| dc.description.abstract | Structural polymorphism in DNA was envisaged from fiber diffraction studies, though all forms were initially believed to be monotonous right-handed. However, subsequent theoretical studies suggested that the DNA helix could deviate significantly from the right-handed structure, and later single-crystal X-ray studies of synthetic oligonucleotides substantiated the sequence dependence of DNA structure and led to speculations that such structures might have important roles in protein–nucleic acid interactions. Among the structures that received considerable attention from the structural viewpoint are cruciforms, left-handed Z-DNA, bent DNA, and triplexes. Unlike the usual B-DNA structure, these unusual DNA structures require a specific sequence motif. Furthermore, supercoiling of DNA has been found to be obligatory for the stabilization of some of these structures under physiological conditions. Characteristic exposed bases in these unusual structures were shown to be the preferred sites for single-strand nucleases and certain chemical probes. Identification of certain proteins that recognize and stabilize these structures indicated that some of these unusual DNA structures per se may be sites for genetic regulation. Indeed, later studies showed that some of the cis-acting regulatory elements of transcription occurring in the flanking regions as well as within the genes met the necessary sequence criteria required for adopting unusual DNA structures. Identification of single-strand specific nuclease and DNase I hypersensitive sites corresponding to these regions of transcriptionally active genes further raised the possibility that such unusual DNA structures might play an important role in transcriptional regulation. Thus, the questions addressed in this thesis concern the role of template DNA structure in transcription control, both at the level of initiation and elongation.
At the onset of this work, our laboratory was engaged in designing and probing unusual DNA structures in various systems and attempting to understand the DNA structure–function relationship. Preparation of form V DNA, the model system used for in vitro transcriptional studies in this thesis, was well standardized. A topologically unlinked (Lk = 0), highly supercoiled molecule, form V DNA, is reconstituted by annealing complementary single-stranded circles. Topological constraints require that every right-handed helical turn be compensated by a left-handed turn or a negative supercoil. Thus, a large portion of the form V molecule is forced to adopt unusual DNA conformations such as cruciforms, left-handed structures, melted regions, etc. About 35–40% of the sequence adopts left-handed helical structure, a major portion of which exists in Z-conformation. Precise locations of these altered structures in pBR322 form V DNA have been previously mapped in our laboratory exploiting the structure specificity of restriction endonucleases and methylases. These structures have been shown not only to be distributed both within and around genes but also to be folded uniquely. Thus, pBR322 form V DNA, a natural DNA sequence, was an obvious choice for the in vitro studies aimed at understanding the role of unusual DNA structures in transcription control.
For the in vivo studies, we developed a ‘structural cassette’ approach which involves knowledge-based design of oligonucleotide sequences having the potential to adopt unusual DNA structures. The oligonucleotides were synthesized, and functional assays were performed after introducing these cassettes into an appropriate expression vector, either in the promoter region or within the gene. The various questions addressed in this thesis are:
(a) Does unusual DNA structure play a role in regulation of transcription If so, at what level
(b) Is it feasible to employ a multicutter restriction enzyme with a single-hit approach to probe the altered structures in pBR322 form V DNA If so, do the flanking sequences influence structural alteration as suggested previously
(c) Is it possible to use highly supercoiled form V DNA as a template for E. coli RNA polymerase If so, how efficiently does it transcribe
(d) How does RNA polymerase respond to the unusual DNA structures present within or around the genes in pBR322 form V DNA
(e) Is it possible to stabilize designed topologically unlinked DNA structures comprising right-handed B and left-handed Z helical conformations alternating after every half turn of the helix under physiological conditions
(f) Will such topologically unlinked structures, present within the promoter region despite having consensus promoter sequences, facilitate transcription initiation in vivo
(g) Is it possible to design structural elements within the gene without altering the native amino acid sequence using degenerate codons
(h) Do these designed structural elements modulate transcription elongation
Biochemical and molecular biology methods were employed to seek probable answers.
Chapter 1 provides a brief account of major advances in our understanding of supercoil-stabilized DNA structures, followed by a discussion on prokaryotic transcriptional machinery and its various regulatory aspects. A concise description of the effect of DNA supercoiling and supercoil-stabilized DNA structures on the transcription process concludes Chapter 1.
Chapter 2 deals with the preparation and purification of pBR322 form V DNA. In brief, preparation of form V DNA involves controlled digestion of supercoiled DNA with DNase I to generate a single nick per molecule. The complementary single-stranded circular DNA molecules are isolated on an alkaline sucrose density gradient and annealed to generate paranemically coiled form V DNA. The single-hit approach was extended to 26 (CCGG) sites recognized by HpaII restriction endonuclease. The choice of HpaII was based on the fact that one of the (CCGG) sites was present at the junction region of a potential cruciform structure, previously shown to cause transcriptional arrest within the rep gene in pBR322 form V. In good agreement with previous studies, results suggested that the flanking sequences influence the propensity of (CCGG) sites to adopt B or non-B conformation. Compilation of all structural studies on pBR322 form V DNA is presented, and logical deductions are drawn.
Chapter 3 describes in vitro studies using pBR322 form V DNA as a template for E. coli RNA polymerase. Despite the high level of superhelical density, efficient transcription was observed. Analysis of transcriptional products showed the presence of only one short transcript corresponding to the rep gene. The transcript initiated accurately but was aborted due to an elongation block within the gene. This elongation block could be relieved once the supercoiling force was alleviated. Based on MspI methylation sensitivity of the pause site and the HpaII single-hit analysis, a stabilized cruciform structure was shown to be the likely candidate for the observed transcriptional elongation block. Earlier enzymatic probing showed that the promoter region of the tet^ gene comprised alternating B and non-B structures, which could explain the absence of initiation at this promoter under in vitro conditions. Results suggest that structural alterations within natural DNA can act as transcriptional blocks at both initiation and elongation stages under in vitro conditions.
Chapter 4 shows that a polymer of (CGCGCGATCGAT), comprising Z- and B-helicogenic segments, adopts alternating right- and left-handed DNA structures after every half turn of the helix. A topologically unlinked structure, mimicking a short segment of form V DNA, could indeed be stabilized in a supercoiled plasmid under physiological conditions. Three units (36 bp) of this structural cassette were cloned into the ClaI site of pBR322 within the tetracycline promoter region to generate recombinant plasmid pRM36. Two-dimensional chloroquine gel, anti Z-DNA antibody binding, and fine mapping using S1 nuclease showed the presence of an alternating B-Z structure with sharp B-Z junctions. Alignment of the tetracycline promoter in pBR322 with the putative promoter region of pRM36 showed significant homology at the -35 consensus sequence (Pribnow box retained), although the distance between -35 and -10 was 15 bp instead of the optimum 17 bp. E. coli cells harboring pRM36 showed sensitivity to tetracycline. A modified plasmid, pRB38, retained structural features but restored optimal -35/-10 spacing; however, cells remained sensitive, indicating that the altered promoter structure itself precludes transcription initiation in vivo.
Chapter 5 employed a ‘structural cassette’ approach to redesign the N-terminal region of the -galactosidase gene. Using a computer program developed in our laboratory, the EcoRI-HindIII segment of the -galactosidase gene in pUC19 was redesigned with degenerate codons to introduce an inverted repeat capable of forming a cruciform. Minor amino acid changes were introduced to stabilize the structure, without affecting -galactosidase activity. E. coli JM109 cells harboring recombinant plasmid pSBCl-9 showed several-fold reduction in -galactosidase activity. Cruciform extrusion was confirmed by two-dimensional chloroquine gel electrophoresis. A control plasmid, pSBmCl-4, with codon-shuffled sequences but identical amino acids, restored activity, confirming the role of inverted repeat sequences with cruciform potential in transcription attenuation in vivo. Dot blot analysis of total RNA further substantiated this.
Results from these model systems provide early evidence that unusual DNA structures can act in cis to regulate gene expression in vivo. | |
