The protein components that direct and activate accurate transcription by rat RNA polymerase I were studied in extracts of Novikoff hepatoma ascites cells. A minimum of at least two components, besides RNA polymerase I, that are necessary for efficient utilization of templates were identified. The first factor, rat SL-1, is required for species-specific recognition of the rat RNA polymerase I promoter and may be sufficient to direct transcription by pure RNA polymerase I. Rat SL-1 directed the transcription of templates deleted to -31, the 5' boundary of the core promoter element (+ 1 being the transcription initiation site). The second factor, rUBF, increased the efficiency of template utilization. Transcription of deletion mutants indicated that the 5' boundary of the domain required for rUBF lay between -137 and -127. Experiments using block substitution mutants confirmed and extended these observations. Transcription experiments using those mutants demonstrated that two regions within the upstream promoter element were required for optimal levels of transcription in vitro. The first region was centered on nucleotides -129 and -124. The 5' boundary of the second domain mapped to between nucleotides -106 and -101. DNase footprint experiments using highly purified rUBF indicated that rUBF bound between -130 and -50. However, mutation of nucleotides -129 and -124 did not affect the rUBF footprint. These results indicate that basal levels of transcription by RNA polymerase I may require only SL-1 and the core promoter element. However, higher transcription levels are mediated by additional interactions of rUBF, and possibly SL-1, bound to distal promoter elements.Eucaryotic rRNA genes (rDNAs) code for three of the four rRNA molecules (18S, 5.8S, and 28S rRNAs). These three RNAs are products of degradative processing of a larger precursor (40S to 47S pre-rRNA). The genes that serve as the template for this transcription are present in multiple copies (approximately 200 per haploid genome) and are organized as clusters of tandem repeats. In interphase cells, the genes are localized to the nucleolus, where they are transcribed by RNA polymerase I. Each repeat consists of a transcribed portion and a nontranscribed spacer (reviewed in references 27 and 28). In at least two species, Xenopus laevis and Drosophila melanogaster, the nontranscribed spacer is transcribed (9,23,41), and at least part of the nontranscribed spacer of the rat repeat is transcribed from a spacer promoter (6), suggesting that the name of this region needs to be changed.The transcription initiation sites of several mammalian rRNA genes have been identified and sequenced. Functional analysis of the promoters of the mammalian rRNA genes indicates that despite significant sequence differences, the promoters apparently consist of elements with similar functions. That region of the promoter (--31 to -+6) required for transcription in vitro (5,12,16,26, 43) is referred to as the core promoter element (CPE). Under more stringent conditions, a requirement...

We have previously demonstrated that the protein encoded by the retinoblastoma susceptibility gene (Rb) functions as a regulator of transcription by RNA polymerase I (rDNA transcription) by inhibiting UBFmediated transcription. In the present study, we have examined the mechanism by which Rb represses UBFdependent rDNA transcription and determined if other Rb-like proteins have similar eects. We demonstrate that authentic or recombinant UBF and Rb interact directly and this requires a functional A/B pocket. DNase footprinting and band-shift assays demonstrated that the interaction between Rb and UBF does not inhibit the binding of UBF to DNA. However, the formation of an UBF/Rb complex does block the interaction of UBF with SL-1, as indicated by using the 48 kDa subunit as a marker for SL-1. Additional evidence is presented that another pocket protein, p130 but not p107, can be found in a complex with UBF. Interestingly, the cellular content of p130 inversely correlated with the rate of rDNA transcription in two physiological systems, and overexpression of p130 inhibited rDNA transcription. These results suggest that p130 may regulate rDNA transcription in a similar manner to Rb. Oncogene (2000) 19, 4988 ± 4999.

The phosphorylation, DNA-binding and dimerization properties of both forms of the RNA polymerase I transcription factor UBF were studied and compared. Tryptic peptide maps of in vivo 32P-labeled UBF contained four phospho-peptides. Two of these peptides are predicted to derive from the serine-rich, carboxyl-terminal of UBF. This region contains nine consensus phosphorylation sites for casein kinase II, and is one of the regions phosphorylated in vitro by casein kinase II. Analysis of the DNA-binding properties of recombinant forms of UBF1 and UBF2 by Southwestern blots revealed: (1) a role for the NH2-terminal 102 amino acid domain of UBF1/UBF2 in DNA-binding; (2) the importance of the bases from -106 to -101 of the rat ribosomal DNA promoter for the binding of UBF; and (3) functional differences between UBF1 and UBF2. Glutaraldehyde cross-linking and overlay assays using recombinant forms of UBF1 and UBF2 demonstrated that the molecules can form both homodimers and heterodimers. These assays also demonstrated that the NH2-terminal 102 amino acids of UBF plays a significant role in dimerization and that other domains contribute to dimerization. The dimerization properties of recombinant forms of UBF1 and UBF2 were different, suggesting that the HMG box 2 of UBF1, which is partially deleted in UBF2, also contributes to UBF dimerization.

Rat cells contain a DNA-binding polymerase I transcription factor, rUBF, with properties similar to UBF homologs that have been purified from both human (hUBF) and frog (xUBF) cells. In this note we report the affinity purification of rUBF to apparent homogeneity and show that UBFs from both rat and frog have identical footprinting characteristics on templates from either species. Furthermore, xUBF was able to stimulate transcription from rat RNA polymerase I promoters in a partially fractionated rat extract that was UBF dependent. These results strengthen the conclusion that all vertebrate cells contain a UBF homolog whose DNA-binding specificity and function have been strongly conserved.

The spacer promoter of the rat rDNA repeat consists of two functional domains: a core (proximal) element that is sufficient for transcription in vitro, and an upstream (distal) promoter element that increases the efficiency of transcription. Two of the transcription factors that interact with the 45S promoter also interact with the spacer promoter. Rat SL-1, is required for transcription of the spacer promoter by heterologous extracts, e.g. human, and rat SF-1 is required for efficient transcription in vitro. Order-of-addition experiments demonstrated that the preinitiation complex formed by these factors on the spacer promoter is not as stable as the complex formed on the 45S promoter. DNase 1 footprinting experiments demonstrated binding sites for rat SL-1 and SF-1 on the distal element of the spacer promoter. The topology of the domains of the spacer promoter may explain both the reduced stability of the preinitiation complex formed on that promoter and the lower efficiency of transcription of that promoter when compared to the 45S promoter.