In vitro transcription by purified yeast RNA polymerase II. Coarse promoter mapping on homologous cloned genes

The active role of DNA as a chromatin organizer

Histone octamers (hos) and DNA topoisomerase I contribute, along with other proteins, to the higher order structure of chromatin. Here we report on the similar topological requirements of these two protein model systems for their interaction with DNA. Both histone octamers and topoisomerase I positively and consistently respond to DNA supercoiling and curvature, and to the spatial accessibility of the preferential interaction sites. These findings (1) point to the relevance of the topology-related DNA conformation in protein interactions and define the particular role of the helically phased rotational information; and (2) help to solve the apparent paradoxical behaviour of ubiquitous and abundant proteins that interact with defined DNA sites in spite of the lack of clear sequence consensuses. Considering firstly, that the interactions with DNA of both DNA topoisomerase I and histone octamers are topology-sensitive and that upon their interaction the DNA conformation is modified; and secondly, that similar behaviours have also been reported for DNA topoisomerase II and histone H1, a topology-based functional correlation among all these determinants of the higher order structure of chromatin is here suggested.

Effects of DNA topology in the interaction with histone octamers and DNA topoisomerase I

Several simple proteins and complex protein systems exist which do not recognize a defined sequence but--rather--a specific DNA conformation. We describe experiments and principles for two of these systems: nucleosomes and eukaryotic DNA topoisomerase I. Evidences are summarized that describe the effects of negative DNA supercoiling on nucleosome formation and the influence of DNA intrinsic curvature on their localization. The function of the DNA rotational information in nucleosome positioning and in the selection of multiple alternative positions on the same helical phase are described. This function suggests a novel genetic regulatory mechanism, based on nucleosome mobility and on the correlation between in vitro and in vivo positions. We observe that the same rules that determine the in vitro localization apply to the in vivo nucleosome positioning, as determined by a technique that relies on the use of nystatin and on the import of active enzymes in living yeast cells. The sensitivity of DNA topoisomerase I to the topological condition of the DNA substrate is reviewed and discussed taking into account recent experiments that describe the effect of the DNA tridimensional context on the reaction. These topics are discussed in the following order: (i) Proteins that look for a consensus DNA conformation; (ii) Nucleosomes; (iii) Negative supercoiling and nucleosomes; (iv) DNA curvature/bending and nucleosomes; (v) Multiple positioning; (vi) Multiple nucleosomes offer a contribution to the solution of the linking number paradox; (vii) Rotational versus translational information; (viii) A regulatory mechanism; (ix) DNA topoisomerase I; (x) DNA topoisomerase I and DNA supercoiling: a regulation by topological feedback; (xi) DNA topoisomerase I and DNA curvature; (xii) The in-and-out problem in the accessibility of DNA information; (xiii) The integrating function of the free energy of supercoiling.

Assembly of RNA polymerase II preinitiation complexes before assembly of nucleosomes allows efficient initiation of transcription on nucleosomal templates

We have previously shown that assembly of nucleosomes on the DNA template blocks transcription initiation by RNA polymerase II in vitro. In the studies reported here, we demonstrate that assembly of a complete RNA polymerase II preinitiation complex before nucleosome assembly results in nucleosomal templates which support initiation in vitro as efficiently as naked DNA. Control experiments prove that our observations are not the result of slow displacement of nucleosomes by the transcription machinery during chromatin assembly, nor are they an artifact of inefficient nucleosome deposition on templates already bearing an RNA polymerase. Thus, the RNA polymerase II preinitiation complex appears to be resistant to disruption by subsequent nucleosome assembly.

Assembly of RNA polymerase II preinitiation complexes before assembly of nucleosomes allows efficient initiation of transcription on nucleosomal templates

We have previously shown that assembly of nucleosomes on the DNA template blocks transcription initiation by RNA polymerase II in vitro. In the studies reported here, we demonstrate that assembly of a complete RNA polymerase II preinitiation complex before nucleosome assembly results in nucleosomal templates which support initiation in vitro as efficiently as naked DNA. Control experiments prove that our observations are not the result of slow displacement of nucleosomes by the transcription machinery during chromatin assembly, nor are they an artifact of inefficient nucleosome deposition on templates already bearing an RNA polymerase. Thus, the RNA polymerase II preinitiation complex appears to be resistant to disruption by subsequent nucleosome assembly.

Structural features of the DNA template required for transcription in vitro by yeast RNA polymerase B (II)

Yeast RNA polymerase II initiates in vitro transcription at two sites located within the vector DNA and the cloned promoter, on a recombinant plasmid DNA containing the yeast iso1 cytochrome c promoter. Both initiation sites are found within a DNA fragment hypersensitive to osmium tetroxide modification. Using a series of yeast iso1 cytochrome c promoter deletions, we have characterized an upstream DNA sequence required for optimal transcription from this site and shown in this case a correlation between osmium sensitivity and the capacity of RNA polymerase to initiate. However, perturbation of the double helix is not sufficient to generate a transcription initiation site. Insertion of 28 alternating AT residues at the EcoRV site of pBR322 generates an site hypersensitive to osmium tetroxide modification, that does not serve as a transcription start site.

Eukaryotic RNA polymerases

This review will attempt to cover the present information on the multiple forms of eukaryotic DNA-dependent RNA polymerases, both at the structural and functional level. Nuclear RNA polymerases constitute a group of three large multimeric enzymes, each with a different and complex subunit structure and distinct specificity. The review will include a detailed description of their molecular structure. The current approaches to elucidate subunit function via chemical modification, phosphorylation, enzyme reconstitution, immunological studies, and mutant analysis will be described. In vitro reconstituted systems are available for the accurate transcription of cloned genes coding for rRNA, tRNA, 5 SRNA, and mRNA. These systems will be described with special attention to the cellular factors required for specific transcription. A section on future prospects will address questions concerning the significance of the complex subunit structure of the nuclear enzymes; the organization and regulation of the gene coding for RNA polymerase subunits; the obtention of mutants affected at the level of factors, or RNA polymerases; the mechanism of template recognition by factors and RNA polymerase.

Mapping of promoter-proximal regions by in vitro transcription of two Agrobacterium tumefaciens Ti-plasmids

Transcription initiation sites were mapped on both the octopine pTi Ach5 and the nopaline pTI C58 plasmids of Agrobacterium tumefaciens. Transcription ternary complexes were subjected to electrophoresis on agarose gels, prior and/or subsequent to restriction endonuclease digestion of the DNA template, evidenced by autoradiography and located on the restriction maps of the two tumor-inducing plasmids. A. tumefaciens RNA polymerase promptly recognizes and starts transcription on a few sequences which include the T-DNA.

Selective transcription of a cloned cauliflower mosaic virus DNA fragment in vitro by soybean RNA polymerase II in the presence of dinucleotide primers

Transcription of a cloned cauliflower mosaic virus (CaMV) DNA fragment (plasmid pCa 8) was studied at a low enzyme: DNA ratio. Preincubation with purine nucleoside triphosphates leads to essentially random transcription, while in the presence of a dinucleoside monophosphate and a purine nucleoside triphosphate in the preincubation medium certain combinations prime preferential transcription of the eucaryotic moiety of the chimeric plasmid. Characterisation of transcription primed by the most efficient combination, ApG + ATP, shows that a low enzyme: DNA ratio is absolutely essential for selective initiation. Interestingly the presence of the eucaryotic insertion is essential for the transcription of vector sequences. Analysis of RNA primed by ApG + ATP and of short chains synthesised in the presence of the GTP analogue 3'-OMeGTP shows a high degree of selectivity of transcription initiation sites. Hybridisation of primed RNA to restriction fragments of pCa8 shows that initiation occurs within a limited region of the inserted CaMV fragment.

RNA polymerase II ternary transcription complexes generated in vitro

Ternary transcription complexes have been formed with a HeLa cell extract, a specific DNA template, and nucleoside triphosphates. The assay depends on the formation of sarkosyl-resistant initiation complexes which contain RNA polymerase II, template DNA, and radioactive nucleoside triphosphates. Separation from the other elements in the in vitro reaction is achieved by electrophoresis in agarose - 0.25% sarkosyl gels. The mobility of the ternary complexes in this system cannot be distinguished from naked DNA. Formation of this complex is dependent on all parameters necessary for faithful in vitro transcription. Complexes are formed with both the plasmid vector and the specific adenovirus DNA insert containing a eucaryotic promoter. The formation of the complex on the eucaryotic DNA is sequence-dependent. An undecaribonucleotide predicted from the template DNA sequence remains associated with the DNA in the ternary complex and can be isolated if the chain terminator 3'-0-methyl GTP is used, or after T1 ribonuclease treatment of the RNA, or if exogenous GTP is omitted from the in vitro reaction. This oligonucleotide is not detected in association with the plasmid vector. Phosphocellulose fractionation of the extract indicates that at least one of the column fractions required for faithful runoff transcription is required for complex formation. A large molar excess of abortive initiation events was detected relative to the level of productive transcription events, indicating a 40-fold higher efficiency of transcription initiation vs. elongation.

Topological modifications and template activation are induced in chimaeric plasmids by inserted sequences

The effect of the insertion of foreign genes or gene systems in closed DNA domains has been investigated in vitro in purified systems. We observe that in chimaeric plasmids two apparently independent classes of modifications, (1) functional and (2) topological, do take place in defined instances. (1) Among the screened yeast gene systems, examples have been found of DNA sequences that upon insertion cause activation of in vitro transcription of distant genes. (2) Foreign DNA sequences may lead to new topological features of the harbouring plasmids; it is shown that more than one S1-sensitive secondary structure may be contemporaneously present on the same chimaeric plasmid. DNA superhelicity is a prerequisite of these modifications. The two classes of effects (1) functional and (2) topological are not a priori directly related one to the other but appear to be two independent consequences of the same cause: the insertion of foreign DNA sequences into closed DNA domains. These observations suggest a regulatory model of gene expression based on alternative topologies of closed DNA domains.