Nucleic Acids Research, 2000, Vol. 28, No. 14 2726-2735
© 2000 Oxford University Press
Discrete promoter elements affect specific properties of RNA polymerase II transcription complexes
Department of Biochemistry and Biophysics, Texas A&M University, 2128 TAMU, College Station, TX 77843-2128, USA
Received as resubmission April 4, 2000; Accepted May 17, 2000.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONRESULTSDISCUSSIONREFERENCES |
|---|
The frequency of transcription initiation at specific RNA polymerase II promoters is, in many cases, related to the ability of the promoter to recruit the transcription machinery to a specific site. However, there may also be functional differences in the properties of assembled transcription complexes that are promoter-specific or regulator-dependent and affect their activity. Transcription complexes formed on variants of the adenovirus major late (AdML) promoter were found to differ in several ways. Mutations in the initiator element increased the sarkosyl sensitivity of the rate of elongation and decreased the rate of early steps in initiation as revealed by a sarkosyl challenge assay that exploited the resistance of RNA synthesis to high concentrations of sarkosyl after formation of one or two phosphodiester bonds. Similar, but clearly distinct, effects were also observed after deletion of the binding site for upstream stimulatory factor from the AdML promoter. In contrast, deletion of binding sites for nuclear factor 1 and Oct-1, as well as mutations in the recognition sequence for initiation site binding protein, were without apparent effect on transcription complexes on templates containing the mouse mammary tumor virus promoter.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONREFERENCES |
|---|
The efficiency of promoter-specific transcription by eukaryotic RNA polymerase II (pol II) is subject to control by transcriptional regulatory proteins. In many cases, these proteins have been shown to affect assembly of a transcription complex containing pol II and the general transcription factors (TF) IIA, IIB, IID, IIE, IIF and IIH on template DNA in vitro (1–4) and in vivo (5). However, the action of transcriptional regulators is likely not limited to effects on assembly. For example, the level of transcription of several genes, including Drosophila heat shock gene hsp70 (6,7) and human immunodeficiency virus (8), are regulated by effects on transcription elongation related to release of paused polymerases at promoter-proximal sites. In addition, promoter-specific differences in the ratio of productive to abortive initiation events have been described (9) and control of this ratio provides an attractive target for post-assembly regulation.
In this study, we have used several assays to compare functional properties of assembled pol II transcription complexes on two promoters and their derivatives (Fig. 1). The adenovirus 2 major late (AdML) promoter is the prototype pol II promoter; it contains a TATA element, an associated initiator (10) and a binding site for upstream stimulatory factor (USF) (11–13). The AdML promoter becomes active at the onset of viral DNA replication and directs initiation of a 25 kb transcription unit, the differentially processed products of which serve as mRNAs for viral structural proteins (14). The mouse mammary tumor virus (MMTV) promoter contains a TATA element, a binding site for nuclear factor 1 (NF-1), two adjacent binding sites for the octamer-binding transcription factor Oct-1 (15) and a binding site for a protein we have termed initiation site binding protein (ISBP) (16). The MMTV promoter is subject to positive control by several classes of steroid hormones (17) and a mammary cell-specific enhancer (18,19), as well as negative control by proximal and distal negative regulatory elements (20–22).
|
Our results indicate that transcription complexes assembled on templates containing different promoters can be functionally distinct. Transcription complexes on templates containing the AdML and MMTV promoters show differential sensitivity to the anionic detergent sarkosyl. In addition, mutations in the initiator element of the AdML promoter alter the sensitivity of elongation to inhibition by sarkosyl, as well as the kinetics of early steps in transcription initiation. Deletion of the USF site in the AdML promoter also affects the sarkosyl sensitivity of transcription.
MATERIALS AND METHODS
Plasmids
The pBML
12 template, which contains the wild-type AdML promoter, was derived from pML(C2AT)19 (23). The region of pML(C2AT)19 containing the promoter and G-free cassette was amplified by PCR using pUC forward and reverse universal primers. The amplified product was digested with SmaI, which cuts 3' of the G-free cassette, and HaeIII, which cuts promoter sequences at –61, and the resulting fragment was ligated into the SmaI site of pUC19 to generate pBML(C2AT)19. The G-free cassette of pBML(C2AT)19 was shortened by linearizing the plasmid with KpnI, which cuts 3' of the G-free cassette, and digesting with Bal31 exonuclease. Shortened plasmids were made blunt with the Klenow fragment of DNA polymerase I and recircularized with T4 DNA ligase. Several independent clones were isolated, sequenced and tested in in vitro transcription reactions. pBML12 contains wild-type AdML promoter sequences from –61 to +10 linked to a shortened G-free cassette and generates a G-free transcript of 178 nt (see Fig. 1).
pAdML(Inrmut), containing mutations that alter the initiator element, was constructed from pBML12. Two separate PCR reactions were used to introduce the mutations into the initiator element. The first PCR reaction amplified the promoter region of pBML12 using the pUC reverse universal primer and a primer (5'-GAGGGGAAGGTAGTACTGACGAACGC-3') that hybridized to pBML12 between –12 and +14 with mismatches at positions –3, –2, –1, +4 and +5. The second PCR reaction amplified the G-free cassette of pBML12 and used a primer (5'-GCGTTCGTCAGTACTACCTTCCCCTC-3') that also hybridized to pBML12 between –12 and +14 with mismatches at positions –3, –2, –1, +4 and +5 along with a primer (5'-GAGTGCACCATATGCGGTGT-3') that hybridized downstream of the G-free cassette in pBML12 in a region that contains a NdeI site (CATATG). Both PCR reactions created a ScaI restriction site (AGTACT) within the initiator element of the AdML promoter. Each PCR product was digested with ScaI, the two fragments were ligated together and the ligated product was isolated by polyacrylamide gel electrophoresis. The isolated fragment was digested with XbaI and NdeI, ligated into XbaI- and NdeI-digested pUC19 and transformed into Escherichia coli HB101. The final construct was verified by DNA sequencing (24). A G-free transcript of 178 nt can be generated from pAdML(Inrmut).
Plasmid pAdML(41), containing AdML promoter sequences from –41 to +10, was constructed from pBML12 through PCR amplification of the promoter and G-free cassette. One PCR primer (5'-GGCGAATCTAGAAGGGGGGCTATAA-3') hybridized to pBML12 between –41 and –27 and contained a 5'-overhang with a XbaI restriction site (TCTAGA) and the second primer was the NdeI site primer described above. The PCR product was digested with XbaI and NdeI and the promoter-containing fragment was isolated by polyacrylamide gel electrophoresis. pBML12 was also digested with XbaI and NdeI and the large vector fragment was isolated by agarose gel electrophoresis. Vector and PCR product were then ligated and transformed into E.coli HB101. The final construct was verified by DNA sequencing (24). A G-free transcript of 178 nt can be generated from pAdML(41).
The MMTV promoter template pMBPT3 was described previously (25). It contains MMTV promoter sequences from –109 to +14 and generates a U-free transcript of 172 nt. MMTV promoter templates with a 5'-deletion to –38 (pTF38) or clustered point mutations in the ISBP site [pLS(+6/+8), termed ISBPmut in this report] have also been previously described (25).
Template DNAs were amplified in E.coli strain HB101 and isolated from 1 l bacterial cultures by alkaline extraction (26). Plasmid DNA was purified by CsCl gradient centrifugation followed by size exclusion chromatography on Biogel A-5M (Bio-Rad) and ethanol precipitation. Template concentrations were determined by absorbance at 260 nm.
Nuclear extracts
HeLa cells were maintained at 5–10 x 105 cells/ml in JMEM supplemented with 5% fetal bovine serum (Life Technologies), 0.2 mM L-glutamine, 100 U/ml penicillin G (sodium salt), 100 µg/ml streptomycin sulfate (Life Technologies) and buffered with 10 mM HEPES (pH 7.4). Nuclear extracts were prepared according to the method of Dignam (27), with previously described modifications (28,29). After ammonium sulfate precipitation, the nuclear extract was resuspended in 4–6 ml TM0.5 buffer (50 mM Tris–HCl, pH 7.9, 1 mM EDTA, 12.5 mM MgCl2, 20% glycerol, 500 mM KCl) containing 4 mM DTT and dialyzed at 4°C against four changes of TM0.5 for a minimum of 2 h for each change. The dialyzed nuclear extract was added to 2 ml of DEAE–Sepharose (Amersham Pharmacia Biotech) equilibrated with TM0.5 and mixed gently for 30 min at 4°C. The mixture was centrifuged for 5 min at 3000 r.p.m. at 4°C and the supernatant was dialyzed against TM0.1 (as TM0.5 but with 100 instead of 500 mM KCl) and frozen in 100 µl aliquots for storage at –70°C. This DEAE treatment removes contaminating nucleotides present in the extract. Protein concentration was determined using the method of Bradford (30) with bovine serum albumin as the standard.
In vitro transcription assays
Prior to use in in vitro transcription reactions, thawed nuclear extract (in TM0.1 storage buffer) was treated with DTT (4 mM) for 15 min on ice. We have found that such treatment restores any loss of activity that may have occurred during storage of the extract (data not shown). Transcription complex assembly was initiated by combining 60 µg nuclear extract protein with 1 µg plasmid DNA in a final volume of 42 µl. Transcription complexes were allowed to assemble for 60 min at 30°C. RNA synthesis was initiated by addition of 8 µl of a mixture containing NTPs that included 10 µCi [
-32P]CTP (800 Ci/mmol; NEN Life Sciences). Unless otherwise specified in the figure legends, the final NTP concentrations for reactions with AdML promoter templates were 0.6 mM ATP, 0.6 mM UTP and 20 µM CTP. For reactions with MMTV promoter templates, 0.6 mM GTP replaced the UTP. Sarkosyl (Sigma-Aldrich) was present in some reactions as indicated in the figure legends. Reactions with the AdML promoter also contained RNase T1 (final concentration 1 U/µl), which eliminated low levels of transcription beyond the G-free cassette that were occasionally observed in its absence. RNA synthesis proceeded for 30 min at 30°C unless indicated otherwise in the figure legend. The reactions were terminated by addition of 350 µl of stop buffer [50 mM Tris–HCl, pH 7.5, 1% SDS, 5 mM EDTA, 25 µg/ml tRNA (Sigma-Aldrich)] supplemented with 1000–5000 c.p.m. of a 32P-labeled recovery control RNA synthesized by SP6 RNA polymerase as described below. Proteinase K (Roche Molecular Biochemicals) was added to a final concentration of 0.1 mg/ml and the samples were incubated for 15 min at 37°C. After extraction with 400 µl of phenol:chloroform (1:1) saturated with a solution containing 100 mM Tris–HCl, pH 7.5, 10 mM sodium acetate, 0.1 M NaCl and 1 mM EDTA, the RNA was ethanol precipitated. Samples were centrifuged in a microfuge for 20 min, washed with 70% ethanol, dissolved in 53% deionized formamide and loaded on 8% denaturing polyacrylamide gels containing 8 M urea. Dried gels were exposed to NEF-496 film (Dupont) for 10–18 h with an intensifying screen at –70°C and quantitated using a Fuji Phosphorimager. Individual lanes were normalized by subtracting background and comparison to the labeled RNA recovery control.
Synthesis of recovery control RNA
Recovery control RNA was synthesized using the SP6 promoter of pGEM (Promega Biosciences). The plasmid was linearized by digestion with PvuII, which cuts at a unique site 98 bp downstream of the transcription start site. Typically, 1 µg linear template was added to transcription buffer consisting of 40 mM Tris–HCl, pH 7.5, 6 mM MgCl2, 2 mM spermidine, 10 mM NaCl and 10 mM DTT supplemented with 0.5 mM ATP, GTP and UTP, 0.25 mM CTP, 1 U/ml RNasin (Promega Biosciences) and 10 µCi [-32P]CTP (800 Ci/mmol). Synthesis was initiated by addition of 10 U SP6 RNA polymerase (Life Technologies) and continued for 60 min at 37°C. Reactions were terminated by addition of 380 µl of stop buffer and processed as described above for in vitro transcription reactions. After ethanol precipitation, RNA was resuspended in DEPC-treated water at 1000–5000 c.p.m./ml. In vitro transcription reactions contained 1000–5000 c.p.m. of RNA added with the stop solution as a recovery control.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONREFERENCES |
|---|
The transcription templates that we employed (Fig. 1) were based on constructs containing either the AdML promoter (–61 to +10) linked to a G-free cassette that generates a G-free transcript of 178 nt (pBML
12) or the MMTV promoter (–109 to +14) linked to a T-free cassette that generates a U-free transcript of 172 nt (pMBPT3; 25).
Templates containing derivatives of the AdML promoter included a deletion to –41, which removed the USF site (41). In vitro transcription assays in which wild-type and 41 templates were quantitatively compared routinely gave ~5-fold more transcription from the wild-type template (data not shown). Another derivative (Inrmut) contained mutations in the AdML initiator element. The AdML initiator can be recognized by a number of proteins, including YY1 (31), HIP1/E2F (32,33), TFII-I (34), USF (35) and one or more subunits of TFIID (36–38). It is not clear which of these proteins is most important for the activity of the initiator in vivo (see Discussion). We introduced mutations into the initiator both upstream and downstream of the transcription start site (Fig. 1A) that were predicted to alter the function of the initiator (39,40) and, consistent with this prediction, the overall level of transcription supported by the Inrmut template was ~50% of that obtained from a template containing the wild-type AdML promoter (data not shown).
Templates containing derivatives of the MMTV promoter included a deletion to –38, which removed the NF-1 and octamer sites (38) and mutations that affect binding of ISBP (16) (ISBPmut). These alterations resulted in decreases in the overall level of in vitro transcription to ~30% (38) and 50% (ISBPmut) of that obtained from a template containing the wild-type MMTV promoter (16,25,41,42).
Rate of RNA synthesis from transcription preinitiation complexes
In an initial experiment to characterize transcription in our in vitro system, we examined the rate at which RNA transcripts were synthesized from assembled preinitiation complexes (Fig. 2). Transcription complexes were assembled by incubation of supercoiled template DNA with a HeLa cell nuclear extract for 60 min, nucleoside triphosphates (including [-32P]CTP) were added at time 0 and reactions were terminated at time t (Fig. 2A). Products were fractionated by polyacrylamide gel electrophoresis and specific transcripts were quantitated with a phosphorimager relative to a 32P-labeled internal standard RNA added with the stop solution, which served to normalize for recovery during preparation for electrophoresis (Fig. 2B and C). Under the conditions employed in this experiment, full-length transcripts from templates containing the AdML wild-type, 41 and Inrmut promoters appeared with kinetics that could be closely approximated by a first order reaction with a half-time of ~2 min (curve in Fig. 2C). These results were not specific for transcription on AdML promoter-containing templates, as RNA synthesis by complexes assembled on a template containing the wild-type MMTV promoter occurred with essentially identical kinetics. It should be noted that the data in Figure 2C were separately normalized to the maximum signal obtained from each template; thus, differences in the number of transcripts obtained from the various templates are not apparent in this presentation of the results. On an absolute scale with the wild-type AdML promoter template set at 100, transcript levels from the 41, Inrmut and MMTV wild-type templates were ~25, 50 and 20, respectively. In addition, there is no indication that reinitiation occurred on any of the templates; reinitiation would have been observed as a slow increase in transcript levels after an initial first round of RNA synthesis from each active transcription complex (43) or as a time-dependent appearance of a ladder of near full-length transcripts generated as polymerases from reinitiation events collide with those already stalled at the end of the G-free cassette (44). Based on these results, our assays apparently allow only a single round of initiation from each transcription complex.
|
Sarkosyl inhibition of RNA synthesis
Sarkosyl is an anionic detergent that has been exploited as an inhibitor of transcription by pol II. Sarkosyl appears to function in at least two ways (43,45). At low concentrations (0.01–0.02%) sarkosyl inhibits assembly of a transcription preinitiation complex. RNA synthesis from assembled preinitiation complexes is resistant to these low concentrations but can be inhibited by higher concentrations of sarkosyl. After synthesis of the first one or two phosphodiester bonds, transcription becomes completely resistant to sarkosyl (45,46) and synthesis of a full-length transcript can occur even in the presence of very high concentrations (up to at least 1%; 46).
Several experiments described below depend on the ability of sarkosyl to inhibit RNA synthesis. To verify the sarkosyl concentration necessary for inhibition in our experimental system, we assembled transcription complexes on AdML and MMTV promoter-containing templates in HeLa cell nuclear extract and, after assembly was complete, added varying concentrations of sarkosyl along with the nucleoside triphosphate substrates needed for RNA synthesis (Fig. 3). Concentrations >0.15% sarkosyl completely inhibited RNA synthesis by transcription complexes on both AdML and MMTV promoters. Interestingly, preinitiation complexes on the AdML template were significantly more sensitive to sarkosyl inhibition than those on the MMTV template. Transcription from the MMTV promoter was completely resistant to sarkosyl up to ~0.05%, while transcription from the AdML promoter was partially inhibited at concentrations as low as 0.02%.
|
Differential effects of sarkosyl on the kinetics of RNA synthesis
The rate of transcription elongation is known to be inhibited by sarkosyl, at least in part due to the sarkosyl sensitivity of some pol II elongation factors (47–49; see Discussion). We examined the effect of a low concentration of sarkosyl (0.02%) that is sufficient to block transcription complex assembly (43,45; data not shown) but not prevent RNA synthesis (Fig. 3) on the kinetics of RNA synthesis. Using templates containing several derivatives of the AdML promoter, the time course of synthesis of full-length RNA under these conditions showed a distinct lag (Fig. 4B). The lag was significantly longer for both the
41 and Inrmut templates relative to the template containing the wild-type AdML promoter. The times at which half of the final levels of RNA were reached were ~7 min for the template containing the wild-type promoter and ~9 and 14 min for the 41 and Inrmut templates, respectively. This difference was not observed in the absence of sarkosyl, where RNA synthesis from all three templates followed apparent first order kinetics with a half-time of ~2 min (Fig. 2C, dashed line in Fig. 4B). In contrast, the time course of RNA synthesis from templates containing the MMTV promoter was only minimally affected by sarkosyl (Fig. 4C).
|
Autoradiographs that display the AdML transcripts present as the time course of RNA synthesis progresses are shown in Figure 5. Apparent elongation intermediates smaller than the expected 178 nt transcript are visible with all three AdML promoter templates; these RNAs appear within ~2 min after addition of nucleotides and are chased into full-length products. Based on the sizes of the AdML transcript (178 nt) and the recovery control RNA standard (98 nt), we estimate that the smallest RNAs seen on these autoradiographs are ~50 nt in length. The time courses for the three AdML promoter templates differ in the rates at which the small RNAs are extended to full-length transcripts. The period over which elongation intermediates are seen for the AdML wild-type template extends from ~2 to ~8–10 min (Fig. 5A). The corresponding period for the Inrmut template (Fig. 5C) is distinctly wider, extending from ~4 to >15 min, while that for the
41 template (Fig. 5B) appears intermediate between the two others. It should be emphasized that the transcribed G-free cassettes of the wild-type and 41 templates are identical and the Inrmut template differs only at positions +4 and +5 (Fig. 1A). Overall, these results suggest that the average rate of transcription elongation is differentially affected by sarkosyl on the three AdML promoter templates. This conclusion is considered in more detail in the Discussion.
|
Sarkosyl challenge assay
Differential inhibition of the average rate of elongation is not the only possible cause of template-specific changes in the rate of RNA synthesis observed in Figure 4B. Other steps in transcription, including open complex formation and promoter clearance, could also be differentially sensitive to sarkosyl. One assay that we have used to begin to assess effects on these early steps in transcription initiation is based on the observation that even extremely high concentrations of sarkosyl (up to at least 1%) do not affect the ability of elongation complexes to complete a transcript once the first one or two phosphodiester bonds are formed (43,45,46). In this assay, which we have termed the sarkosyl challenge assay, transcription complexes are assembled and at time 0 nucleoside triphosphates necessary for RNA synthesis are added. At time t, sarkosyl is added to a final concentration of 0.2%, a concentration that, if added at time 0, would completely inhibit transcription (Fig. 3). RNA synthesis is then allowed to proceed for 30 min and the number of full-length transcripts produced is a measure of the number of transcription complexes that had formed one or two phosphodiester bonds (and thus become sarkosyl-resistant) at time t.
Under our standard assay conditions (which included 20 µM CTP) the transition to sarkosyl resistance occurred very rapidly for transcription from both the AdML and MMTV promoter templates (data not shown). Consistent with published results (43,45), the transition to sarkosyl resistance for transcription from the AdML promoter was complete in <1 min (data not shown). We also found that the MMTV promoter completed the transition to sarkosyl resistance in <1 min under these conditions (data not shown). However, when the CTP concentration was lowered from 20 to 5 µM, the rate of transition to sarkosyl resistance for transcription from the AdML template was significantly decreased. For example, in a sarkosyl challenge experiment with the wild-type AdML promoter template in which 0.02% sarkosyl (the same concentration used in the experiments in Figs 4 and 5) was added with the nucleotide substrates at time 0 (Fig. 6A), transition to resistance to 0.2% sarkosyl followed apparent first order kinetics with a half-time of ~1 min (Fig. 6B, solid circles). The rate of this transition was not affected by the presence of 0.02% sarkosyl during the initial phase of the reaction (see below). Transition to sarkosyl resistance remained very fast with all of the MMTV promoter templates under any conditions tested (Fig. 6C).
|
The AdML
41 and Inrmut promoter templates were also used in the sarkosyl challenge assay (Fig. 6B). It is important to note that the transcribed regions of the wild-type and 41 promoters are identical and the mutations introduced into the Inrmut promoter do not change the first 3 nt of the transcript (Fig. 1B); thus, the nucleotides required to be transcribed to attain resistance to 0.2% sarkosyl in the challenge assay are the same in all three promoters. Transcription complexes on both the 41 and Inrmut templates were significantly slower in attaining sarkosyl resistance than those on the wild-type template (Fig. 6B) and the transitions could be reasonably fitted by first order curves with half-times of ~3–4 min. The relative transcription levels in Figure 6B were separately normalized for each promoter to emphasize the rate difference and decreases in the overall level of transcription caused by the mutations are not seen in this presentation of the data. Representative autoradiographs for the sarkosyl challenge assays graphed in Figure 6 are shown in Figure 7 and the slower rate of transition to sarkosyl resistance of the 41 and Inrmut templates is evident. These autoradiographs were exposed to give approximately equal intensities of the specific transcripts and promoter-specific differences in overall transcription levels are not seen.
|
In order to assess whether the template-specific differences observed in Figure 6B were dependent on the low concentration of sarkosyl added with the nucleotide substrates (Fig. 6A), we performed an additional set of sarkosyl challenge assays with the AdML promoter templates in which no sarkosyl was present until the 0.2% sarkosyl challenge at time t (Fig. 8A). Under these conditions, the
41 and wild-type templates behaved essentially identically, while transition to sarkosyl resistance was somewhat slower for the Inrmut template. Interestingly, the kinetics for the AdML wild-type promoter template were independent of the presence of 0.02% sarkosyl (compare Figs 6B and 8B). The curves in Figure 8B represent first order reactions, but the data for the Inrmut template do not fit particularly well; the apparent plateau attained in the first 2 min of the reaction (which was quite reproducible) suggests that there may be two types of complexes formed, one that attains sarkosyl resistance very rapidly and another that is much slower. This observation is considered more fully in the Discussion.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONREFERENCES |
|---|
Template-specific sarkosyl sensitivity of elongation
One of our most striking observations is the reduced rate of RNA synthesis on the
41 and Inrmut templates in the presence of sarkosyl (Figs 4B and 5). For the Inrmut template it could be argued that the slower kinetics of RNA synthesis might result entirely from sequence-specific pausing near the initiation site, the only place where the transcribed regions of the wild-type and Inrmut templates differ (Fig. 1A). According to this idea, the reduced rate of RNA synthesis would not be related to initiator function, but rather to sequence-specific effects of the initial transcribed sequence. It should be noted that, as CTP is the labeled nucleotide in our reactions and is present at limiting concentrations (20 µM in the experiments in Figs 4 and 5), changes in the Inrmut template do not introduce additional C residues into the initial transcribed region; one C is introduced at +5, but the wild-type C at +4 is replaced with an A. While we cannot totally discount potential effects of pausing in the initial transcribed region of the Inrmut template, the slower rate at which nascent transcripts were chased into full-length product (Fig. 5) clearly indicates that the rate of elongation in the presence of a low concentration of sarkosyl is decreased with this template. It seems most likely that the absence of a functional initiator results in a transcription complex with increased sensitivity to sarkosyl.
RNA synthesis with the 41 template was also significantly slower than with the wild-type template in the presence of sarkosyl (Fig. 4B). The wild-type and 41 templates are identical in the transcribed G-free cassette and, therefore, such differences cannot be invoked to account for the differential sarkosyl sensitivity of the rates of RNA synthesis on wild-type and 41 templates. It seems most likely that the absence of USF results in a transcription complex with increased sensitivity to sarkosyl.
For all three of the templates shown in Figure 5, a small fraction (2–3%) of the transcription complexes are able to make a full-length transcript within 1–2 min. This may be the result of a limiting concentration of a specific factor in our extract (48,50) that decreases the sarkosyl sensitivity of elongation.
Template-specific rates of early steps in initiation
Template-specific differences in the AdML promoter templates were also observed with an assay sensitive to early steps in transcription initiation. This assay, which we have termed the sarkosyl challenge assay, measures the transition to a complex that is resistant to high concentrations (0.2%) of sarkosyl and requires the formation of one or two phosphodiester bonds of the nascent transcript (45,46). All three AdML promoter templates are identical at positions +1, +2 and +3 (Fig. 1A) and, therefore, the nucleotides required to be incorporated into the RNA before attaining sarkosyl resistance are the same in each template. In the presence of 0.02% sarkosyl, transition to resistance to the higher sarkosyl concentration was significantly slower for the 41 and Inrmut templates relative to the template containing the wild-type promoter (Fig. 6B). The transition kinetics could be approximated by first order reactions with half-times of ~1 min for the wild-type template and 3–4 min for the 41 and Inrmut templates. Processes required for the transition to sarkosyl resistance could include ATP-dependent steps, such as open complex formation (51,52) and phosphorylation of the C-terminal domain of the largest subunit of pol II (53), as well as phosphodiester bond formation. For the Inrmut template, we cannot eliminate the possibility that the sarkosyl sensitivity of one or more of these processes is sequence-dependent in a manner that is not related to initiator function. However, for the 41 template, sequences near the initiation site are identical to those in the wild-type template and the most plausible cause for their different behavior relates to whether USF is present on the template.
Interestingly, the kinetics of the sarkosyl challenge assay differ for the wild-type and Inrmut templates even in the absence of 0.02% sarkosyl (Fig. 8). While the data in Figure 8B were fitted to first order reactions, the fit was not particularly good for the Inrmut template. The apparent plateau between 0.5 and 1.5 min, which was very reproducible, could be indicative of two classes of transcription complexes that differ in the rate at which they attain resistance to sarkosyl; the distinction between wild-type and Inrmut templates might then be considered as a difference in the relative abundance of the two classes of complexes. It should be emphasized that the slower kinetics are not related to the number of C residues present in the initial transcribed region; the wild-type and Inrmut templates have the identical number and the alterations in the Inrmut template actually change the position of one C from +4 to +5, where it is probably less likely to influence the kinetics of the sarkosyl challenge assay.
Transcription complexes with different functional properties
If, as suggested by the differential activities of the various templates, transcription complexes with distinct functional properties are present on AdML wild-type, Inrmut and 41 templates, the altered function(s) could have several causes. The absence of specific protein binding at the USF or initiator sites could result, directly or indirectly, in altered function of some component of the transcription complex. Such components could include TFIIH, which contains helicase and kinase activities (54) necessary for early steps in initiation (55–59), TFIIE, which has been implicated in transcription initiation (60,61), or pol II itself. Other activities might also be affected, such as the elongation factor SII, which is known to be sarkosyl sensitive (47). This sensitivity might be altered by protein–protein interactions within the preinitiation complex.
While USF is likely to be the protein whose specific association with the transcription complex is lost in the 41 template, it is not clear what functional DNA–protein interactions might be disrupted in the Inrmut template. A number of proteins have been reported to recognize initiator elements, including YY1 (31), HIP1/E2F (32,33), TFII-I (34), USF (35) and TFIID (36–38). However, there is no general correlation between YY1 (39) or HIP1/E2F (39,62) binding with initiator activity. Furthermore, a degenerate initiator consensus sequence derived from the analysis of mutated initiator elements [PyPyAN(T/A)PyPy, with initiation at the A at position 3] does not fit well with consensus USF and TFII-I binding sites (39,40), although both of these proteins recognize the AdML promoter initiator and can affect transcription by binding to this site (34,35,63). It has been proposed (39) that some initiators, such as that from the AdML promoter, could interact with TFIID, the binding of which is strictly dependent on the most conserved nucleotides in the initiator consensus (36,37,40), as well as promoter-specific factors such as USF or TFII-I. The mutations introduced into the Inrmut template (Fig. 1A) change two pyrimidines in the initiator consensus to purines, but they also affect nucleotides that are part of the E-box binding site for USF and TFII-I, which even in the wild-type promoter is not a particularly good match to the consensus (CACGTG; 64).
Transcription from AdML and MMTV promoter templates is differentially sensitive to sarkosyl
Two obvious differences distinguish the behavior of templates containing AdML and MMTV promoters in our experiments. First, RNA synthesis on a template containing the AdML promoter is significantly more sensitive to sarkosyl, as seen in the lower concentration of sarkosyl required to inhibit RNA synthesis from the wild-type AdML template (Fig. 3), as well as in the effect of sarkosyl on the kinetics of RNA synthesis (Fig. 4). In addition, the kinetics of the sarkosyl challenge assay were too rapid to measure in all of the experiments with MMTV promoter templates (Fig. 6C). While it is tempting to speculate that such differences reflect inherent properties of these promoters, it is also possible that they are related to sequence-specific effects of the G-free (AdML) or T-free (MMTV) cassettes present on the templates. At present we cannot distinguish these possibilities.
Effect of transcriptional activators on steps subsequent to preinitiation complex assembly
Many experiments support the idea that pol II transcriptional activators stimulate transcription complex assembly in vivo (5) and in vitro (4,65). However, assembly is not the sole determinant of transcriptional efficiency. Quantitation of transcription complex assembly independent of RNA synthesis has shown that differences in promoter strength can result from differences in initiation or promoter clearance (65). Transcription can also be limited by elongation efficiency. Nuclear run-on experiments with a number of genes have shown that promoter-proximal sequences have a much higher density of elongating polymerase than promoter-distal sequences in vivo (7,66–69) and the apparent processivity of elongation can be increased by some, but not all, transcriptional activators (66–68,70). Several activator proteins have been demonstrated to fall into one of three classes on the basis of whether their primary effect is on initiation/assembly, elongation or both, and the ability of an activator to affect elongation was shown to correlate with its ability to interact with TFIIH in vitro (70). For at least some of the genes in which elongation appears to be a regulated event, pol II does not proceed beyond ~25–30 nt from the initiation site in the absence of activation (6,67), suggesting that steps related to promoter clearance might be targets for regulation. Indeed, several in vitro studies have shown that pol II does not efficiently polymerize >10–15 nt in the absence of ATP and TFIIH (56,58,61,71) and, in some cases, phosphorylation of the C-terminal domain of the largest subunit of pol II has been implicated as important for transcription of promoter-distal sequences (72–75). None of the transcriptional regulators recognized as being important for transcription from the AdML or MMTV promoters have been included in any of these studies and it is therefore difficult to determine whether our observations are mechanistically related to some of these published results.
Sarkosyl has figured prominently in many of the studies dealing with regulation of elongation. Stalled polymerases near the 5'-end of the Drosophila hsp70 gene could not be detected in the absence of sarkosyl or high salt (7), both of which could strip away histones and non-specific DNA-binding proteins from the template and also inhibit the activity of elongation factors (48). In addition, sarkosyl or sarkosyl-washed transcription complexes have been used to define promoter-proximal limits to elongation on the c-fos gene (69), as well as the TFIIH-dependent release of paused polymerases mediated by Gal4–VP16 (58). Thus, sarkosyl dependence of some of the effects we observe does not distinguish our results from many other experiments that have addressed similar questions. While the specific effects of sarkosyl may not be directly relevant in vivo, the differential sensitivity of transcription complexes to inhibitors like sarkosyl is clearly revealing altered functional properties, which are likely to have an impact on transcription under physiological conditions.
|
ACKNOWLEDGEMENTS |
|---|
We thank Katherine Beifuss for construction of the plasmid pBML
12. We are also indebted to Robert Roeder for generously providing the plasmid pML(C2AT)19 and Lisa Dandridge for critical reading of the manuscript. This work was partially supported by Public Health Service Grant R01 CA32695 from the National Cancer Institute. The support of the Texas Agricultural Experiment Station is also gratefully acknowledged.
|
FOOTNOTES |
|---|
* To whom correspondence should be addressed. Tel: +1 979 845 0953; Fax +1 979 845 9274; Email: dopeterson@tamu.edu Present addresses: John W. Steinke, University of Virginia, PO Box 801355, Charlottesville, VA 22908-1355, USA Stephan J. Kopytek, Department of Chemistry, Havemeyer Hall, MC 3111, Columbia University, New York, NY 10027, USA
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONREFERENCES |
|---|
-
1 Lin,Y.-S. and Green,M.R. (1991) Cell, 64, 971–981.[Web of Science][Medline]
2 Kim,T.K. and Roeder,R.G. (1994) Proc. Natl Acad. Sci. USA, 91, 4170–4174.
3 Choy,B. and Green,M.R. (1993) Nature, 366, 531–536.[Medline]
4 Gaudreau,L., Adam,M. and Ptashne,M. (1998) Mol. Cell, 1, 913–916.[Web of Science][Medline]
5 Keaveney,M. and Struhl,K. (1998) Mol. Cell, 1, 917–924.[Web of Science][Medline]
6 O’Brien,T. and Lis,J.T. (1991) Mol. Cell. Biol., 11, 5285–5290.
7 Rougvie,A.E. and Lis,J.T. (1988) Cell, 54, 795–804.[Web of Science][Medline]
8 Jones,K.A. and Peterlin,B.M. (1994) Annu. Rev. Biochem., 63, 717–743.[Web of Science][Medline]
9 Jacob,G.A., Luse,S.W. and Luse,D.S. (1991) J. Biol. Chem., 266, 22537–22544.
10 Smale,S.T. and Baltimore,D. (1989) Cell, 57, 103–113.[Web of Science][Medline]
11 Sawadogo,M. and Roeder,R.G. (1985) Cell, 43, 165–175.[Web of Science][Medline]
12 Chodosh,L.A., Carthew,R.W. and Sharp,P.A. (1986) Mol. Cell. Biol., 6, 4723–4733.
13 Sawadogo,M., Van Dyke,M.E., Gregor,P.D. and Roeder,R.G. (1988) J. Biol. Chem., 263, 11985–11993.
14 Ziff,E.B. (1980) Nature, 287, 491–499.[Medline]
15 Toohey,M.G., Lee,J.W., Huang,M. and Peterson,D.O. (1990) J. Virol., 64, 4477–4488.
16 Pierce,J., Fee,B.E., Toohey,M.G. and Peterson,D.O. (1993) J. Virol., 67, 415–424.
17 Günzburg,W.H. and Salmons,B. (1992) Biochem. J., 283, 625–632.
18 Lefebvre,P., Berard,D.S., Cordingley,M.G. and Hager,G.L. (1991) Mol. Cell. Biol., 11, 2529–2537.
19 Yanagawa,S., Tanaka,H. and Ishimoto,A. (1991) J. Virol., 65, 526–531.
20 Morley,K.L., Toohey,M.G. and Peterson,D.O. (1987) Nucleic Acids Res., 15, 6973–6989.
21 Lee,J.W., Moffitt,P.G., Morley,K.L. and Peterson,D.O. (1991) J. Biol. Chem., 266, 24101–24108.
22 Ostrowski,M.C., Huang,A.L., Kessel,M., Wolford,R.G. and Hager,G.L. (1984) EMBO J., 3, 1891–1900.[Web of Science][Medline]
23 Sawadogo,M. and Roeder,R.G. (1985) Proc. Natl Acad. Sci. USA, 82, 4394–4398.
24 Chen,E.Y. and Seeburg,P.H. (1985) DNA, 4, 165–170.[Web of Science][Medline]
25 Bral,C.M., Steinke,J.W., Kang,C.-J. and Peterson,D.O. (1998) Gene Expr., 7, 191–204.[Web of Science][Medline]
26 Birnboim,H.C. and Doly,J. (1979) Nucleic Acids Res., 7, 1513–1523.
27 Dignam,J.D., Martin,P.L., Shastry,B.S. and Roeder,R.G. (1983) Methods Enzymol., 101, 582–598.
28 Kuo,W.L., Vilander,L.R., Huang,M. and Peterson,D.O. (1988) J. Virol., 62, 2394–2402.
29 Kang,C.-J. and Peterson,D.O. (1999) Virology, 264, 211–219.[Web of Science][Medline]
30 Bradford,M.M. (1976) Anal. Biochem., 72, 248–254.[Web of Science][Medline]
31 Seto,E., Shi,Y. and Shenk,T. (1991) Nature, 354, 241–245.[Medline]
32 Means,A.L. and Farnham,P.J. (1990) Mol. Cell. Biol., 10, 653–661.
33 Blake,M.C. and Azizkhan,J.C. (1989) Mol. Cell. Biol., 9, 4994–5002.
34 Roy,A.L., Du,H., Gregor,P.D., Novina,C.D., Martinez,E. and Roeder,R.G. (1997) EMBO J., 16, 7091–7104.[Web of Science][Medline]
35 Du,H., Roy,A.L. and Roeder,R.G. (1993) EMBO J., 12, 501–511.[Web of Science][Medline]
36 Kaufmann,J. and Smale,S.T. (1994) Genes Dev., 8, 821–829.
37 Purnell,B.A., Emanuel,P.A. and Gilmour,D.S. (1994) Genes Dev., 8, 830–842.
38 Verrijzer,C.P., Yokomori,K., Chen,J.-L. and Tjian,R. (1994) Science, 264, 933–941.
39 Javahery,R., Khachi,A., Lo,K., Zenzie-Gregory,B. and Smale,S.T. (1994) Mol. Cell. Biol., 14, 116–127.
40 Lo,K. and Smale,S.T. (1996) Gene, 182, 13–22.[Web of Science][Medline]
41 Kim,M.H. and Peterson,D.O. (1995) J. Virol., 69, 4717–4726.[Abstract]
42 Kim,M.H. and Peterson,D.O. (1995) J. Biol. Chem., 270, 27823–27828.
43 Hawley,D.K. and Roeder,R.G. (1987) J. Biol. Chem., 262, 3452–3461.
44 Szentirmay,M.N. and Sawadogo,M. (1994) Nucleic Acids Res., 22, 5341–5346.
45 Hawley,D.K. and Roeder,R.G. (1985) J. Biol. Chem., 260, 8163–8172.
46 Cai,H. and Luse,D.S. (1987) J. Biol. Chem., 262, 298–304.
47 Wiest,D.K., Wang,D. and Hawley,D.K. (1992) J. Biol. Chem., 267, 7733–7744.
48 Kephart,D.D., Marshall,N.F. and Price,D.H. (1992) Mol. Cell. Biol., 12, 2067–2077.
49 Toohey,M.G. and Jones,K.A. (1989) Genes Dev., 3, 265–282.
50 Marshall,N.F. and Price,D.H. (1992) Mol. Cell. Biol., 12, 2078–2090.
51 Wang,W., Carey,M. and Gralla,J.D. (1992) Science, 255, 450–453.
52 Tantin,D. and Carey,M. (1994) J. Biol. Chem., 269, 17397–17400.
53 Dahmus,M.E. (1996) J. Biol. Chem., 271, 19009–19012.
54 Seroz,T., Hwang,J.R., Moncollin,V. and Egly,J.M. (1995) Curr. Opin. Genet. Dev., 5, 217–221.[Medline]
55 Dvir,A., Garrett,K.P., Chalut,C., Egly,J.M., Conaway,J.W. and Conaway,R.C. (1996) J. Biol. Chem., 271, 7245–7248.
56 Dvir,A., Conaway,R. and Conaway,J. (1997) Proc. Natl Acad. Sci. USA, 94, 9006–9010.
57 Dvir,A., Tan,S., Conaway,J.W. and Conaway,R.C. (1997) J. Biol. Chem., 272, 28175–28178.
58 Kumar,K.P., Akoulitchev,S. and Reinberg,D. (1998) Proc. Natl Acad. Sci. USA, 95, 9767–9772.
59 Moreland,R.J., Tirode,F., Yan,Q., Conaway,J.W. and Conaway,R.C. (1999) J. Biol. Chem., 274, 22127–22130.
60 Goodrich,J.A. and Tjian,R. (1994) Cell, 77, 145–156.[Web of Science][Medline]
61 Kugel,J.F. and Goodrich,J.A. (1998) Proc. Natl Acad. Sci. USA, 95, 9232–9237.
62 Kollmar,R. and Farnham,P.J. (1993) Proc. Soc. Exp. Biol. Med., 203, 127–139.[Medline]
63 Roy,A.L., Meisterernst,M., Pognonec,P. and Roeder,R.G. (1991) Nature, 354, 245–248.[Medline]
64 Baxevanis,A.D. and Vinson,C.R. (1993) Curr. Opin. Genet. Dev., 3, 278–285.[Medline]
65 Jacob,G.A., Kitzmiller,J.A. and Luse,D.S. (1994) J. Biol. Chem., 269, 3655–3663.
66 Yankulov,K., Blau,J., Purton,T., Roberts,S. and Bentley,D.L. (1994) Cell, 77, 749–759.[Web of Science][Medline]
67 Krumm,A., Hickey,L.B. and Groudine,M. (1995) Genes Dev., 9, 559–572.
68 Bentley,D.L. (1995) Curr. Opin. Genet. Dev., 5, 210–216.[Medline]
69 Pinaud,S. and Mirkovitch,J. (1998) J. Mol. Biol., 280, 785–798.[Web of Science][Medline]
70 Blau,J., Xiao,H., McCracken,S., O’Hare,P., Greenblatt,J. and Bentley,D. (1996) Mol. Cell. Biol., 16, 2044–2055.[Abstract]
71 Dvir,A., Conaway,R.C. and Conaway,J.W. (1996) J. Biol. Chem., 271, 23352–23356.
72 Akhtar,A., Faye,G. and Bentley,D.L. (1996) EMBO J., 15, 4654–4664.[Web of Science][Medline]
73 Marshall,N.F., Peng,J.M., Xie,Z. and Price,D.H. (1996) J. Biol. Chem., 271, 27176–27183.
74 Parada,C.A. and Roeder,R.G. (1996) Nature, 384, 375–378.[Medline]
75 Song,C.Z. (1996) Biochem. Biophys. Res. Commun., 229, 810–816.[Web of Science][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  H. Tang, M. B. Veldman, and D. Goldman Characterization of a Muscle-specific Enhancer in Human MuSK Promoter Reveals the Essential Role of Myogenin in Controlling Activity-dependent Gene Regulation J. Biol. Chem., February 17, 2006; 281(7): 3943 - 3953. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||