MOLECULAR AND CELLULAR BIOLOGY, May 2008, p. 3290–3300
0270-7306/08/$08.00⫹0 doi:10.1128/MCB.02224-07
Copyright © 2008, American Society for Microbiology. All Rights Reserved.
Vol. 28, No. 10
NELF and GAGA Factor Are Linked to Promoter-Proximal Pausing
at Many Genes in Drosophila䌤†
Chanhyo Lee,1 Xiaoyong Li,2 Aaron Hechmer,2 Michael Eisen,2,3 Mark D. Biggin,2 Bryan J. Venters,1
Cizhong Jiang,1 Jian Li,1 B. Franklin Pugh,1 and David S. Gilmour1*
Center for Gene Regulation, Department of Biochemistry and Molecular Biology, The Pennsylvania State University,
University Park, Pennsylvania 168021; Genomics Division, Lawrence Berkeley National Laboratory, Berkeley,
California 947202; and Center for Integrative Genomics, Department of Molecular and Cell Biology,
University of California, Berkeley, California 947203
Received 15 December 2007/Returned for modification 8 January 2008/Accepted 26 February 2008
immunoprecipitation (ChIP) experiments first detected Pol II
in the promoter region of this gene prior to induction (19), and
subsequent studies established that Pol II initiates transcription but pauses in the region 20 to 40 nucleotides downstream
from the transcription start site (17, 48, 50). NELF and DSIF
cause Pol II to pause in the promoter-proximal region of hsp70
(59, 61).
Induction of hsp70 relies on two proteins, GAGA factor and
HSF, that bind upstream from the core promoter region (35).
GAGA factor functions before heat shock in establishing the
paused Pol II and a promoter architecture conducive to binding by HSF (31, 52). Heat shock rapidly induces HSF to associate with hsp70 (5). This leads to P-TEFb being recruited to
hsp70, thus providing a mechanism for alleviating NELF-mediated inhibition (34). Notably, the HSF activation domain
appears to be unable to activate transcription without the prior
action of the GAGA factor (58). Thus, induction of hsp70
relies on two types of sequence-specific DNA binding proteins,
namely, GAGA factor, to promote initiation and establish the
transcriptional potential of the gene, and HSF, to promote
both release of transcriptionally engaged Pol II and rapid reinitiation.
Recent genome-wide analyses of the distribution of Pol II
indicate that Pol II is concentrated in the promoter regions of
thousands of genes, even though many of these genes appear
to be transcribed infrequently or quiescent (23, 38, 66). Hence,
transcriptional control targeting a step after initiation may be
far more common than previously known. To increase our
understanding of promoter-proximal pausing, we analyzed the
distributions of NELF and GAGA factor in Drosophila cells by
using ChIP-microarray chip (ChIP-chip) analysis. We also examined the association of Pol II at numerous promoters by
Based on studies of the transcription inhibitor 5,6-dichloro-1␤-D-ribobenzimidazole (DRB), it was proposed that transcription
elongation by RNA polymerase II (Pol II) in animal cells is
controlled a short distance downstream from the transcription
start site by the opposing actions of positive and negative elongation factors, dubbed P-TEF and N-TEF, respectively (36). Biochemical analyses resulted in a model for regulation of transcription elongation in the promoter-proximal region (42). Acting as
N-TEFs, the proteins NELF and DSIF associate with the Pol II
elongation complex to inhibit elongation (57, 64). This inhibition
is overcome by a kinase called P-TEFb (37). P-TEFb phosphorylates Pol II, DSIF, and NELF to alleviate the inhibitory action of
NELF and DSIF, thus allowing Pol II to productively transcribe
a gene. Inhibition of transcription elongation by DRB is currently
ascribed to its ability to inhibit P-TEFb. P-TEFb, DSIF, and
NELF are likely to affect transcription of many genes, since inhibitors of P-TEFb repress transcription of over half of the genes
in animal cells (10, 30).
The vast majority of studies of transcriptional regulatory
mechanisms have emphasized the importance of proteins involved in the recruitment of Pol II to a promoter. Early studies
with animal cells, however, provided evidence that transcriptional regulation of some genes occurs at a stage after Pol II
initiates transcription (51, 55). The best-studied example is
hsp70 in Drosophila melanogaster (33). hsp70 is rapidly and
dramatically induced in response to heat shock. Chromatin
* Corresponding author. Mailing address: Penn State University,
403 S. Frear, University Park, PA 16802. Phone: (814) 863-8905. Fax:
(814) 863-7024. E-mail: dsg11@psu.edu.
† Supplemental material for this article may be found at http://mcb
.asm.org/.
䌤
Published ahead of print on 10 March 2008.
3290
Downloaded from mcb.asm.org at Penn State Univ on October 23, 2009
Recent analyses of RNA polymerase II (Pol II) revealed that Pol II is concentrated at the promoters of many
active and inactive genes. NELF causes Pol II to pause in the promoter-proximal region of the hsp70 gene in
Drosophila melanogaster. In this study, genome-wide location analysis (chromatin immunoprecipitation-microarray
chip [ChIP-chip] analysis) revealed that NELF is concentrated at the 5ⴕ ends of 2,111 genes in Drosophila cells.
Permanganate genomic footprinting was used to determine if paused Pol II colocalized with NELF. Forty-six of 56
genes with NELF were found to have paused Pol II. Pol II pauses 30 to 50 nucleotides downstream from transcription start sites. Analysis of DNA sequences in the vicinity of paused Pol II identified a conserved DNA sequence that
probably associates with TFIID but detected no evidence of RNA secondary structures or other conserved sequences
that might directly control elongation. ChIP-chip experiments indicate that GAGA factor associates with 39% of the
genes that have NELF. Surprisingly, NELF associates with almost one-half of the most highly expressed genes,
indicating that NELF is not necessarily a repressor of gene expression. NELF-associated pausing of Pol II might be
an obligatory but sometimes transient checkpoint during the transcription cycle.
VOL. 28, 2008
PROMOTER-PROXIMAL PAUSING IN DROSOPHILA
3291
using permanganate genomic footprinting to assess the relationship between NELF, GAGA factor, and promoter-proximal pausing. Our findings link NELF and GAGA factor to
promoter-proximal pausing at many genes and suggest that
NELF-directed pausing of Pol II may represent a widespread
checkpoint during the transcription cycle.
MATERIALS AND METHODS
Microarray analysis of DNA isolated by ChIP. ChIP-chip analyses of NELF-B,
NELF-E, and GAGA factor were carried out with Schneider 2 cells. DNA
cross-linked to proteins was isolated by immunoprecipitation as previously described (59).
The NELF-B and -E samples were analyzed by the Berkeley Drosophila
Transcription Network Project. The ChIP and control ChIP DNA samples, along
with DNAs purified from input chromatin samples, were amplified using a
modified random prime-based PCR amplification protocol. The amplified DNAs
were fragmented with DNase I, biotinylated, and hybridized to Affymetrix Drosophila Tiling 1.0F arrays. The ChIP-chip data were analyzed using TiMAT
programs developed by the Berkeley Drosophila Transcription Network Project
(http://bdtnp.lbl.gov/TiMAT/TiMAT2/index.html). Peaks of NELF were located
using Mpeak (27). All parameters for Mpeak were set at the default, except for
“largest search range,” which was reduced to 1.0 kb. This change was made
because the sonicated DNAs generated for ChIP were in the size range of 0.3 to
1.0 kb.
The GAGA factor ChIP DNA was analyzed by the Penn State Cartography
Project. DNA cross-linked to GAGA factor was isolated by immunoprecipitation
as previously described (59). The GAGA factor ChIP and mock ChIP (preimmune GAGA factor serum) DNA samples were amplified by ligation-mediated
Downloaded from mcb.asm.org at Penn State Univ on October 23, 2009
FIG. 1. Comparison of the distributions of NELF and paused Pol II. In each panel, the bar graphs for NELF-E and NELF-B represent ChIP
signals at individual probe coordinates. Regions lacking bars represent no statistically significant signal above the preimmune control signal. The
MnO4⫺ graph in each panel is derived from SAFA analysis (13) of the permanganate footprinting data shown in Fig. 3. The enhancement in
permanganate reactivity that occurred in cells was determined by subtracting the band intensities for naked DNA from the corresponding band
intensities obtained for treated cells. The regions analyzed by permanganate footprinting are limited to regions of approximately 100 nucleotides.
The graphs were aligned with each other and with an annotated view of the corresponding region of the genome by using the UCSC browser.
Numbers at the top of each panel designate coordinates on the chromosome indicated to the left of the panel. CG14709, CG5807, and CG31531
are examples where the densities of NELF are significantly elevated in a 500- to 1,000-base-pair region at the 5⬘ end of a gene. Bap170, Cctgamma,
and Nipsnap are examples where the densities of NELF are above a 1% false discovery rate but evenly distributed over 2 to 3 kb; no permanganate
footprint was evident in these promoter regions (Fig. 3). The solid horizontal bars delineate transcription units, and the arrowheads indicate the
direction of transcription. Results for 59 regions are provided in File S1 in the supplemental material.
3292
MOL. CELL. BIOL.
LEE ET AL.
RESULTS
Peaks of NELF are detected near the promoters of 2,111
genes. High-density Affymetrix tiling microarrays were used to
analyze two independent DNA samples (biological replicates)
that had been isolated by ChIP from Drosophila cells with
antisera against the B and E subunits of NELF. Figure 1 shows
three cases where NELF is concentrated near the transcription
start site and three cases where it is more evenly distributed—
inspection of all of the data reveals a continuum between these
extremes (see Files S2 and S3 in the supplemental material).
NELF-B and NELF-E were each detected in over 4,000 separate regions, with a 1% false discovery rate. At least 80% of
the NELF-B and NELF-E regions overlap with each other by
a minimum of 1 kb, consistent with the two functioning as part
of a complex (Fig. 2A). The Mpeak algorithm (27) was used to
identify regions with concentrated peaks of NELF. Most peaks
of NELF-B and NELF-E mapped to within 500 base pairs of
an annotated transcription start site (Fig. 2B). A total of 2,111
genes were found to have peaks of both NELF-B and NELF-E
within 500 base pairs of the start site (see Fig. 5C; see also File
S6 in the supplemental material).
Inspection of the peaks located farther than 500 base pairs
FIG. 2. Analysis of distribution of NELF. (A) Venn diagram showing the number of NELF-E regions that overlap NELF-B regions by at
least 1 kb and visa versa. Regions correspond to places where contiguous probes on the microarray yielded signals above a 1% false discovery rate and include regions that contain peaks as well as low even
distributions, as shown in Fig. 1. The P value for the chance occurrence
of this overlap is ⬍10⫺300. The number not in parentheses (3,829) is
the number of NELF-B regions overlapping NELF-E regions. The
number in parentheses (3,934) is the number of NELF-E regions
overlapping NELF-B regions. In some cases, two separate regions for
one protein overlap one region for the other protein. (B) Positions of
NELF peaks relative to transcription start sites. NELF-B (n ⫽ 2,624)
and NELF-E (n ⫽ 2,987) peaks were located using Mpeak. Note that
there are fewer peaks than regions (A) because many regions lack a
discernible peak of NELF.
from a start site did not reveal any obvious relationship to
other aspects of genes, such as sites of RNA processing. Recently, NELF was shown to be involved in 3⬘-end processing of
the nonpolyadenylated histone mRNAs (41). Since repeated
genes like histone genes were not included on the microarray,
we could not analyze NELF’s relationship to these genes. For
single-copy sequences, there was no indication that NELF
peaked near the 3⬘ ends of genes, although low levels of NELF
can be detected at the ends of some genes.
Paused Pol II is frequently located near peaks of NELF. To
determine if peaks of NELF contained paused Pol II, we experimentally interrogated a subset of cases by using permanganate genomic footprinting (8). The genes were chosen based
on inspection of the ChIP-chip data, selecting mainly NELFenriched genes but also several NELF-deficient controls. Permanganate reacts preferentially with thymines in the singlestranded transcription bubble associated with transcriptionally
engaged Pol II. Since some thymines in double-stranded DNA
Downloaded from mcb.asm.org at Penn State Univ on October 23, 2009
PCR (24). Using a GeneChip WT (whole transcript) double-stranded DNA
terminal labeling kit (Affymetrix), the amplified DNA containing dUTP was
fragmented with uracil DNA glycosylase and apyrimidinic/apurinic endonuclease, biotinylated, and finally hybridized to Affymetrix Drosophila Tiling 1.0R
arrays. Signal analysis, interval analysis, and peak calling were performed using
software for model-based analysis of tiling arrays (26). A bandwidth and maximum gap value of 150 bp was used for the interval analysis. Peaks were called
using a 1% false discovery rate threshold.
Analysis of tabulated information. The probe intensities from the microarray
analysis were obtained in tabulated form (see Files S2, S3, and S4 in the supplemental material). Analysis of tabulated data was done using tools provided at
the Penn State Galaxy website (http://g2.trac.bx.psu.edu/) to evaluate the relationship between the locations of protein peaks and transcription start sites (see
Fig. 2) and to identify genes associated with protein peaks (see Fig. 5 and 7; also
see File S6 in the supplemental material). Galaxy was also used to prepare and
transfer tabulated data for viewing in the UCSC browser (http://genome.ucsc
.edu/). The chi-square test function in Excel was used to calculate P values
associated with the Venn diagrams (65).
Permanganate genomic footprinting. Drosophila S2 cells (serum-free medium
[SFM]-adapted D.mel-2 cells; Invitrogen) (2 ml) in a six-well plate containing
Drosophila-SFM supplemented with 90 ml/liter of 200 mM L-glutamine were
grown to a density of approximately 5 ⫻ 106 cells/ml at 25°C. The medium was
removed, and the plate was placed on ice. Cells were treated with permanganate
by adding 800 ␮l of ice-cold 10 mM KMnO4 dissolved in phosphate-buffered
saline (PBS) and incubating them for 60 seconds. The reactions were stopped by
the addition of 200 ␮l of 5⫻ stop solution (50 mM Tris-Cl, pH 7.5, 50 mM NaCl,
50 mM Na2EDTA, pH 8.0, 2.5% sodium dodecyl sulfate, 1 M 2-mercaptoethanol). The cell lysate was transferred to a 1.5-ml tube and then incubated overnight at 37°C with 50 ␮g of proteinase K. Samples were extracted twice with
phenol-chloroform-isoamyl alcohol (25:24:1), and the permanganate cleavage
pattern was analyzed as previously described (8).
To prepare permanganate-treated naked DNA and GA markers, cells were
treated with 800 ␮l of PBS followed by 200 ␮l of 5⫻ stop solution. DNA was
purified as described above. One microgram of DNA was diluted to a final
volume of 100 ␮l with cold PBS. DNA samples were treated with 100 ␮l of 20
mM KMnO4 in PBS on ice. Individual samples were incubated with permanganate on ice for 0, 30, or 60 seconds and then stopped by the addition of 200 ␮l
of 2⫻ stop solution. DNA was ethanol precipitated and dissolved in 90 ␮l of
water. GA markers were generated as previously described (8).
Primers and annealing temperatures used in the ligation-mediated PCR analyses are listed in File S8 in the supplemental material.
Microarray data accession number. The raw and processed microarray data
were submitted to ArrayExpress (www.ebi.ac.uk/arrayexpress/) under accession
number E-MEXP-1547.
VOL. 28, 2008
PROMOTER-PROXIMAL PAUSING IN DROSOPHILA
3293
are hyperreactive in the absence of protein, the patterns of
permanganate reactivity occurring in cells were compared to
the patterns of reactivity occurring in purified DNA. Figure 3
shows results for the six regions presented in Fig. 1, and results
for a total of 59 different regions, along with ChIP-chip data,
are provided in File S1 in the supplemental material. Forty-six
of the 59 regions analyzed had a clearly discernible permanganate footprint, and all 46 footprints were located within 500
base pairs of a NELF peak. This is clearly exemplified by
aligning the pattern of permanganate reactivity with the ChIPchip data (Fig. 1; see File S1 in the supplemental material) and
suggests that paused Pol II colocalizes with NELF.
Interestingly, we found 10 cases of those interrogated where
no permanganate footprint was evident even though NELF
was present (Bap170, Cctgamma, and Nipsnap are examples
shown in Fig. 1, and the remainder are shown in File S1 in the
supplemental material). ChIP-chip data for Pol II from the
work of Muse et al. (38) show that various levels of Pol II are
present at the promoter regions of all 10 of these genes. A
possible explanation for this discrepancy is that Pol II does not
reach a sufficient concentration in a restricted region of these
genes to be detected with permanganate.
Transcription bubbles are concentrated in the region 20 to
50 nucleotides downstream from an initiator element. The
locations of the permanganate footprints were analyzed because these locations could have important implications concerning the mechanism of the pause. When we aligned permanganate footprints relative to annotated transcription start
sites, designated by thick blue vertical lines in Fig. 4A, the edge
of each footprint closest to the start site was at least 15 nucleotides downstream from the start site for 36 of 46 cases. A
distance of 15 nucleotides corresponds to Pol II molecules
having transcribed at least 26 nucleotides, given that the size of
the transcription bubble is 12 nucleotides and the 3⬘ end of the
nascent transcript is located 1 nucleotide behind the leading
edge of the bubble (21, 28).
The locations of the permanganate footprints for several
cases seemed unusual. The annotated start sites of CG6014,
Downloaded from mcb.asm.org at Penn State Univ on October 23, 2009
FIG. 3. Permanganate genomic footprints for selected genes. Drosophila cells were subjected to permanganate genomic footprinting, and the
resulting patterns of reactivity were analyzed on DNA sequencing gels. The rightmost lane in each panel shows the pattern of reactivity occurring
in cells; the leftmost lane shows GA markers. The center lanes show controls done with purified DNA (naked DNA). The numbers to the right
of each panel identify locations of thymines relative to the transcription start sites annotated in the Drosophila genome. Some thymines have not
been marked to simplify the display; all are labeled in File S1 in the supplemental material. The thymines deemed hyperreactive are printed in
black; the rest are shown in gray. The short vertical lines adjacent to CG14709 (at positions ⫺22), CG5807 (between positions ⫹30 and ⫹45), and
CG31531 (at position ⫹3) delineate initiator elements identified in Fig. 4. The footprinting results are for those genes displayed in Fig. 1.
Footprinting results for an additional 53 loci are provided in File S1 in the supplemental material.
MOL. CELL. BIOL.
LEE ET AL.
3294
Downloaded from mcb.asm.org at Penn State Univ on October 23, 2009
PROMOTER-PROXIMAL PAUSING IN DROSOPHILA
3295
corto, LanA, and rpr were located within the permanganate
footprint (Fig. 4A, bottom 4 rows; note the locations of the
annotated start site [in blue] and the permanganate reactive
sites [in black]). A search for conserved sequence elements in
the vicinity of the permanganate footprints, however, revealed
that the unusual cases could be explained by the presence of
nearby transcription start sites that had not yet been identified.
The program MEME (3) identified conserved nucleotides
within a 35-nucleotide region in 28 of 46 of the footprints (Fig.
4B; see File S7 in the supplemental material) (genes with this
conserved motif are marked with asterisks in Fig. 4A). The
conserved nucleotides in the regions designated ⫺2 to 4 and 28
to 32 resemble the previously identified initiator element (Inr)
and downstream promoter element (DPE), which are recognized by TFIID (7, 44, 54) and are shown in Fig. 4C. In
Drosophila, transcription usually begins at or near the conserved adenine in the Inr (16). We surmise that the unusually
positioned permanganate footprints resulted from Pol II molecules that used a nearby Inr to initiate transcription. In support of this, we confirmed that transcription initiates in vitro
within the Inr of CG14709 (data not shown), and the transcription start site for Tollo was previously mapped to the Inr (32).
To measure where Pol II typically resides when paused on a
promoter, we plotted the frequency of reactive thymines as a
function of the distance from the adenine in the initiator. The
start site proximal edge of the composite footprint begins approximately 20 nucleotides from the transcription start site and
extends for approximately 30 nucleotides (Fig. 4D).
GAGA factor frequently colocalizes with paused Pol II and
NELF. Our MEME analysis of sequences encompassing the
permanganate footprints identified another conserved sequence called the GAGA element (Fig. 5A). Twenty-seven of
the 46 promoters with permanganate footprints contained at
least one GAGA element within regions spanning from 200 bp
upstream to 100 nucleotides downstream of the Inr elements
shown in Fig. 4A (designated with the symbol “#”). GAGA
factor has been implicated in promoter-proximal pausing on
hsp70 (31, 52, 58).
To assess the relationship between GAGA factor, NELF,
and paused Pol II, we determined GAGA factor’s distribution
across the entire genome by using ChIP-chip analysis. GAGA
factor was detected at 2,907 regions, with a false discovery rate
of 1%. Figure 5B displays examples of the locations of GAGA
factor along with the locations of NELF-B and permanganate
footprints, and the remaining cases that were also analyzed
with permanganate are provided in File S1 in the supplemental
material. We selected tai as one example because this region is
approximately 9 kb downstream from the annotated start site.
The presence of NELF, GAGA factor, paused Pol II, and a
conserved initiator element indicates that this is probably a
promoter. Like its mammalian counterpart (43), D-fos has
paused Pol II, and this could be linked to the presence of
GAGA factor. Most of the places where we detected paused
Pol II by permanganate footprinting (40/46 cases) were within
1 kb of a peak of GAGA factor.
To assess the potential link between GAGA factor and
NELF, we compared the genes with a peak of GAGA factor
within 1 kb of the transcription start site to genes with NELF-B
and NELF-E peaks within 500 base pairs of the transcription
start (Fig. 5C). A total of 2,111 genes were found to associate
with both NELF-B and NELF-E, and 39% of these genes were
found to associate with GAGA factor. Thus, GAGA factor is
detected near a significant number of NELF-associated genes
(P ⬍ 10⫺300).
The relationship between NELF, pausing, and gene expression levels. Since NELF inhibits transcription elongation, we
anticipated that it might correlate with transcriptional repression. To address this possibility, we ranked approximately
14,000 genes according to their previously determined levels of
expression (49) and binned them in 10 percentile increments.
Next, the percentage of genes in each bin that had NELF
located within 500 bp was determined. The frequency of genes
associated with NELF increased as the level of expression
increased (Fig. 6). Surprisingly, almost half of the genes expressed above the 90th percentile appeared to associate with
NELF. In accordance with this trend, 22 of 46 genes exhibiting
a permanganate footprint were ranked in the top quartile for
expression, while only 3 of the 46 were ranked in the bottom
quartile (data not shown). Given that NELF is a negative
regulator of elongation and its association correlates with
pausing, these results indicate that pausing could be an integral
but transient part of the transcription cycle.
DISCUSSION
Approximately 2,000 genes have NELF at the 5ⴕ end, and
many of these genes are likely to have paused Pol II. We
identified 2,111 genes in the Drosophila genome that had peaks
FIG. 4. Sequence conservation and alignment of permanganate footprints with initiator elements. (A) Schematics of regions encompassing
permanganate footprints were aligned relative to putative initiator elements. The putative initiator elements matched at least four of five positions
of previously characterized Drosophila initiator elements (9, 32). Those tagged with an asterisk contain the conserved motif shown in panel B.
Those tagged with the symbol “#” contain the conserved GAGA element displayed in Fig. 5A. Three of the promoters (scyl, ␤1-tubulin, and hsp26)
lack the conserved initiator, so they were aligned on the annotated start site. The annotated start sites from FlyBase are marked with thick blue
vertical lines; note that several cases do not align with an initiator element. Those cases lacking a thick blue line have annotated transcription start
sites that fall outside the illustrated region. Black vertical lines mark the locations of hyperreactive thymines that constitute the permanganate
footprint, and red vertical lines mark the positions of the remaining thymines. (B) Conserved element detected in the vicinity of the permanganate
footprints (MnO4⫺ footprint-linked conservation). Sequence conservation was based on the MEME results shown in File S7 in the supplemental
material and displayed using WebLogo (http://weblogo.berkeley.edu/). P values for regions containing this motif were 5 ⫻ 10⫺7 or lower.
(C) Composite sequence of putative TFIID binding sites (Inr, MTE, and DPE) derived from a statistical analysis of 3,393 nonredundant Drosophila
promoters (16). (D) Average percentage of hyperreactive thymines at each position relative to the initiator element. Thymines in the regions
encompassing the permanganate footprints were given a score of 1 if they were judged to be hyperreactive and a score of 0 if not. The percentage
of hyperreactive thymines in a three-nucleotide window was calculated, and the percentage at the center nucleotide of the three-nucleotide window
was plotted as a function of distance from the adenine in the initiator.
Downloaded from mcb.asm.org at Penn State Univ on October 23, 2009
VOL. 28, 2008
3296
LEE ET AL.
MOL. CELL. BIOL.
FIG. 5. Comparison of NELF and GAGA factor distributions.
(A) Conserved sequence motif identified in the vicinity of regions with a
permanganate footprint. MEME analysis of the region from 200 nucleotides upstream to 100 nucleotides downstream from the Inr elements
detected enrichment of the GAGA element. P values for regions containing this motif were 3.76 ⫻ 10⫺5 or lower. (B) Permanganate footprints
and distributions of NELF-B and GAGA factor on the tai and D-fos
genes. The footprinted region of tai is approximately 9 kb downstream
from the annotated transcription start site. Nevertheless, the presence of
an initiator element (Fig. 4) indicates that this could be a transcription
start site. (C) Venn diagram displaying the relationships between genes
associated with NELF-B, NELF-E, and GAGA factor. Separate lists were
of both NELF-B and -E within 500 bp of the transcription start
(Fig. 5C; see File S6 in the supplemental material). The broad
distribution of NELF throughout the genome revealed by
ChIP-chip analysis is in excellent agreement with the broad
distribution detected on Drosophila polytene chromosomes
(59, 61). The close association of NELF with promoter regions
raises the possibility that NELF is involved in the expression of
many genes.
We investigated the relationship between NELF and paused
Pol II by performing permanganate genomic footprinting on
the promoter regions of 58 genes and one region that exhibited
an obvious peak of NELF located 9 kb downstream from the
annotated start site. Of 56 NELF-associated genes, 46 had a
clearly discernible footprint (see File S1 in the supplemental
material). Our results can be extrapolated to predict the number and identity of genes that are likely to have Pol II paused
in the promoter-proximal region. NELF-associated genes were
ranked according to the level of NELF reflected by the Mpeak
value. Forty-four of the genes we analyzed are within the
collection of genes with the top 1,000 Mpeak values, and 42 of
these had permanganate footprints. Thus, these top 1,000
genes represent cases where we anticipate detection of paused
Pol II with permanganate.
Comparison of our data with recent Pol II ChIP-chip data
obtained for Drosophila S2 cells provides another measure of
the relationship between NELF and promoter-proximal pausing. Muse et al. (38) identified 5,403 genes with Pol II, and the
majority of these had Pol II concentrated at the 5⬘ end. Based
on the unusually high density of Pol II found at the 5⬘ end,
1,014 genes were judged to have stalled Pol II at the 5⬘ end.
compiled for genes that had a peak of NELF-B or NELF-E within 500
base pairs of the transcription start site or of GAGA factor within
1,000 base pairs of the transcription start site (see File S6 in the
supplemental material).
Downloaded from mcb.asm.org at Penn State Univ on October 23, 2009
FIG. 6. Relationship between gene expression level and NELF.
Data from the work of Rehwinkel et al. (49) were used to rank 13,701
genes according to expression level. The percentage of genes in each
bin of 10 percentile points with a NELF-B peak within 500 bp of the
start was plotted against the percentile of expression. Each bin contains approximately 1,370 genes.
VOL. 28, 2008
These Pol II were described as stalled rather than paused
because there is uncertainty in the conformation of the elongation complex (see Discussion in the work of Muse et al. [38]).
Nevertheless, since these conformations are related to pausing,
the relationships between genes with NELF, GAGA factor,
and stalled Pol II were assessed. Thirty-nine of the 46 genes
with permanganate footprints were found among the genes
with stalled Pol II, indicating that a significant portion of the
stalled Pol II are indeed transcriptionally engaged in the promoter-proximal region. The Venn diagram in Fig. 7 shows the
concordance between our NELF-associated genes and those
with stalled Pol II. Eighty percent of the genes with stalled Pol
II coincide with our NELF-associated genes. Collectively, the
data indicate that the presence of NELF, the presence of
paused Pol II, and the presence of stalled Pol II are linked to
each other. The interdependence of these gene features is
further supported by the finding that the distribution of Pol II
was altered on 115 of 200 genes with stalled Pol II when NELF
was depleted with RNA interference (38).
Importantly, it is also evident that pausing is not dictated
solely by NELF. We detected 10 genes that lacked a permanganate footprint yet clearly associated with NELF. Muse et al.
(38) observed that 85 of 200 genes with stalled Pol II displayed
no significant change in the distribution of Pol II when NELF
was depleted with RNA interference. It is noteworthy that
NELF slows but does not stop elongation in vitro (11, 64).
Thus, the density of Pol II in the promoter-proximal region is
probably influenced by additional factors. These could include
gene-specific factors that influence initiation rates and the duration of the pause. Chromatin structure could impede elongation, as appears to be the case for hsp70 in human cells (6,
12). Some cases could involve Pol II in a preinitiated state, as
observed by Dellino et al. (14). Other cases could involve
premature termination, as observed by Zhang et al. (68). Further analysis of individual genes is necessary to explore these
possibilities.
NELF associates with highly expressed genes. Based on
previous work (1, 2, 61), we anticipated that NELF would
repress transcription. Surprisingly, we observed that approxi-
3297
mately 80% of the NELF-associated genes were among the
upper half of genes ranked by expression level (Fig. 6). Almost
one-half of the genes above the 90th percentile in expression
associated with NELF. In addition, 35 of the 46 genes for
which we detected a permanganate footprint ranked in the top
50th percentile for expression.
As proposed by Sims et al. (53), NELF could function as a
checkpoint during an early stage in elongation. The delay in
elongation could allow time for both elongation factors and
RNA processing factors to associate with Pol II. The checkpoint could also serve to attenuate the level of expression of
some genes whose expression levels might otherwise be too
high for the biological functions of the genes. This has been
proposed for some estrogen-induced genes, wherein the estrogen receptor appears to recruit NELF at the same time that it
is activating transcription (2).
Many genes with paused Pol II and NELF associate with
GAGA factor. The Venn diagram in Fig. 7 shows that GAGA
factor associates with a significant number of genes that associate with NELF or a stalled Pol II. GAGA factor was previously implicated in pausing Pol II on hsp70 (31, 52). GAGA
factor interacts with TFIID, the chromatin remodeling factor
NURF, and the histone chaperone FACT (20, 39, 62). Thus,
GAGA factor could function in initiation by recruiting TFIID
and by establishing a nucleosome-free region that allows access
of the transcription machinery to the promoter. It is unclear if
GAGA factor directly impacts elongation. Pausing could be a
default outcome of GAGA factor’s inability to overcome the
action of NELF, possibly because it does not recruit P-TEFb to
the gene.
Approximately 40% of the GAGA factor-associated genes
do not associate with NELF (Fig. 7). More investigation is
needed to determine what GAGA factor does at these genes.
GAGA factor associates with polycomb response elements
(25), which are involved in transcription repression, so GAGA
factor could be involved in organizing a repressive chromatin
structure that blocks transcription initiation.
Six genes for which we detected a permanganate footprint
lack GAGA factor, and approximately one-half of the genes
identified by Muse et al. (38) lack GAGA factor (Fig. 7). This
raises the possibility of the existence of other sequence-specific
DNA binding proteins that function similarly to GAGA factor.
These could be proteins that participate in initiation but are
unable to directly impact elongation. Such proteins were previously implicated in an earlier study (4), but a direct link to
promoter-proximal pausing and NELF remains to be investigated.
NELF-associated genes with paused Pol II are enriched for
a combination of TFIID binding sites. DNA sequence analysis
of the regions encompassing the permanganate footprints of 46
genes identified a 35-nucleotide region with three conserved
patches of nucleotides, centered at positions ⫹1, ⫹18, and ⫹29
relative to known or putative transcription start sites. These
conserved patches align with three contacts made by TFIID in
the first 30 nucleotides of the hsp70, hsp26, and histone H3
genes (44, 45), strongly suggesting that the three conserved
patches are recognized by TFIID. This conclusion is further
supported by the similarity between the sequences of the initiator and DPE, which are recognized by TFIID (7, 44), and
Downloaded from mcb.asm.org at Penn State Univ on October 23, 2009
FIG. 7. Relationships between NELF-associated genes, GAGA
factor-associated genes, and genes with stalled Pol II. NELF-associated genes are those genes that have both NELF-B and NELF-E peaks
within 500 bp of the transcription start. GAGA factor-associated genes
were identified as described in the legend to Fig. 5. Genes with stalled
Pol II were identified by Muse et al. (38). Only 983 of the 1,014 genes
with stalled Pol II could be matched up with our set of 13,745 genes.
File S6 in the supplemental material provides a list of genes with
associated factors.
PROMOTER-PROXIMAL PAUSING IN DROSOPHILA
3298
LEE ET AL.
and RNA secondary structure. NELF was previously shown to
associate with a variety of RNAs, including some with substantial secondary structure and some without it (63). Our conclusion contrasts with studies of the human immunodeficiency
virus (HIV) provirus, for which sequence-specific association
of NELF-E with the TAR element has been detected and is
thought to control elongation (15, 46). Recent analysis of an
HIV provirus, however, indicates that NELF associates with
the elongation complex and causes Pol II to pause at position
⫹45, well before the TAR element has been transcribed and
well within the range where we find Pol II pausing on Drosophila genes (67).
Transcriptional control in animal cells. Knowledge of transcriptional control is based largely on studies with yeast and
cell-free systems. These studies have tended to emphasize the
regulation of transcription initiation. New data provided by
genome-wide analyses of Pol II significantly increase the number of genes whose transcription appears to be controlled after
initiation in animal cells (23, 38, 66). Thus, a full understanding
of transcriptional control in animal cells requires investigation
of molecular events occurring both before and after initiation
in the promoter regions. Knowledge of the genome-wide distribution of NELF and GAGA factor combined with recent
data on the distribution of Pol II provides a foundation for
future studies of promoter-proximal pausing in Drosophila.
ACKNOWLEDGMENTS
We thank Joe Reese and Song Tan for their comments on the
manuscript and Anton Nekrutenko for assistance with Galaxy.
This work was supported by NIH grant GM47477 to D.S.G., NIH grant
HG004160 to B.F.P., and grant GM704403 awarded to the Berkeley
Drosophila Transcription Network Project from NIGMS and NHGRI.
The work at Lawrence Berkeley National Laboratory was conducted
under Department of Energy contract DE-AC02-05CH11231.
REFERENCES
1. Aida, M., Y. Chen, K. Nakajima, Y. Yamaguchi, T. Wada, and H. Handa.
2006. Transcriptional pausing caused by NELF plays a dual role in regulating
immediate-early expression of the junB gene. Mol. Cell. Biol. 26:6094–6104.
2. Aiyar, S. E., J. L. Sun, A. L. Blair, C. A. Moskaluk, Y. Z. Lu, Q. N. Ye, Y.
Yamaguchi, A. Mukherjee, D. M. Ren, H. Handa, and R. Li. 2004. Attenuation of estrogen receptor alpha-mediated transcription through estrogen-stimulated recruitment of a negative elongation factor. Genes
Dev. 18:2134–2146.
3. Bailey, T. L., N. Williams, C. Misleh, and W. W. Li. 2006. MEME: discovering and analyzing DNA and protein sequence motifs. Nucleic Acids Res.
34:W369–W373.
4. Blau, J., H. Xiao, S. McCracken, P. O’Hare, J. Greenblatt, and D. Bentley.
1996. Three functional classes of transcriptional activation domain. Mol.
Cell. Biol. 16:2044–2055.
5. Boehm, A. K., A. Saunders, J. Werner, and J. T. Lis. 2003. Transcription
factor and polymerase recruitment, modification, and movement on dhsp70
in vivo in the minutes following heat shock. Mol. Cell. Biol. 23:7628–7637.
6. Brown, S. A., A. N. Imbalzano, and R. E. Kingston. 1996. Activator-dependent regulation of transcriptional pausing on nucleosomal templates. Genes
Dev. 10:1479–1490.
7. Burke, T. W., and J. T. Kadonaga. 1996. Drosophila TFIID binds to a
conserved downstream basal promoter element that is present in many
TATA-box-deficient promoters. Genes Dev. 10:711–724.
8. Cartwright, I. L., D. E. Cryderman, D. S. Gilmour, L. A. Pile, L. L. Wallrath,
J. A. Weber, and S. C. Elgin. 1999. Analysis of Drosophila chromatin structure in vivo. Methods Enzymol. 304:462–496.
9. Chalkley, G. E., and C. P. Verrijzer. 1999. DNA binding site selection by
RNA polymerase II TAFs: a TAF(II)250-TAF(II)150 complex recognizes
the initiator. EMBO J. 18:4835–4845.
10. Chao, S. H., and D. H. Price. 2001. Flavopiridol inactivates P-TEFb and
blocks most RNA polymerase II transcription in vivo. J. Biol. Chem. 276:
31793–31799.
11. Cheng, B., and D. H. Price. 2007. Properties of RNA polymerase II elongation complexes before and after the P-TEFb-mediated transition into productive elongation. J. Biol. Chem. 282:21901–21912.
Downloaded from mcb.asm.org at Penn State Univ on October 23, 2009
the sequences centered at positions ⫹1 and ⫹29 in our conserved element (compare Fig. 4B and C).
While it is not surprising for the promoters with paused Pol
II to have TFIID binding sites, it is unusual to have such a large
proportion of promoters with multiple binding sites. Statistical
analysis of 3,393 nonredundant Drosophila promoters revealed
that only 12% of the promoters contained an Inr and a DPE
(16), whereas we found this combination in 61% (28/46 cases)
of the cases with paused Pol II. These two regions, in combination with a third conserved patch of nucleotides near position ⫹18, further distinguish this collection of promoters with
paused Pol II from the rest of the promoters in the genome.
The combination of TFIID contacts in the regions at positions ⫹1, ⫹18, and ⫹29 could contribute to pausing by increasing the affinity of TFIID for the core promoter region, thus
shifting the rate-limiting step from the binding of TFIID to a
subsequent step, such as promoter-proximal pausing. In addition, the contacts of TFIID downstream from the transcription
start might help to establish a promoter architecture that is
conducive to pausing. DNase I footprinting and UV crosslinking analyses showed that TFIID contacts DNA as far as 45
nucleotides downstream from the transcription start (56, 60).
Pol II pauses in the region 30 to 50 nucleotides from the
start site of NELF-associated genes, but the pause does not
appear to be dictated by sequence-specific nucleic acid contacts with the elongation complex. Our permanganate footprinting results indicate that Pol II pauses after transcribing
approximately 30 to 50 nucleotides. This agrees well with highresolution mapping of nascent transcripts associated with
paused Pol II on the hsp70 and hsp27 genes (47) and with the
locations of permanganate footprints detected on several other
genes (1, 18, 29, 43, 67). Because the footprints at several
different promoters extend further than the size of a bubble
associated with a single Pol II molecule, the location of the
paused Pol II must often be heterogeneous. This also agrees
with previous results for the hsp70 and hsp27 genes (47).
The distance where Pol II begins to pause downstream from
the start site is similar to the distance where NELF was found
to inhibit elongation in a cell-free system (64). This strengthens
our conclusion that NELF is involved in concentrating Pol II in
the promoter-proximal region. Moreover, our results fit well
with a proposed pausing mechanism in which the NELF-E
subunit associates with the nascent transcript and slows elongation by restricting the extrusions of the nascent transcript
from the elongation complex (40). This model requires that
Pol II transcribe at least 18 nucleotides so that the transcript is
exposed outside the elongation complex (22).
The location of paused Pol II provides a landmark for sequence comparisons directed at identifying sequences involved
in promoter-proximal pausing. The location where Pol II
pauses does not correlate with conserved sequences or conserved RNA structures. The conserved element that is likely to
be recognized by TFIID is located various distances from
where Pol II pauses, so it seems unlikely that this element
directly influences Pol II. Mfold analyses (69) of the portions
of nascent transcripts that would be exposed from the elongation complexes failed to detect evidence of substantial secondary structure.
We propose that NELF’s interaction with the nascent transcript associated with paused Pol II is independent of sequence
MOL. CELL. BIOL.
VOL. 28, 2008
36.
37.
38.
39.
40.
41.
42.
43.
44.
45.
46.
47.
48.
49.
50.
51.
52.
53.
54.
55.
56.
57.
58.
59.
60.
61.
3299
p. 459–475. In R. C. Conaway and J. W. Conaway (ed.), Transcription:
mechanisms and regulation. Raven Press, Ltd., New York, NY.
Marshall, N. F., and D. H. Price. 1992. Control of formation of two distinct
classes of RNA polymerase II elongation complexes. Mol. Cell. Biol. 12:
2078–2090.
Marshall, N. F., and D. H. Price. 1995. Purification of P-TEFb, a transcription factor required for the transition into productive elongation. J. Biol.
Chem. 270:12335–12338.
Muse, G. W., D. A. Gilchrist, S. Nechaev, R. Shah, J. S. Parker, S. F.
Grissom, J. Zeitlinger, and K. Adelman. 2007. RNA polymerase is poised for
activation across the genome. Nat. Genet. 39:1507–1511.
Nakayama, T., K. Nishioka, Y. X. Dong, T. Shimojima, and S. Hirose. 2007.
Drosophila GAGA factor directs histone H3.3 replacement that prevents the
heterochromatin spreading. Genes Dev. 21:552–561.
Narita, T., Y. Yamaguchi, K. Yano, S. Sugimoto, S. Chanarat, T. Wada, D. K.
Kim, J. Hasegawa, M. Omori, N. Inukai, M. Endoh, T. Yamada, and H.
Handa. 2003. Human transcription elongation factor NELF: identification of
novel subunits and reconstitution of the functionally active complex. Mol.
Cell. Biol. 23:1863–1873.
Narita, T., T. M. Yung, J. Yamamoto, Y. Tsuboi, H. Tanabe, K. Tanaka, Y.
Yamaguchi, and H. Handa. 2007. NELF interacts with CBC and participates
in 3⬘ end processing of replication-dependent histone mRNAs. Mol. Cell
26:349–365.
Peterlin, B. M., and D. H. Price. 2006. Controlling the elongation phase of
transcription with P-TEFb. Mol. Cell 23:297–305.
Plet, A., D. Eick, and J. M. Blanchard. 1995. Elongation and premature
termination of transcripts initiated from c-fos and c-myc promoters show
dissimilar patterns. Oncogene 10:319–328.
Purnell, B. A., P. A. Emanuel, and D. S. Gilmour. 1994. TFIID sequence
recognition of the initiator and sequences farther downstream in Drosophila
class II genes. Genes Dev. 8:830–842.
Purnell, B. A., and D. S. Gilmour. 1993. Contribution of sequences downstream of the TATA element to a protein-DNA complex containing the
TATA-binding protein. Mol. Cell. Biol. 13:2593–2603.
Rao, J. N., L. Neumann, S. Wenzel, K. Schweimer, P. Rosch, and B. Wohrl.
2006. Structural studies on the RNA-recognition motif of NELF E, a cellular
negative transcription elongation factor involved in the regulation of HIV
transcription. Biochem. J. 400:449–456.
Rasmussen, E. B., and J. T. Lis. 1993. In vivo transcriptional pausing and cap
formation on three Drosophila heat shock genes. Proc. Natl. Acad. Sci. USA
90:7923–7927.
Rasmussen, E. B., and J. T. Lis. 1995. Short transcripts of the ternary
complex provide insight into RNA polymerase II elongational pausing. J.
Mol. Biol. 252:522–535.
Rehwinkel, J., P. Natalin, A. Stark, J. Brennecke, S. M. Cohen, and E.
Izaurralde. 2006. Genome-wide analysis of mRNAs regulated by Drosha and
Argonaute proteins in Drosophila melanogaster. Mol. Cell. Biol. 26:2965–
2975.
Rougvie, A. E., and J. T. Lis. 1988. The RNA polymerase II molecule at the
5⬘ end of the uninduced hsp70 gene of D. melanogaster is transcriptionally
engaged. Cell 54:795–804.
Saunders, A., L. J. Core, and J. T. Lis. 2006. Breaking barriers to transcription elongation. Nat. Rev. Mol. Cell. Biol. 7:557–567.
Shopland, L. S., K. Hirayoshi, M. Fernandes, and J. T. Lis. 1995. HSF access
to heat shock elements in vivo depends critically on promoter architecture
defined by GAGA factor, TFIID, and RNA polymerase II binding sites.
Genes Dev. 9:2756–2769.
Sims, R. J., III, R. Belotserkovskaya, and D. Reinberg. 2004. Elongation by
RNA polymerase II: the short and long of it. Genes Dev. 18:2437–2468.
Smale, S. T., and J. T. Kadonaga. 2003. The RNA polymerase II core
promoter. Annu. Rev. Biochem. 72:449–479.
Spencer, C. A., and M. Groudine. 1990. Transcription elongation and eukaryotic gene regulation. Oncogene 5:777–785.
Sypes, M. A., and D. S. Gilmour. 1994. Protein/DNA crosslinking of a TFIID
complex reveals novel interactions downstream of the transcription start.
Nucleic Acids Res. 22:807–814.
Wada, T., T. Takagi, Y. Yamaguchi, A. Ferdous, T. Imai, S. Hirose, S.
Sugimoto, K. Yano, G. A. Hartzog, F. Winston, S. Buratowski, and H.
Handa. 1998. DSIF, a novel transcription elongation factor that regulates
RNA polymerase II processivity, is composed of human Spt4 and Spt5
homologs. Genes Dev. 12:343–356.
Wang, Y. V., H. Tang, and D. S. Gilmour. 2005. Identification in vivo of
different rate-limiting steps associated with transcriptional activators in the
presence and absence of a GAGA element. Mol. Cell. Biol. 25:3543–3552.
Wu, C. H., C. Lee, R. Fan, M. J. Smith, Y. Yamaguchi, H. Handa, and D. S.
Gilmour. 2005. Molecular characterization of Drosophila NELF. Nucleic
Acids Res. 33:1269–1279.
Wu, C. H., L. Madabusi, H. Nishioka, P. Emanuel, M. Sypes, I. Arkhipova,
and D. S. Gilmour. 2001. Analysis of core promoter sequences located
downstream from the TATA element in the hsp70 promoter from Drosophila
melanogaster. Mol. Cell. Biol. 21:1593–1602.
Wu, C. H., Y. Yamaguchi, L. R. Benjamin, M. Horvat-Gordon, J. Washinsky,
Downloaded from mcb.asm.org at Penn State Univ on October 23, 2009
12. Corey, L. L., C. S. Weirich, I. J. Benjamin, and R. E. Kingston. 2003.
Localized recruitment of a chromatin-remodeling activity by an activator in
vivo drives transcriptional elongation. Genes Dev. 17:1392–1401.
13. Das, R., A. Laederach, S. M. Pearlman, D. Herschlag, and R. B. Altman.
2005. SAFA: semi-automated footprinting analysis software for highthroughput quantification of nucleic acid footprinting experiments. RNA
11:344–354.
14. Dellino, G. I., Y. B. Schwartz, G. Farkas, D. McCabe, S. C. Elgin, and V.
Pirrotta. 2004. Polycomb silencing blocks transcription initiation. Mol. Cell
13:887–893.
15. Fujinaga, K., D. Irwin, Y. Huang, R. Taube, T. Kurosu, and B. M. Peterlin.
2004. Dynamics of human immunodeficiency virus transcription: P-TEFb
phosphorylates RD and dissociates negative effectors from the transactivation response element. Mol. Cell. Biol. 24:787–795.
16. Gershenzon, N. I., E. N. Trifonov, and I. P. Ioshikhes. 2006. The features of
Drosophila core promoters revealed by statistical analysis. BMC Genomics
7:161.
17. Giardina, C., and J. T. Lis. 1993. Polymerase processivity and termination on
Drosophila heat shock genes. J. Biol. Chem. 268:23806–23811.
18. Giardina, C., M. Perez-Riba, and J. T. Lis. 1992. Promoter melting and
TFIID complexes on Drosophila genes in vivo. Genes Dev. 6:2190–2200.
19. Gilmour, D. S., and J. T. Lis. 1986. RNA polymerase II interacts with the
promoter region of the noninduced hsp70 gene in Drosophila melanogaster
cells. Mol. Cell. Biol. 6:3984–3989.
20. Giot, L., J. S. Bader, C. Brouwer, A. Chaudhuri, B. Kuang, Y. Li, Y. L. Hao,
C. E. Ooi, B. Godwin, E. Vitols, G. Vijayadamodar, P. Pochart, H. Machineni, M. Welsh, Y. Kong, B. Zerhusen, R. Malcolm, Z. Varrone, A. Collis, M.
Minto, S. Burgess, L. McDaniel, E. Stimpson, F. Spriggs, J. Williams, K.
Neurath, N. Ioime, M. Agee, E. Voss, K. Furtak, R. Renzulli, N. Aanensen,
S. Carrolla, E. Bickelhaupt, Y. Lazovatsky, A. DaSilva, J. Zhong, C. A.
Stanyon, R. L. Finley, Jr., K. P. White, M. Braverman, T. Jarvie, S. Gold, M.
Leach, J. Knight, R. A. Shimkets, M. P. McKenna, J. Chant, and J. M.
Rothberg. 2003. A protein interaction map of Drosophila melanogaster.
Science 302:1727–1736.
21. Gnatt, A. L., P. Cramer, J. Fu, D. A. Bushnell, and R. D. Kornberg. 2001.
Structural basis of transcription: an RNA polymerase II elongation complex
at 3.3 A resolution. Science 292:1876–1882.
22. Gu, W., M. Wind, and D. Reines. 1996. Increased accommodation of nascent
RNA in a product site on RNA polymerase II during arrest. Proc. Natl.
Acad. Sci. USA 93:6935–6940.
23. Guenther, M. G., S. S. Levine, L. A. Boyer, R. Jaenisch, and R. A. Young.
2007. A chromatin landmark and transcription initiation at most promoters
in human cells. Cell 130:77–88.
24. Harbison, C. T., D. B. Gordon, T. I. Lee, N. J. Rinaldi, K. D. Macisaac, T. W.
Danford, N. M. Hannett, J. B. Tagne, D. B. Reynolds, J. Yoo, E. G. Jennings,
J. Zeitlinger, D. K. Pokholok, M. Kellis, P. A. Rolfe, K. T. Takusagawa, E. S.
Lander, D. K. Gifford, E. Fraenkel, and R. A. Young. 2004. Transcriptional
regulatory code of a eukaryotic genome. Nature 431:99–104.
25. Horard, B., C. Tatout, S. Poux, and V. Pirrotta. 2000. Structure of a polycomb response element and in vitro binding of polycomb group complexes
containing GAGA factor. Mol. Cell. Biol. 20:3187–3197.
26. Johnson, W. E., W. Li, C. A. Meyer, R. Gottardo, J. S. Carroll, M. Brown,
and X. S. Liu. 2006. Model-based analysis of tiling-arrays for ChIP-chip.
Proc. Natl. Acad. Sci. USA 103:12457–12462.
27. Kim, T. H., L. O. Barrera, M. Zheng, C. Qu, M. A. Singer, T. A. Richmond,
Y. Wu, R. D. Green, and B. Ren. 2005. A high-resolution map of active
promoters in the human genome. Nature 436:876–880.
28. Kireeva, M. L., N. Komissarova, D. S. Waugh, and M. Kashlev. 2000. The
8-nucleotide-long RNA:DNA hybrid is a primary stability determinant of the
RNA polymerase II elongation complex. J. Biol. Chem. 275:6530–6536.
29. Krumm, A., T. Meulia, M. Brunvand, and M. Groudine. 1992. The block to
transcriptional elongation within the human c-myc gene is determined in the
promoter-proximal region. Genes Dev. 6:2201–2213.
30. Lam, L. T., O. K. Pickeral, A. C. Peng, A. Rosenwald, E. M. Hurt, J. M.
Giltnane, L. M. Averett, H. Zhao, R. E. Davis, M. Sathyamoorthy, L. M.
Wahl, E. D. Harris, J. A. Mikovits, A. P. Monks, M. G. Hollingshead, E. A.
Sausville, and L. M. Staudt. 2001. Genomic-scale measurement of mRNA
turnover and the mechanisms of action of the anti-cancer drug flavopiridol.
Genome Biol. 2:41.1–41.11.
31. Lee, H., K. W. Kraus, M. F. Wolfner, and J. T. Lis. 1992. DNA sequence
requirements for generating paused polymerase at the start of hsp70. Genes
Dev. 6:284–295.
32. Lim, C. Y., B. Santoso, T. Boulay, E. Dong, U. Ohler, and J. T. Kadonaga.
2004. The MTE, a new core promoter element for transcription by RNA
polymerase II. Genes Dev. 18:1606–1617.
33. Lis, J. 1998. Promoter-associated pausing in promoter architecture and
postinitiation transcriptional regulation. Cold Spring Harbor Symp. Quant.
Biol. 63:347–356.
34. Lis, J. T., P. Mason, J. Peng, D. H. Price, and J. Werner. 2000. P-TEFb
kinase recruitment and function at heat shock loci. Genes Dev. 14:792–803.
35. Lis, J. T., and C. Wu. 1994. Transcriptional regulation of heat shock genes,
PROMOTER-PROXIMAL PAUSING IN DROSOPHILA
3300
62.
63.
64.
65.
LEE ET AL.
E. Enerly, J. Larsson, A. Lambertsson, H. Handa, and D. Gilmour. 2003.
NELF and DSIF cause promoter proximal pausing on the hsp70 promoter in
Drosophila. Genes Dev. 17:1402–1414.
Xiao, H., R. Sandaltzopoulos, H. M. Wang, A. Hamiche, R. Ranallo, K. M.
Lee, D. Fu, and C. Wu. 2001. Dual functions of largest NURF subunit
NURF301 in nucleosome sliding and transcription factor interactions. Mol.
Cell 8:531–543.
Yamaguchi, Y., N. Inukai, T. Narita, T. Wada, and H. Handa. 2002. Evidence
that negative elongation factor represses transcription elongation through
binding to a DRB sensitivity-inducing factor/RNA polymerase II complex
and RNA. Mol. Cell. Biol. 22:2918–2927.
Yamaguchi, Y., T. Takagi, T. Wada, K. Yano, A. Furuya, S. Sugimoto, J.
Hasegawa, and H. Handa. 1999. NELF, a multisubunit complex containing
RD, cooperates with DSIF to repress RNA polymerase II elongation. Cell
97:41–51.
Zanton, S. J., and B. F. Pugh. 2004. Changes in genomewide occupancy of
MOL. CELL. BIOL.
66.
67.
68.
69.
core transcriptional regulators during heat stress. Proc. Natl. Acad. Sci. USA
101:16843–16848.
Zeitlinger, J., A. Stark, M. Kellis, J. W. Hong, S. Nechaev, K. Adelman, M.
Levine, and R. A. Young. 2007. RNA polymerase stalling at developmental
control genes in the Drosophila melanogaster embryo. Nat. Genet. 39:1512–
1516.
Zhang, Z., A. Klatt, D. S. Gilmour, and A. J. Henderson. 2007. Negative
elongation factor NELF represses human immunodeficiency virus transcription by pausing the RNA polymerase II complex. J. Biol. Chem. 282:16981–
16988.
Zhang, Z., A. Klatt, A. J. Henderson, and D. S. Gilmour. 2007. Transcription
termination factor Pcf11 limits the processivity of Pol II on an HIV provirus
to repress gene expression. Genes Dev. 21:1609–1614.
Zuker, M. 2003. Mfold web server for nucleic acid folding and hybridization
prediction. Nucleic Acids Res. 31:3406–3415.
Downloaded from mcb.asm.org at Penn State Univ on October 23, 2009