcriptional elements to cause overexpression in red blood cells. To discriminate between a cis-acting and a trans-acting mutation, we took advantage of a highly polymorphic TAAA repeat located at the tail end of an Alu repeat approximately 1.1 kb upstream of the ADA gene. Using PCR to amplify this region, we identified five different alleles in 19 members of the family. All 11 affected individuals had an ADA allele with 12 TAAA repeats, whereas none of the 8 normal individuals did. We conclude that this disorder results from a cis-acting mutation in the vicinity of the ADA gene.

Antitermination of vaccinia virus early transcription: possible role of RNA secondary structure

Transcription of vaccinia early genes by the viral RNA polymerase terminates downstream of a signal sequence TTTTTNT in the nontemplate DNA strand. Signal recognition occurs at the level of the sequence UUUUUNU in nascent RNA and depends on a virus-encoded termination factor (VTF). The presence of TTTTTNT elements within protein encoding regions of some early genes requires that these 5' proximal signals be ignored in order to achieve early expression of the full-sized proteins. In the case of the A18R gene, which contains a proximal terminator that is not utilized in vivo (Pacha et al., J. Virol. 64, 3853-3863 (1990)), the TTTTTNT sequence can be folded into a potential hairpin structure such that UUUUUNU would be part of a duplex stem in the nascent RNA. We find that the A18R putative hairpin is unable to promote factor-dependent termination in a purified in vitro transcription system. Sequence manipulations that abrogate the potential to form an RNA hairpin restore the activity of the TTTTTNT motif. The in vitro studies suggest that antitermination at the proximal site of the A18R gene may be mediated by secondary structure in the nascent RNA, and that early termination involves recognition by VTF and/or RNA polymerase of the UUUUUNU sequence in single-stranded form.