Single molecule microscopy reveals mechanistic insight into RNA polymerase II preinitiation complex assembly and transcriptional activity

Abstract

Transcription by RNA polymerase II (Pol II) is a complex process that requires general transcription factors and Pol II to assemble on DNA into preinitiation complexes that can begin RNA synthesis upon binding of NTPs (nucleoside triphosphate). The pathways by which preinitiation complexes form, and how this impacts transcriptional activity are not completely clear. To address these issues, we developed a single molecule system using TIRF (total internal reflection fluorescence) microscopy and purified human transcription factors, which allows us to visualize transcriptional activity at individual template molecules. We see that stable interactions between polymerase II (Pol II) and a heteroduplex DNA template do not depend on general transcription factors; however, transcriptional activity is highly dependent upon TATA-binding protein, TFIIB and TFIIF. We also found that subsets of general transcription factors and Pol II can form stable complexes that are precursors for functional transcription complexes upon addition of the remaining factors and DNA. Ultimately we found that Pol II, TATA-binding protein, TFIIB and TFIIF can form a quaternary complex in the absence of promoter DNA, indicating that a stable network of interactions exists between these proteins independent of promoter DNA. Single molecule studies can be used to learn how different modes of preinitiation complex assembly impact transcriptional activity.

Figure 1.

Transcriptional activity can be monitored by the release of an oligo containing a quencher molecule. (A) Transcription from a 3-piece chimeric heteroduplex template. Reactions were performed in the absence of GTP, which halted Pol II after synthesis of a 35 nt RNA due to the presence of the first C in the template strand at +36. ³²P-labeled RNA transcripts were resolved on a 20% denaturing polyacrylamide gel. (B) Schematic of the three-piece quenched construct used to visualize transcription. (C) PICs migrate slower than Pol II/DNA complexes in native gels. Migration of the fluorescent three-piece heteroduplex template was monitored with Pol II and the GTFs or with Pol II alone. Shown is a 4% native gel scanned for Cy3 fluorescence. (D) Release of the quencher oligo in the three piece DNA construct pictured in panel B is NTP dependent. Shown is a 4% native gel scanned for both Cy3 (left panel) and Cy5 (right panel) fluorescence. The positions at which free DNA and PICs migrate are indicated.

Figure 2.

Pol II binding to heteroduplex DNA is not highly dependent upon GTFs. (A) Schematic of the surface used to immobilize Pol II on a glass coverslip. (B) The level of binding heteroduplex DNA by surface-immobilized Pol II did not significantly change with addition of TBP, TFIIB or TFIIF. Each GTF was individually omitted from reactions, and individually added to Pol II, as indicated. Binding events were measured as the number of colocalized (Cy3 and Cy5) spot pairs on the surface. The bars represent the average of five separate movies and errors bars represent one standard deviation. The average number of spot pairs in the absence of 8WG16 (1.2 ± 0.6) and Pol II (0.8 ± 0.7) are not visible on this scale.

Figure 3.

Transcriptional activity at the single molecule level is dependent upon GTFs. Pol II was immobilized and the three-piece quenched construct was used. The percentage of Cy3 molecules that colocalized with Cy5 was determined before and after the addition of NTPs (gray versus black bars). Each GTF was individually omitted from reactions, and individually added to Pol II, as indicated. The bars represent the average of five separate movies and errors bars represent one standard deviation.

Figure 4.

The order of assembly of PICs affects their activity. (A) TFIIF can incorporate into complexes last and yield transcriptionally active PICs. The bars labeled ‘All factors’ and ‘-TFIIF’ were taken from Figure 3, as a point of comparison. The bars represent the average of five separate movies and errors bars represent one standard deviation. (B) TFIIB is unable to incorporate into and activate complexes containing immobilized Pol II/TFIIF/TBP/DNA. Bars labeled ‘All factors’ and ‘-TFIIB’ were taken from Figure 3, as a point of comparison. The bars represent the average of five separate movies and errors bars represent one standard deviation. (C) TFIIB can activate Pol II/TBP/DNA complexes in the absence of TFIIF. Bars labeled ‘-TFIIF’ and ‘-TFIIF, -TFIIB’ were taken from Figure 3. The bars represent the average of five separate movies and errors bars represent one standard deviation.

Figure 5.

Combinations of GTFs and Pol II form stable complexes in the absence of DNA. (A) The indicated complexes of Pol II and GTFs were pre-assembled on the surface. NTP-dependent transcriptional activity was evaluated after the remaining GTFs and promoter DNA were added. The bar labeled ‘All factors’ was taken from Figure 3, as a point of comparison. The bars represent the average of five separate movies and errors bars represent one standard deviation. (B) Pol II, TFIIB, TFIIF and TBP are capable of forming a stable complex in the absence of DNA. The schematic shows the 18 bp Alexa555/Alexa647 labeled piece of TATA DNA used to probe the presence of TBP in complexes assembled on surface-immobilized Pol II. With the exception of bar 2 (-TBP) and bar 6, TBP and DNA were pre-incubated prior to addition to the slide. The bars represent the average of five separate movies and errors bars represent one standard deviation.