An initiation element in the yeast CUP1 promoter is recognized by RNA polymerase II in the absence of TATA box-binding protein if the DNA is negatively supercoiled

Abstract

Purified RNA polymerase II initiated transcription from the yeast CUP1 promoter fused to a C-less cassette if the DNA was negatively supercoiled. Relaxed plasmid was not transcribed. Transcription did not require addition of any other transcription factors. TATA box-binding protein (TBP) was not detectable in the polymerase preparation and the TATA box was not required. Deletion analysis of the CUP1 promoter revealed that a 25-bp element containing the initiation region was sufficient for recognition by polymerase. Two transcription start sites were mapped, one of which is identical to one of the two major start sites observed in vivo. Our observations can be accounted for by using a theoretical analysis of the probability of DNA melting within the plasmid as a function of superhelix density: the CUP1 initiation element is intrinsically unstable to superhelical stress, permitting entry of the polymerase, which then scans the DNA to locate the start site. In support of this analysis, the CUP1 promoter was sensitive to mung bean nuclease. These observations and a previous theoretical analysis of yeast genes support the idea that promoters are stress points within the DNA superhelix. The role of transcription factors might be to mark the promoter and to regulate specific melting of promoter DNA.

Figure 1

Purification of histidine-tagged Pol II from yeast. (A) Outline of purification scheme for 6-histidine-tagged Pol II. (B) Analysis of proteins in the Pol II preparation by SDS-PAGE. The subunits of Pol II (Rpb1–12) are indicated. (C) Western blot with anti-yTBP antibody. Twenty micrograms (about 15 pmol) purified Pol II was loaded onto the gel. The indicated amounts of purified recombinant yTBP were loaded onto the gel as quantitation standards.

Figure 2

Transcription of negatively supercoiled plasmids containing the CUP1 promoter by Pol II. (A) Map of the insert in pCUP1C-less(410). 3′URF, sequence from the termination region of the gene neighboring CUP1 in the yeast genome; UAS, upstream activating sequences (binding sites for Ace1 and HSF); T, putative TATA boxes; IE, initiation element (contains two point mutations as described in the text); synthetic cassette, synthetic C-less cassette. (B) Analysis of supercoiled pCUP1C-less(410) preparations in two-dimensional chloroquine gels. (Right) A mixture of plasmids prepared at superhelix densities: 0, −0.08, −0.12, and −0.16 (corresponding to 0, 26, 40, and 53 negative supercoils in a plasmid of 3474 bp). (Left) As the right panel but with plasmid prepared at native superhelix density (about −0.055) also added. (C) Transcription of negatively supercoiled pCUP1C-less(410) by purified Pol II. Analysis of radiolabeled transcripts in a sequencing gel. -C, -G: transcripts synthesized in the absence of CTP or GTP, respectively. Markers: pBR322 DNA digested with MspI. Superhelix densities (σ) are indicated. (D) Plot of amount of transcript initiated at the CUP1 promoter (phosphorimager units) against the square of the superhelix density for the data in C.

Figure 3

The start sites within the initiation element of CUP1 are similar to those reported in vivo. Transcripts synthesized by Pol II from pCUP1C-less(100) [as pCUP1C-less(410) but with a shorter synthetic cassette] analyzed in a sequencing gel; sizes indicated at right. Markers: run-off transcripts synthesized by T7 RNA polymerase from pBluescript II SK(+) cut with XbaI, EagI, or BstXI. The sequence of the initiation region of CUP1 (wt) with the start sites in vivo marked (Karin) is shown above the mutated sequences used in pCUP1C-less(100) and pCUP1C-less(T1/T10) with observed start sites.

Figure 4

Deletion analysis of the CUP1 promoter: transcription by Pol II requires only the initiation element. (A) Deletion constructs tested (abbreviations as in legend to Fig. 3). (B) Transcripts synthesized from deletion constructs by Pol II analyzed in a sequencing gel. Only the top part of the gel is shown. The bar separates two different experiments.

Figure 5

The CUP1 promoter is intrinsically unstable to superhelical stress: SIDD analysis and mung bean nuclease sensitivity. (A) SIDD analysis: Probability of melting (Left) and free energy of destabilization (Right) for the sequence of pCUP1C-less(410) as a function of superhelix density. bla, β-lactamase gene; 3′U, URF; P, CUP1 promoter; C, synthetic cassette. (B) Mapping of sites sensitive to mung bean nuclease in pCUP1C-less(410) as a function of superhelix density. pCUP1C-less(410) at the superhelix densities indicated was incubated with mung bean nuclease and cleavage sites were identified by indirect end-labeling using the AatII–NdeI (252 bp) fragment as probe. A Southern blot of an alkaline agarose gel is shown. Markers: pCUP1C-less(410) cut with AatII (3474 bp) and XmnI (3155 bp), SapI (1998 bp), HindIII (1304 bp), SpeI (1103 bp), MfeI (853 bp), SacI (471 bp), or NdeI (252 bp).