Molecular Cell, Vol. 20, 213–223, October 28, 2005, Copyright ©2005 by Elsevier Inc. DOI 10.1016/j.molcel.2005.08.023
A Posttranscriptional Role for the Yeast
Paf1-RNA Polymerase II Complex Is Revealed
by Identification of Primary Targets
Kristi L. Penheiter, Taylor M. Washburn,
Stephanie E. Porter, Matthew G. Hoffman,
and Judith A. Jaehning*
Department of Biochemistry and Molecular
Molecular Biology Program
University of Colorado, Denver and
Health Sciences Center, B121
4200 East Ninth Avenue
Denver, Colorado 80262
The yeast Paf1 complex (Paf1C: Paf1, Cdc73, Ctr9,
Rtf1, and Leo1) is associated with RNA Polymerase II
(Pol II) at promoters and coding regions of transcrip-
tionally active genes, but transcript abundance for
only a small subset of genes is altered by loss of
Paf1. By using conditional and null alleles of PAF1
and microarrays, we determined the identity of both
primary and secondary targets of the Paf1C. Neither
primary nor secondary Paf1C target promoters were
responsive to loss of Paf1. Instead, Paf1 loss altered
poly(A) site utilization of primary target genes SDA1
and MAK21, resulting in increased abundance of 3?-
extended mRNAs. The 3?-extended MAK21 RNA is
sensitive to nonsense-mediated decay (NMD), as re-
vealed by its increased abundance in the absence of
Upf1. Therefore, although the Paf1C is associated
with Pol II at initiation and during elongation, these
critical Paf1-dependent changes in transcript abun-
dance are due to alterations in posttranscriptional
RNA Pol II undergoes a complex series of interactions
during the transcription cycle. Pol II is recruited to the
promoter by the general transcription factors (GTFs)
and the mediator complex (Hahn, 2004; Kornberg,
2005), which are necessary for accurate initiation at
promoters and critical for response to transcriptional
activators and repressors. During the transition from
initiation to elongation, GTF and mediator interactions
with Pol II are exchanged for interactions with capping
and elongation factors and histone modifying enzymes
(Dvir et al., 2001; Gerber and Shilatifard, 2003; Zorio
and Bentley, 2004). Pol II is joined during elongation by
factors involved in 3# end formation, polyadenylation,
and nuclear export (Jensen et al., 2003; Zorio and Bent-
ley, 2004). These dynamic interactions are mediated in
part by changes in the phosphorylation state of the
C-terminal domain (CTD) of the largest subunit of Pol
II. The unphosphorylated CTD associates with initiation
factors and the promoter, phosphorylation at serine 5
of the CTD (Ser5) is required for promoter escape and
association with capping factors, and phosphorylation
at serine 2 (Ser2) is necessary for interactions with his-
tone modifying enzymes and the cleavage and poly-
adenylation factors (Buratowski, 2003; Palancade and
Bensaude, 2003; Zorio and Bentley, 2004).
There is extensive communication between these dif-
ferent stages of transcription and posttranscriptional
processing. In particular, the “quality” of the mRNA is
checked at the site of transcription before the tran-
script is licensed for export (Jensen et al., 2003, 2004;
Neugebauer, 2002). Defective transcripts are rapidly
degraded by the nuclear exosome. Transcripts ex-
ported to the cytoplasm are subject to another quality
control mechanism: that of mRNA surveillance or NMD.
First identified as a process that degrades mRNAs with
premature termination codons (Peltz et al., 1993), NMD
also degrades transcripts with abnormally extended
3# untranslated regions (UTRs) (Muhlrad and Parker,
The yeast Paf1 complex (Paf1C), comprised of Paf1,
Ctr9, Cdc73, Rtf1, and Leo1, is associated with Pol II
throughout the transcription cycle (Mueller and Jaehn-
ing, 2002; Mueller et al., 2004; Squazzo et al., 2002).
The Paf1C from yeast and humans associates with the
unphosphorylated form of Pol II at promoters and the
Ser2 and Ser5 phosphorylated forms present during
elongation (Rozenblatt-Rosen et al., 2005; Wade et al.,
1996; M.G.H., unpublished data). Although the Paf1C
does not require Ser2 phosphorylation for its associa-
tion with chromatin (Ahn et al., 2004), loss of Paf1 re-
sults in reduced Ser2 phosphorylation with a concomi-
tant reduction of Ser2-dependent factors (Mueller et al.,
2004; M.G.H., unpublished data). Consistent with its
presence at all stages of transcription, the Paf1C has
been found to associate with initiation and elongation
factors (TFIIB, TFIIF, Spt5, and Dst1 [Krogan et al.,
2002; Mueller and Jaehning, 2002; Squazzo et al., 2002;
Wade et al., 1996]), chromatin remodeling and modify-
ing factors (Chd1 and Set1 [Gerber and Shilatifard,
2003; Simic et al., 2003]), and factors involved in mRNA
processing, export, and quality control (Hpr1, Sub2,
and hSki8 [Chang et al., 1999; Mueller et al., 2004; Zhu
et al., 2005]). Because the Paf1C is biochemically dis-
tinct from the Srb/mediator form of Pol II, it must join
the transcription complex after the initial recruitment
event (Shi et al., 1997). From that point on, the Paf1C
accompanies the RNA polymerase, possibly serving as
a “platform” for the dynamic association of additional
factors during elongation (Gerber and Shilatifard, 2003).
Although the Paf1C appears to be associated with
Pol II at all transcriptionally active yeast genes (Ahn et
al., 2004; Mueller et al., 2004; Simic et al., 2003), its
loss results in transcript abundance changes for only a
small subset (Porter et al., 2002; Shi et al., 1997). How-
ever, isogenic paf1D and ctr9D strains exhibit severe
and nearly indistinguishable growth defects (Betz et al.,
2002; Koch et al., 1999). In contrast, loss of Cdc73 or
Rtf1 results in less severe phenotypes that are a subset
of those seen in paf1D and ctr9D; leo1D strains have
few detectable deficiencies (Betz et al., 2002). There-
fore, Paf1 and Ctr9 appear to have the most important
and essentially identical roles in the activity of the
To further investigate the role for the Paf1C in gene
expression, we have compared the genome-wide tran-
script expression patterns of paf1D and ctr9D mutants
to those of wild-type (wt) cells. Consistent with the
identical phenotypes of paf1D and ctr9D mutants, we
found substantial overlap in their microarray expression
profiles. To identify primary Paf1C targets, we created
a conditional, tetracycline-controlled form of Paf1 and
followed patterns of gene expression as Paf1 was de-
pleted from cells. By using both primary and secondary
Paf1 targets, we investigated the transcriptional stage
affected by loss of the Paf1C. We found that Paf1 loss
does not affect initiation, nor does it correlate with al-
terations in elongation, but instead appears to lie in a
later stage of the transcription and processing cycle,
resulting in alternative poly(A) site utilization and in
some cases generating transcripts that are targets of
Paf1 Is Required for the Full Expression
of a Subset of Genes
Loss of Paf1 results in transcript abundance changes
for a subset of yeast genes (Porter et al., 2002; Shi et
al., 1997). In addition, loss of either Paf1 or Ctr9 results
in similar phenotypes, and a double paf1D ctr9D mu-
tant is indistinguishable from either single mutant (Betz
et al., 2002). To identify all of the Paf1-dependent tran-
scripts, and confirm that the in vivo roles of Paf1 and
Ctr9 are similar, we performed Affymetrix oligonucleo-
tide microarray analyses comparing RNA abundance
patterns of wt yeast to isogenic paf1D and ctr9D
strains. The analysis of the microarray data was carried
out as described in the Experimental Procedures. Tran-
scripts with a detectable signal were considered Paf1
or Ctr9 dependent if they increased or decreased at
least 2-fold relative to wt RNA.
Deletion of PAF1 affects the expression of several
hundred yeast genes. 213 genes decreased and 599
genes increased at least 2-fold in paf1D when com-
pared to wt (Table S1 available in the Supplemental
Data with this article online). In addition to affecting the
expression of many cell cycle-regulated genes (Porter
et al., 2002), deletion of PAF1 results in the altered ex-
pression of a substantial number of genes involved in
cell wall biosynthesis and maintenance, and RNA Pol II
transcription. This is consistent with our previous ob-
servation that paf1D mutants display cell wall defects,
detected by sensitivity to cell wall-damaging agents
like caffeine and SDS (Betz et al., 2002; Chang et al.,
1999). We also found that expression of several tran-
scription factors is altered upon deletion of PAF1, indi-
cating that Paf1 indirectly, through other transcription
factors, affects the expression of many genes. Finally,
as expected by the relatively severe phenotype associ-
ated with loss of Paf1, many of the genes in this data
set are stress factor responsive (Causton et al., 2001).
Consistent with the similarity of paf1D and ctr9D phe-
notypes, we found a substantial overlap in Paf1- and
Figure 1. Paf1 and Ctr9 Targets Form an Extensively Overlapping
(A) Venn diagram depicting the overlap of Paf1 and Ctr9-dependent
genes from the microarray analysis. Target gene expression de-
creased at least 2-fold in paf1D orctr9D.
(B) Venn diagram as in (A) for genes whose expression increases
in paf1D orctr9D. For a complete list of Paf1 and Ctr9-dependent
genes, see Table S1.
(C) Promoters from Paf1 target genes do not confer sensitivity to
loss of Paf1. Constructs containing CLN1, FKS1, and ACB1 pro-
moters (down in paf1D) were transformed into wt (YJJ662) and
paf1D (YJJ664) as described (Porter et al., 2002). The construct
containing the TAD2 promoter (down in paf1D) was integrated into
the URA3 locus of wt (YJJ1366) and paf1D (YJJ1368). Extracts
were prepared and luciferase activity measured as described in the
Experimental Procedures. The data are presented as the average ±
the SD from the mean of quadruplicate assays.
Ctr9-dependent genes (Figures 1A and 1B). 65% of the
transcripts that decrease in paf1D also decrease in
ctr9D, and 70% of the transcripts that increase in paf1D
also increase in ctr9D. This confirms that Paf1 and Ctr9
function together in the full expression of a subset of
Gene Expression Driven by Promoters of Paf1-
Dependent Genes Is Not Sensitive to Loss of Paf1
Paf1C components copurify with the unphosphory-
lated, initiating form of Pol II and the GTFs TFIIB and
TFIIF (Shi et al., 1997; Wade et al., 1996). Chromatin
immunoprecipitation (ChIP) experiments have placed
the complex at promoter regions of active genes (Ahn
et al., 2004; Mueller et al., 2004), leading us to ask
whether Paf1 plays a role in transcription initiation. We
Normal Poly(A) Site Utilization Requires Paf1
constructed reporter constructs driving luciferase ex-
pression from the promoters of several Paf1 target
genes. The upstream intergenic regions of CLN1, FKS1,
ACB1 (all decreased in paf1D), and TAD2 (increased in
paf1D) were inserted into luciferase reporter vectors
and integrated into the genome of wt and paf1D strains.
We observed no significant change in luciferase activity
driven by these promoters between wt and paf1D (Fig-
ure 1C). These results agree with our previous findings
(Porter et al., 2002) and establish that the changes we
see in Paf1 target gene expression are not promoter de-
Paf1 Expression and Function Are Effectively Turned
Off by Using a Tetracycline-Regulated Promoter
The microarray experiments described above identified
a combination of both primary and secondary Paf1 tar-
gets. To distinguish primary targets (transcripts directly
responsive to loss of Paf1) from secondary targets
(transcripts whose abundance may be affected by
changes in the expression of other transcription fac-
tors), we constructed a yeast strain in which Paf1 is
under the control of a tetracycline-regulated promoter
(Belli et al., 1998; Gari et al., 1997). We replaced the
PAF1 promoter region with the CYC1 promoter contain-
ing a bacterial tetO operator under the control of a tet-
racycline-responsive activator consisting of a tetO
binding moiety, tetR, fused to the viral VP16 activator.
This fusion protein, when bound to the tetO operator,
consisting of two tetR binding sites, activates tran-
scription of PAF1 (Figure 2A, ON). When the tetracy-
cline derivative doxycycline (DOX) is added to the
growth medium, it binds to tetR, displacing it from the
tetO promoter, shutting off transcription of PAF1 (Figure
Phenotypes associated with loss of Paf1 include
slow growth and increased sensitivity to high temper-
ature and cell wall-damaging agents such as caffeine
(Betz et al., 2002; Chang et al., 1999). When the tet-
regulated Paf1 strain (tetO2−Paf1-6HA) is plated on
YPD medium in the absence of DOX, it behaves like wt
(Figure 2B), growing normally at 30°C, 37°C, and not
sensitive to 8 mM caffeine. However, plus DOX it exhib-
its the temperature-sensitive and caffeine-sensitive
phenotypes of paf1D, confirming that Paf1 function is
shut off by the tet-regulated promoter.
Expression of a Small Number of Genes Is Directly
Sensitive to Loss of Paf1
To identify Paf1C primary targets, we shut off Paf1 ex-
pression by using DOX and isolated RNA from samples
taken at time points before and after the loss of Paf1
expression and function. As shown in Figures 2C and
2D, Paf1 function, measured by growth rate, and Paf1
protein, detected by Western blot, were no longer pre-
sent 5 hr post DOX addition. Total RNA from 0, 2, 4, 5,
6, and 8 hr post DOX addition was isolated and pro-
cessed for hybridization to Affymetrix microarrays.
Transcripts with a detectable signal were considered
Paf1 dependent if they increased or decreased at least
2-fold relative to RNA from the 0 time point. To identify
Paf1 primary targets within this dataset, we selected
genes whose expression decreased or increased at
least 2-fold beginning at 4, 5, or 6 hr post DOX addition
and maintained this change in expression throughout
the remainder of the time course. In addition, to be con-
sidered a primary target, we required the transcript to
exhibit at least a 2-fold change in expression in the
paf1D data set (see above) compared to wt. We elimi-
nated the small number of transcripts whose abun-
dance began to change at 2 hr or not until 8 hr post
DOX addition, because these changes did not correlate
with loss of Paf1. The complete list of primary targets
is shown in Figures 3A and 3B.
We confirmed the results of the microarrays by using
Northern analyses. RNA was again isolated from the
tet-regulated Paf1 strain at the 0, 2, 4, 5, 6, and 8 hr
time points, as well as from a paf1D strain, and probed
with several probes for genes in both the decreased
and increased categories. RNA analysis for genes whose
expression decreased (MAK21) or increased (AHA1) is
shown in Figure 3C. Quantifications of MAK21, AHA1,
TRM8, and ENA2 are shown in Figure 3D. For all tran-
scripts tested, the Northern blots confirmed the micro-
Remarkably, we found that 68% (11 of 16) of the Paf1
targets with decreased expression are essential genes.
This is in contrast to the yeast genomic average of
about 20% (Csank et al., 2002). This helps to explain
the severe phenotypes associated with paf1D. In par-
ticular, nine of these essential genes (NOG2, MAK21,
ESF1, CGR1, SDA1, RRP12, NOP53, YJL010C, and
YOR004W) play important roles in nucleolar structure
and function and the processing of rRNA. Consistent
with the reduced expression of these genes, loss of
Paf1 results in rRNA processing defects (Porter et al.,
2005). In marked contrast, none of the 41 target genes
that increase in expression upon loss of Paf1 is
Primary Paf1 Target Promoters Are Not
Sensitive to Loss of Paf1
To ask if Paf1 affects transcription through the promot-
ers of its primary targets, we created promoter/reporter
constructs driving luciferase activity for several of the
genes listed in Figures 3A and 3B. Because the promot-
ers of most of the genes examined are uncharacterized,
we inserted the entire upstream intergenic region into
the reporter vectors. The promoter regions of genes
whose expression decreased (SDA1, TRM8, MAK21,
RRP12, ATF2, and NOG2) and increased (AHA1, ARO10,
and MAG1) were analyzed after integration of the lucif-
erase constructs into the genome of wt and paf1D
strains. We observed that luciferase activity driven by
the TRM8, MAK21, SDA1, RRP12, and ATF2 promoters
was unchanged in paf1D, whereas the NOG2 promoter
activity increased slightly (Figure 4A). Similarly, we ob-
served no increase in luciferase activity driven from the
promoters of primary targets whose expression in-
creased in paf1D (Figure 4B). We searched for common
sequences within the promoter regions of the Paf1 tar-
get genes by using the Regulatory Sequence Analysis
tools program found at http://itzamna.cifn.unam.mx/
wjvanheld/rsa-tools/. By using the default parameters
set by the program, we found no significant consensus
sequences greater than 4 bp in the promoters of the
Figure 2. Shutoff of Paf1 Function Using a Tetracycline-Regulated Promoter
(A) Schematic of the tetracycline-regulated PAF1 expression construct.
(B) Wt (YJJ662), paf1D (YJJ576), and tetO2-PAF1-6HA (YJJ1653) were streaked onto YPD medium with or without DOX and/or caffeine and
grown at the indicated temperatures. Photographs of plates were taken after 4 days incubation.
(C) Growth curve of the indicated strains grown in YPD medium with 30 ?g/ml DOX as described in the Experimental Procedures.
(D) Cells were harvested at the indicated times post DOX addition, and HA-Paf1 was detected with an anti-HA antibody as described in the
primary target genes. These data, taken together with
the results of the promoter/reporter assays, suggest
that the Paf1 requirement for correct expression of its
target genes is not through its promoters.
Transcripts Sensitive to Loss of Paf1 Have
an Unusual Pattern of mRNA Stability
Because the Paf1C is associated with Pol II at promot-
ers and in coding regions of genes, it has been classi-
fied as an elongation factor (Gerber and Shilatifard,
2003). However, loss of Paf1 does not appear to change
the abundance or distribution of Pol II along a gene
as might be expected if the complex were critical for
elongation (Mueller et al., 2004). Hpr1, found associ-
ated with the Paf1C (Chang et al., 1999; Mueller and
Jaehning, 2002), has also been described as an elonga-
tion factor (Chavez and Aguilera, 1997), and its loss has
been linked to defects in the expression of long- or
G + C-rich transcription units (Chavez et al., 2001). We
analyzed both the primary and secondary Paf1 targets
and found no evidence that these genes were on
average either longer or more G + C rich than the yeast
genome average (1400 nucleotides and 40% G + C; D.
Reines, personal communication). However, we did ob-
serve that the half-lives of the primary, but not the sec-
ondary, targets were obviously different than the ge-
nome-wide average. The average yeast mRNA half-life
is about 21 min (Holstege et al., 1998; Wang et al.,
2002). In contrast, the average half-life of transcripts
whose abundance either decreased or increased with-
out Paf1 is 12 min and over 40 min, respectively (as
determined from published data sets [Holstege et al.,
1998; Wang et al., 2002]). These marked differences in
mRNA stability led us to ask if loss of the Paf1C might
result in changes in posttranscriptional rather than tran-
Loss of Paf1 Results in Changes in Poly(A)
We focused on the Paf1 primary targets with decreased
expression, because this unique collection of short-
lived, primarily essential genes serves to explain the
severe phenotypes resulting from loss of Paf1. In par-
ticular, we examined expression of SDA1 and MAK21
(NOC1), both essential genes important for ribosome
biogenesis (Gavin et al., 2002; Milkereit et al., 2001).
Normal Poly(A) Site Utilization Requires Paf1
Figure 3. Identification of Paf1 Primary Tar-
gets after Shut Off of Tet-Regulated Paf1
(A) Genes dependent upon Paf1 for full ex-
pression. RNA isolated from time points at 4,
5, and 6 hr after the addition of DOX to the
tetO2-PAF1-6HA strain was hybridized to Af-
fymetrix microarrays as described in the Ex-
perimental Procedures. Primary Paf1-depen-
dent targets were identified as described in
(B) Genes whose expression increases in the
absence of Paf1. Targets were identified as
(C) Confirmation of the microarray results.
RNA was isolated from a second time course
like that used to isolate RNA for the microar-
rays. Ten micrograms of RNA was analyzed
as described in the Experimental Proce-
dures. A representative 18S rRNA loading
control is also shown.
(D) Quantification of Northern blots. The RNA
expression levels of MAK21 and AHA1, as
well as similar TRM8 and ENA2 blots, were
normalized to 18S rRNA.
Based on a probabilistic prediction of 3# ends (Graber
et al., 2002), both genes have two poly(A) sites in their
3# noncoding region (Figures 5A and 5C). Use of these
sites would result in predicted transcripts of about 2400
and 2600 nucleotides for the 2.3 kb SDA1 gene and
3000 and 3300 nucleotides for the 3 kb MAK21 gene.
When total RNA was analyzed, we observed a single
transcript for each gene (Figures 5B and 5D, left); how-
ever, in poly(A)+RNA samples, we detected transcripts
corresponding to utilization of both predicted poly(A)
sites (Figures 5B and 5D, middle).
The poly(A)+SDA1 transcripts are remarkable for two
Figure 4. Paf1-Primary Target Promoters Do
Not Respond to Loss of Paf1
Luciferase constructs containing the pro-
moters of primary targets (left) whose ex-
pression decreases (TRM8, MAK21, SDA1,
RRP12, ATF2, and NOG2) and (right) whose
expression increases (ARO10, MAG1, and
AHA1) were transformed into wt (YJJ1366)
and paf1D (YJJ1368). Extracts were pre-
pared, and luciferase activity was measured
as described in the Experimental Proce-
dures. The data are presented as the aver-
age ± the SD from the mean of quadrupli-
reasons. First, the paf1D-dependent 2- to 3-fold de-
creased abundance of SDA1 mRNA seen in total RNA
and in the microarrays is not seen in the poly(A)+sam-
ples (Figure 5B, compare left and middle), implying that
a significant fraction of SDA1 mRNA may not be polya-
denylated. Second, in the paf1D poly(A)+samples, not
only is there an increased amount of transcript extend-
ing to the second SDA1 poly(A) site(s) (a cluster of sites
are predicted [Figure 5A] corresponding to the smear
of transcripts marked by the open arrowhead), but we
also detected a 4000 nucleotide RNA consistent with
extension to the downstream LSC2 poly(A) site (gray
Figure 5. Loss of Paf1 Results in Altered Poly(A) Site Utilization
(A) Schematic of the SDA1 gene showing the position of predicted
poly(A) sites (Graber et al., 2002) and probes used for transcript
(B) Total (left) and poly(A)+(middle) RNA was probed for SDA1
mRNA. The filled, open, and gray arrowheads mark bands in the
poly(A)+RNA corresponding to the predicted 3# ends shown in (A).
The asterisk marks a band not corresponding to predicted poly(A)
sites. RT-PCR was used to assay the relative abundance of tran-
scripts extending beyond the SDA1 poly(A) sites (right).
(C) Schematic of the MAK21 gene showing the position of pre-
dicted poly(A) sites (Graber et al., 2002) and probes used for tran-
script analysis. Abbreviation: NP, northern probe.
(D) Total (left) and poly(A)+(middle) RNA was probed for MAK21
mRNA with the indicated NPs. NP1 hybridizes to both transcripts;
NP2, spanning poly(A) site 2, is specific for the longer transcript.
The closed and open arrowheads mark bands in the poly(A)+RNA
corresponding to the predicted 3# ends shown in (A). RT-PCR was
used to assay the relative abundance of transcripts extending be-
yond the proximal MAK21 poly(A) site (right).
arrowhead, Figure 5B, middle). The increased abun-
dance of the transcripts extending beyond the SDA1
poly(A) sites in paf1D is confirmed by the RT-PCR
analysis shown in the right panel of Figure 5B.
For MAK21, we found reduced expression of the
transcript extending to the first poly(A) site in both total
and poly(A)+RNA from paf1D and an increased ratio of
transcripts extending to the second predicted poly(A)
site (open arrowhead, Figure 5D, middle, NP1). By using
a probe spanning the distal poly(A) site, we confirmed
that the longer transcript extends beyond the proximal
site (Figure 5D, middle, NP2). RT-PCR was used to con-
firm the existence of this extended transcript and its
increased abundance in paf1D relative to wt (Figure 5D,
right). These results demonstrate that loss of the Paf1C
results in reduced utilization of some proximal poly(A)
sites and increased production of longer transcripts ex-
tending to distal poly(A) sites.
Loss of Paf1 Creates MAK21 Transcripts
Sensitive to NMD
Loss of Paf1 results in changes in poly(A) site utiliza-
tion, but this does not explain why SDA1 and MAK21
transcript abundance is reduced in total RNA from
paf1D. We asked whether the reduced abundance was
due to the formation of targets for mRNA surveillance
mechanisms. This hypothesis was based in part on the
links between Paf1 and Hpr1 described above. Deletion
of HPR1 results in defects in poly(A)+mRNA export
(Schneiter et al., 1999), leading to nuclear retention and
degradation by the nuclear exonuclease Rrp6 (Libri et
al., 2002; Zenklusen et al., 2002). Because loss of Hpr1
can be partially suppressed by loss of rrp6 (Libri et al.,
2002), we asked whether the decreased abundance of
Paf1 target genes in paf1D could be restored to wt
levels when combined with deletion of RRP6. Mutant
strains were constructed, and the abundance of several
Paf1 target genes was analyzed in isogenic wt, paf1D,
and rrp6D strains and a viable but slow growing paf1D
rrp6D double mutant. Unlike hpr1D, we did not observe
suppression of paf1D-dependent transcript abundance
defects by loss of Rrp6 (data not shown). These results
do not rule out a role for the Paf1C in mRNA export but
prompted us to look at the effects of another mRNA
surveillance pathway on the abundance of Paf1 target
The NMD pathway degrades cytoplasmic mRNAs
with abnormally extended 3# UTRs (Muhlrad and Par-
ker, 1999). Because the loss of Paf1 results in increased
production of 3#-extended SDA1 and MAK21 tran-
scripts, we asked if deletion of UPF1 (NAM7), encoding
an NMD-required helicase (Czaplinski et al., 1995),
would restore normal transcript levels in combination
with paf1D. We compared the abundance of SDA1 and
MAK21 mRNAs in wt, paf1D, upf1D, and paf1D upf1D
strains. Although we did not detect suppression of
paf1D growth defects or SDA1 transcript levels in the
paf1D upf1D double mutant (data not shown), the re-
duced abundance of MAK21 mRNA was suppressed by
loss of UPF1 (Figure 6). This is clear both in Northern
analysis of total RNA shown in Figure 6B, where MAK21
mRNA abundance is restored to levels nearly 3-fold
above those found in wt in the paf1D upf1D double mu-
tant and in the RT-PCR reactions amplifying the 5# and
3# ends of the MAK21 ORF (Figure 6C).
Using RT-PCR to amplify regions beyond the first pre-
dicted MAK21 poly(A) site revealed that there is nor-
mally very low expression of this region in wt cells; loss
of Paf1 increases expression in this region 2-fold, and
abundance is elevated 4- to 5-fold relative to wt in
upf1D and in the paf1D upf1D double mutant (Figure
6D). Finally, to confirm that the extended transcripts
originate in the MAK21 ORF, we queried with RT-PCR
primers extending from the ORF to the downstream po-
ly(A) site; the increased presence of these transcripts
is clear in paf1D, upf1D, and paf1D upf1D strains (Fig-
ure 6E). These results confirm that loss of Paf1 results
in altered poly(A) site utilization and that the extended
transcripts created can be a target of NMD. Therefore,
the transcript changes that we observe for some Paf1
primary targets are not the result of changes in tran-
scription initiation or elongation but are consistent with
posttranscriptional effects resulting from increased for-
mation of unstable, 3#-extended RNAs.
Normal Poly(A) Site Utilization Requires Paf1
Figure 6. Loss of Paf1 Results in Increased
Formation of an Unstable, Extended MAK21
Transcript Sensitive to NMD
(A) An expanded diagram of the MAK21 gene
showing probes and primers used for North-
ern and RT-PCR analysis of MAK21 tran-
(B) Northern analysis of MAK21 RNA. Total
RNA was isolated from wt (YJJ662), paf1D
(YJJ576), upf1D (YJJ1757), and paf1D upf1D
(YJJ1759). Ten micrograms was separated
and analyzed as described in the Experi-
mental Procedures (top). The results were
quantified and normalized to 18S rRNA (bot-
tom). The bars represent the average of two
independent blots, and the error bars indi-
cate the range.
(C–E) RT-PCR reactions were performed as
described in the Experimental Procedures.
The numbers above each panel represent
the primers used for amplification as shown
in (A). 18S rRNA was used for normalization.
(C) RT-PCR analysis of transcripts from the
5# + 3# ends of the MAK21 ORF. (D) RT-PCR
of regions downstream of the proximal
MAK21 poly(A) site. (E) RT-PCR of a region
connecting the extended transcripts to the
MAK21 ORF. In (D) and (E), the data are pre-
sented as the average ± the SD of the mean
from triplicate assays.
Although the Paf1C is associated with pol II at promot-
ers and throughout the coding region of transcribed
yeast genes, its loss affects the abundance of only a
small subset of transcripts. In this work we have dem-
onstrated that Paf1 dependence is not linked to the
promoters of target genes and therefore is not a conse-
quence of changes in initiation. We have also not been
able to detect any gene-specific deficiencies in tran-
scriptional elongation (this work and Mueller et al.
). Instead, for the Paf1-dependent MAK21 tran-
script, abundance changes are caused by decreased
utilization of a proximal poly(A) site and the increased
formation of an extended transcript subject to NMD.
Therefore, the differential effects of the Paf1C on tran-
script abundance are probably linked to a role of the
Paf1C in recruitment or activity of cleavage and poly(A)
factors. This linkage is at least in part through the
requirement of Paf1 for full levels of Pol II CTD Ser2
phosphorylation (Mueller et al., 2004). This modification
is important for the association of cleavage and poly(A)
factors with Pol II (Ahn et al., 2004; Licatalosi et al.,
2002), and loss of Paf1 has been shown to reduce both
CTD Ser2 phosphorylation and association of the
essential cleavage and poly(A) factor Pcf11 with chro-
matin (Mueller et al., 2004).
However, if the only significant consequence of loss
of Paf1 was reduced Ser2 phosphorylation, then its loss
should be equivalent to that of loss of the Ser2 kinase
Ctk1. This supposition predicts that loss of both Paf1
and Ctk1 would be no more deleterious than loss of
just Ctk1. But, mutations in PAF1 and CTK1 are lethal in
combination (Squazzo et al., 2002; M.S.H., unpublished
data), refuting the hypothesis that Paf1’s only function
is in CTD Ser2 phosphorylation. In addition, we have
shown that loss of Rtf1 and the subsequent dissoci-
ation of the Paf1C from Pol II and chromatin (Mueller
et al., 2004) also results in decreased utilization of the
proximal MAK21 poly(A) site under conditions where
Ser2 phosphorylation is only slightly diminished (Kris-
ten Nordick, personal communication). These observa-
tions lead us to propose that the Paf1C is critical for
proper mRNA cleavage and polyadenylation indepen-
dent of its effects on CTD-Ser2 phosphorylation.
The Paf1C has been physically and genetically linked
to Hpr1, a factor connected to both transcriptional
elongation and nuclear export of mRNAs (Chang et al.,
1999; Chavez and Aguilera, 1997; Strasser et al., 2002).
When the Paf1C is detached from Pol II, by loss of
either Rtf1 or Cdc73 (Mueller et al., 2004), it is found in
the nucleolus, possibly due to an association with
mRNAs in the export pathway (Porter et al., 2005). Loss
of Hpr1 results in changes in polyadenylation and
nuclear retention of improperly processed transcripts
(Libri et al., 2002; Schneiter et al., 1999; Zenklusen et
al., 2002). Some hpr1D effects can be suppressed by
loss of the nuclear exosome component Rrp6 (Libri et
al., 2002; Zenklusen et al., 2002). We did not observe
that that loss of Rrp6 reversed the effects of loss of
Paf1 on its target genes. Instead, upf1D suppresses the
paf1D effects on MAK21, indicating that the extended
MAK21 transcripts formed in paf1D must be success-
fully exported to the cytoplasm where they associate
with the ribosome and become targets for NMD. Unlike
MAK21, transcripts from SDA1 and several other Paf1-
primary targets are not sensitive to disruption of NMD
or the nuclear exosome (data not shown). It may be that
loss of Paf1 creates unstable extended transcripts from
these genes that are the target of an as yet unknown
Loss of Paf1 does not affect the distribution of Pol II
on genes (Mueller et al., 2004), and neither the primary
nor the secondary Paf1 targets are longer or more G +
C rich than the genome-wide average, but it remains
possible that the Paf1C directly affects Pol II elonga-
tion. Whether this is a direct effect of the loss of Paf1C
components or a secondary consequence of the reduc-
tion of CTD Ser2 phosphorylation, changes in elonga-
tion rate may be linked to the changes in poly(A) site
utilization that we have observed in Paf1-target genes.
It is interesting to note in this regard that changes in
poly(A) site utilization have been correlated with de-
fects in elongation; however, the pattern observed is
opposite to our observations. Cui and Denis (2003)
found that elongation-deficient mutations in Rpb2,
Dst1, and Spt5 result in increased utilization of proximal
poly(A) sites rather than the decreased utilization we
have observed in the absence of Paf1. This suggests
that if the Paf1C has a role in elongation, its role is dis-
tinct from that of known elongation factors Dst1 and
Paf1 targets are a special subset of yeast genes, possi-
bly with a unique class of poly(A) sites particularly sen-
sitive to loss of CTD Ser2 phosphorylation, reduced as-
sociation with cleavage and poly(A) factors, or loss of
the Paf1C. The genes identified as primary Paf1 targets
are remarkable in that most are essential (11/16), many
are required for nucleolar function and rRNA process-
ing, and they have average half-lives of only about 12
min, much shorter than the genome-wide average of 21
min (Holstege et al., 1998; Wang et al., 2002). A connec-
tion between mRNA stability and ribosome biogenesis
factors has recently been reported (Garcia-Martinez et
al., 2004; Grigull et al., 2004), but there is only a small
overlap between the genes identified in those studies
and the Paf1-dependent genes identified here. It will be
interesting to discover if all of these short-lived mRNAs
share some particular features of their poly(A) sites that
result in the formation of extended and unstable tran-
scripts under conditions of stress, including loss of the
Paf1C. As also concluded by Grigull et al. (2004), we
have not been able to identify any common elements
in the 3# UTRs of the Paf1-dependent genes that help
to explain their particular sensitivity to loss of Paf1. We
speculate that Paf1 sensitivity is the property of a sub-
set of genes that normally have relatively poor utiliza-
tion of their nearest poly(A) site. As we have presented
for the MAK21 gene, transcripts extending to the distal
poly(A) site can be detected even in wt cells. Although
the utilization of the proximal MAK21 poly(A) site is re-
duced without Paf1, we do not see the appearance of
transcripts beyond the distal poly(A) site (data not
shown). In contrast, for SDA1, which also has two pre-
dicted poly(A) sites, we observe transcripts that extend
beyond these sites and through the downstream gene
to the next available poly(A) site. Because we have not
been able to detect extended (and probably highly
unstable) transcripts from all of the Paf1 targets, it re-
mains possible that loss of the Paf1 complex affects
other aspects of RNA processing for some target genes.
The primary target genes whose transcripts are in-
creased in response to loss of Paf1 are also remarkable
in that none is essential and their average half-lives are
over 40 min, twice the genome-wide average. Although
about half of these transcripts are also induced by
some stress conditions (similar to the genome-wide
average [Causton et al., 2001]), they appear to be a
unique subset of genes. Both primary and secondary
target promoter fusions that we tested for Paf1 depen-
dence were negative, implying that the secondary tar-
gets may also be sensitive to posttranscriptional events
rather than initiation or elongation of transcription. The
secondary targets may therefore include an additional
population of genes whose proper 3# end formation de-
pends on the Paf1C.
The current studies emphasize the importance of the
Paf1C for the proper utilization of a subset of poly(A)
sites as a mechanism for observed changes in tran-
script abundance and an explanation for the deleteri-
ous phenotypes associated with loss of Paf1. However,
the Paf1C is associated with Pol II from the promoter
to the poly(A) site of all transcribed genes (Ahn et al.,
2004; Mueller et al., 2004), and its loss affects other
cotranscriptional processes. Therefore, the full spectrum
of the Paf1C’s activities is still to be discovered.
Yeast Strains and Growth Conditions
The isogenic yeast strains used in this study were derived from
YJJ662 (MATa leu2D1 his3D200 ura3-52) (Shi et al., 1997) and in-
clude: YJJ576 (paf1D::HIS3), YJJ664 (paf1D::HIS3),YJJ1197 (ctr9D::
Kanr), YJJ1366 (trp1D::hisG), YJJ1368 (paf1D::HIS3 trpD::hisG),
YJJ1598 (PAF1::PAF16HA[kITRP1]), YJJ1653 (kanr::tetO2-PAF1::
PAF16HA([kITRP1]),YJJ1605 (rrp6D::Kanr), YJJ1607 (rrp6D::Kanr
paf1D::HIS3), YJJ1757 (upf1D::Kanr), and YJJ1759 (upf1D::Kanr
paf1D::HIS3). Kanrdeletion/disruptions of RRP6 and UPF1 were
constructed in a PAF1/paf1D heterodiploid strain (YJJ1035) fol-
lowed by sporulation and identification of haploid single and
double mutants by tetrad dissection (Betz et al., 2002). Details of
the hemagglutinin (HA)-tagged and tetO2-PAF1-6HA strains are de-
scribed below. Strains were grown in yeast extract-peptone-
dextrose (YPD, 4% dextrose) by using standard methods (Guthrie
and Fink, 1991).
HA-Tagged Paf1 Strain Construction
Paf1 was C-terminally epitope tagged as described (Porter et al.,
2005). The tetO2-PAF1-6HA strain was constructed by using the
tetracycline operator-based promoter substitution cassette plas-
mid pCM224 ([Belli et al., 1998], deposited at EUROSCARF, Frank-
Normal Poly(A) Site Utilization Requires Paf1
furt, Germany), containing the kanMX4 selectable marker. The pro-
moter substitution cassette was PCR amplified with a 5# oligo-
nucleotide containing a 40 bp sequence from 200 bp upstream of
the PAF1 initiation codon, followed by 19 bp complementary to the
cassette multiple cloning site. The 3# oligonucleotide contained 40
bp of sequence from upstream, and including, the PAF1 initiation
codon followed by 19 bp of the cassette multiple cloning site.
YJJ1598 was transformed with the PCR product to create YJJ1653,
containing HA-tagged PAF1 controlled by the tetracycline-regu-
Growth of tetO2-PAF1-6HA Strain and Analysis
of Paf1 Expression
Cells were grown in YPD to 1 × 106cells/ml, 30 ?g/ml DOX was
added, and duplicate 10 ml samples were taken at 1 hr intervals.
At 5 × 106cells/ml, the cells were diluted to 1 × 106cells/ml with
fresh medium plus 30 ?g/ml DOX, and samples were taken at 1 hr
intervals for 18 hr. One set of 10 ml aliquots was used for protein
isolation with Y-PER yeast protein extraction reagent (Pierce). 30
?g total protein was separated on a 4%–12% Bis-Tris NuPAGE gel
(Invitrogen), and Paf1 expression was determined by Western blot-
ting using a 12CA5 anti-HA antibody (Roche) (Mueller and Jaehn-
Oligonucleotide Microarray Analysis
The other set of 10 ml time course aliquots and samples from wt,
paf1D, and ctr9D strains were used for RNA isolation by using the
hot phenol method (Schmitt et al., 1990). Total RNA was purified
over RNeasy columns (Qiagen). Duplicate samples of 8 ?g total
RNA were used to prepare Biotin-labeled cRNA probes, which were
hybridized to Affymetrix Yeast Genome S98 GeneChip arrays ac-
cording to Affymetrix protocols. The microarrays were washed and
stained with R-Phycoerythrin-linked streptavidin. The arrays were
scanned, and signals were detected by using a Hewlett Packard
GeneArray scanner. Data were analyzed by using Affymetrix Micro-
array Suite version 5.0 with the set default parameters.
10 ?g total RNA was fractionated by using either the glyoxal or the
formaldehyde RNA denaturation method (Ausubel et al., 1994) or
by 1 × TBE gel electrophoresis as described (Kevil et al., 1997).
RNA was transferred to Zeta-Probe GT membranes by capillary (for
glyoxal and formaldehyde methods) or electrophoretic (Kevil et al.,
1997) transfer.32P-labeled probes were made by PCR amplification
of specific yeast genes from genomic yeast DNA, followed by gel
purification and random prime labeling using a RadPrime Kit (Invit-
rogen). The 18S rRNA oligonucleotide was 5# end labeled (Ausubel
et al., 1994). Hybridization and washing were as described (Porter
et al., 2002). Blots were exposed to a phosphorimaging screen, and
signals were quantitated by using Quantity One imaging software
and normalized to 18S rRNA. Polyadenylated RNA (poly[A]+) was
selected from 750 ?g of purified total RNA by using an Oligotex
mRNA midi kit (Qiagen). 1 ?g of poly(A)+RNA was fractionated and
analyzed by using the formaldehyde denaturation method de-
Reverse transcription reactions were performed by using a Super-
script III Kit (Invitrogen). cDNA synthesis from 2 ?g of total RNA
was primed with random hexamers. PCR amplifications used 1/10
of the reverse transcription reaction as template. Linear range PCR
conditions were determined for each set of primers used, and sig-
nals were corrected for contaminating DNA. RT-PCR reactions also
contained primers to amplify 18S rRNA for normalization.
Promoter/Reporter (Luciferase) Assays
Luciferase reporter constructs used plasmid pMTLuc (pJJ1316)
provided by D. Reines (Emory University) (Shaw and Reines, 2000).
The CEN/ARS-containing ScaI fragment of pMTLuc was replaced
with the ScaI fragment of pRS306 (Sikorski and Boeke, 1991) to
generate an integrating plasmid, pJJ1358, which when cleaved at
the unique NcoI site, can be integrated into the URA3 locus. The
entire upstream intergenic region of ARO10, MAG1, AHA1, TAD2,
TRM8, MAK21, RRP12, ATF2, NOG2, and SDA1 was PCR amplified
from yeast DNA (oligo sequences available on request). Each PCR
product was cloned into pJJ1358, and the resulting plasmid was
cleaved with NcoI and transformed into YJJ1366 and YJJ1368 by
using standard transformation procedures (Guthrie and Fink, 1991).
Luciferase assays were performed according to the Promega lucif-
erase assay system as described (Porter et al., 2002), with the data
presented as the average relative luciferase units per milligram of
total protein (measured using Bio-Rad protein assay reagents) ±
the SD from the mean from quadruplicate assays.
Supplemental Data include one table and are available with this
article online at http://www.molecule.org/cgi/content/full/20/2/213/
We thank T. Blumenthal and D. Bentley for many useful sugges-
tions, D. Reines for a database of yeast gene length and G + C
content and the luciferase reporter construct, and J. Betz and E.
Amiott for critically reading the text. This work was supported by
National Institutes of Health grant RO1-GM38101 to J.A.J.
Received: April 11, 2005
Revised: June 3, 2005
Accepted: August 22, 2005
Published: October 27, 2005
Ahn, S.H., Kim, M., and Buratowski, S. (2004). Phosphorylation of
serine 2 within the RNA polymerase II C-terminal domain couples
transcription and 3# end processing. Mol. Cell 13, 67–76.
Ausubel, F.M., Brent, R., Kingston, R.E., Moore, D.D., Seidman,
J.G., Smith, J.A., and Struhl, K. (1994). Current Protocols in Molecu-
lar Biology (New York: John Wiley & Sons).
Belli, G., Gari, E., Aldea, M., and Herrero, E. (1998). Functional
analysis of yeast essential genes using a promoter-substitution
cassette and the tetracycline-regulatable dual expression system.
Yeast 14, 1127–1138.
Betz, J.L., Chang, M., Washburn, T.M., Porter, S.E., Mueller, C.L.,
and Jaehning, J.A. (2002). Phenotypic analysis of Paf1/RNA poly-
merase II complex mutations reveals connections to cell cycle reg-
ulation, protein synthesis, and lipid and nucleic acid metabolism.
Mol. Genet. Genomics 268, 272–285.
Buratowski, S. (2003). The CTD code. Nat. Struct. Biol. 10, 679–680.
Causton, H.C., Ren, B., Koh, S.S., Harbison, C.T., Kanin, E., Jen-
nings, E.G., Lee, T.I., True, H.L., Lander, E.S., and Young, R.A.
(2001). Remodeling of yeast genome expression in response to en-
vironmental changes. Mol. Biol. Cell 12, 323–337.
Chang, M., French-Cornay, D., Fan, H.Y., Klein, H., Denis, C.L., and
Jaehning, J.A. (1999). A complex containing RNA polymerase II,
Paf1p, Cdc73p, Hpr1p, and Ccr4p plays a role in protein kinase C
signaling. Mol. Cell. Biol. 19, 1056–1067.
Chavez, S., and Aguilera, A. (1997). The yeast HPR1 gene has a
functional role in transcriptional elongation that uncovers a novel
source of genome instability. Genes Dev. 11, 3459–3470.
Chavez, S., Garcia-Rubio, M., Prado, F., and Aguilera, A. (2001).
Hpr1 is preferentially required for transcription of either long or
G+C-rich DNA sequences in Saccharomyces cerevisiae. Mol. Cell.
Biol. 21, 7054–7064.
Csank, C., Costanzo, M.C., Hirschman, J., Hodges, P., Kranz, J.E.,
Mangan, M., O’Neill, K., Robertson, L.S., Skrzypek, M.S., Brooks,
J., and Garrels, J.I. (2002). Three yeast proteome databases: YPD,
PombePD, and CalPD (MycoPathPD). Methods Enzymol. 350,
Cui, Y., and Denis, C.L. (2003). In vivo evidence that defects in the
transcriptional elongation factors RPB2, TFIIS, and SPT5 enhance
upstream poly(A) site utilization. Mol. Cell. Biol. 23, 7887–7901.
Czaplinski, K., Weng, Y., Hagan, K.W., and Peltz, S.W. (1995). Purifi-
cation and characterization of the Upf1 protein: a factor involved in
translation and mRNA degradation. RNA 1, 610–623.
Dvir, A., Conaway, J.W., and Conaway, R.C. (2001). Mechanism of
transcription initiation and promoter escape by RNA polymerase II.
Curr. Opin. Genet. Dev. 11, 209–214.
Garcia-Martinez, J., Aranda, A., and Perez-Ortin, J.E. (2004). Geno-
mic run-on evaluates transcription rates for all yeast genes and
identifies gene regulatory mechanisms. Mol. Cell 15, 303–313.
Gari, E., Piedrafita, L., Aldea, M., and Herrero, E. (1997). A set of
vectors with a tetracycline-regulatable promoter system for modu-
lated gene expression in Saccharomyces cerevisiae. Yeast 13,
Gavin, A.C., Bosche, M., Krause, R., Grandi, P., Marzioch, M.,
Bauer, A., Schultz, J., Rick, J.M., Michon, A.M., Cruciat, C.M., et al.
(2002). Functional organization of the yeast proteome by system-
atic analysis of protein complexes. Nature 415, 141–147.
Gerber, M., and Shilatifard, A. (2003). Transcriptional elongation by
RNA polymerase II and histone methylation. J. Biol. Chem. 278,
Graber, J.H., McAllister, G.D., and Smith, T.F. (2002). Probabilistic
prediction of Saccharomyces cerevisiae mRNA 3#-processing
sites. Nucleic Acids Res. 30, 1851–1858.
Grigull, J., Mnaimneh, S., Pootoolal, J., Robinson, M.D., and
Hughes, T.R. (2004). Genome-wide analysis of mRNA stability using
transcription inhibitors and microarrays reveals posttranscriptional
control of ribosome biogenesis factors. Mol. Cell. Biol. 24, 5534–
Guthrie, C., and Fink, G.R. (1991). Guide to Yeast Genetics and
Molecular Biology, Volume 194 (San Diego, CA: Academic Press).
Hahn, S. (2004). Structure and mechanism of the RNA polymerase
II transcription machinery. Nat. Struct. Mol. Biol. 11, 394–403.
Holstege, F.C., Jennings, E.G., Wyrick, J.J., Lee, T.I., Hengartner,
C.J., Green, M.R., Golub, T.R., Lander, E.S., and Young, R.A. (1998).
Dissecting the regulatory circuitry of a eukaryotic genome. Cell 95,
Jensen, T.H., Dower, K., Libri, D., and Rosbash, M. (2003). Early
formation of mRNP: license for export or quality control? Mol. Cell
Jensen, T.H., Boulay, J., Olesen, J.R., Colin, J., Weyler, M., and Li-
bri, D. (2004). Modulation of transcription affects mRNP quality.
Mol. Cell 16, 235–244.
Kevil, C.G., Walsh, L., Laroux, F.S., Kalogeris, T., Grisham, M.B.,
and Alexander, J.S. (1997). An improved, rapid Northern protocol.
Biochem. Biophys. Res. Commun. 238, 277–279.
Koch, C., Wollmann, P., Dahl, M., and Lottspeich, F. (1999). A role
for Ctr9p and Paf1p in the regulation G1 cyclin expression in yeast.
Nucleic Acids Res. 27, 2126–2134.
Kornberg, R.D. (2005). Mediator and the mechanism of transcrip-
tional activation. Trends Biochem. Sci. 30, 235–239.
Krogan, N.J., Kim, M., Ahn, S.H., Zhong, G., Kobor, M.S., Cagney,
G., Emili, A., Shilatifard, A., Buratowski, S., and Greenblatt, J.F.
(2002). RNA polymerase II elongation factors of Saccharomyces
cerevisiae: a targeted proteomics approach. Mol. Cell. Biol. 22,
Libri, D., Dower, K., Boulay, J., Thomsen, R., Rosbash, M., and Jen-
sen, T.H. (2002). Interactions between mRNA export commitment,
3#-end quality control, and nuclear degradation. Mol. Cell. Biol. 22,
Licatalosi, D.D., Geiger, G., Minet, M., Schroeder, S., Cilli, K.,
McNeil, J.B., and Bentley, D.L. (2002). Functional interaction of
yeast pre-mRNA 3# end processing factors with RNA polymerase
II. Mol. Cell 9, 1101–1111.
Milkereit, P., Gadal, O., Podtelejnikov, A., Trumtel, S., Gas, N., Pet-
falski, E., Tollervey, D., Mann, M., Hurt, E., and Tschochner, H.
(2001). Maturation and intranuclear transport of pre-ribosomes re-
quires Noc proteins. Cell 105, 499–509.
Mueller, C.L., and Jaehning, J.A. (2002). Ctr9, Rtf1, and Leo1 are
components of the Paf1/RNA polymerase II complex. Mol. Cell.
Biol. 22, 1971–1980.
Mueller, C.L., Porter, S.E., Hoffman, M.G., and Jaehning, J.A. (2004).
The Paf1 complex has functions independent of actively transcrib-
ing RNA polymerase II. Mol. Cell 14, 447–456.
Muhlrad, D., and Parker, R. (1999). Aberrant mRNAs with extended
3# UTRs are substrates for rapid degradation by mRNA surveil-
lance. RNA 5, 1299–1307.
Neugebauer, K.M. (2002). On the importance of being co-transcrip-
tional. J. Cell Sci. 115, 3865–3871.
Palancade, B., and Bensaude, O. (2003). Investigating RNA poly-
merase II carboxyl-terminal domain (CTD) phosphorylation. Eur. J.
Biochem. 270, 3859–3870.
Peltz, S.W., Brown, A.H., and Jacobson, A. (1993). mRNA destabili-
zation triggered by premature translational termination depends on
at least three cis-acting sequence elements and one trans-acting
factor. Genes Dev. 7, 1737–1754.
Porter, S.E., Washburn, T.M., Chang, M., and Jaehning, J.A. (2002).
The yeast pafl-RNA polymerase II complex is required for full ex-
pression of a subset of cell cycle-regulated genes. Eukaryot. Cell
Porter, S.E., Penheiter, K.L., and Jaehning, J.A. (2005). Separation
of the Saccharomyces cerevisiae Paf1 complex from RNA polymer-
ase II results in changes in its subnuclear localization. Eukaryot.
Cell 4, 209–220.
Rozenblatt-Rosen, O., Hughes, C.M., Nannepaga, S.J., Shanmu-
gam, K.S., Copeland, T.D., Guszczynski, T., Resau, J.H., and Meyer-
son, M. (2005). The parafibromin tumor suppressor protein is part
of a human paf1 complex. Mol. Cell. Biol. 25, 612–620.
Schmitt, M.E., Brown, T.A., and Trumpower, B.L. (1990). A rapid and
simple method for preparation of RNA from Saccharomyces cere-
visiae. Nucleic Acids Res. 18, 3091–3092.
Schneiter, R., Guerra, C.E., Lampl, M., Gogg, G., Kohlwein, S.D.,
and Klein, H.L. (1999). The Saccharomyces cerevisiae hyperrecom-
bination mutant hpr1D is synthetically lethal with two conditional
alleles of the acetyl coenzyme A carboxylase gene and causes a
defect in nuclear export of polyadenylated RNA. Mol. Cell. Biol. 19,
Shaw, R.J., and Reines, D. (2000). Saccharomyces cerevisiae tran-
scription elongation mutants are defective in PUR5 induction in re-
sponse to nucleotide depletion. Mol. Cell. Biol. 20, 7427–7437.
Shi, X., Chang, M., Wolf, A.J., Chang, C.H., Frazer-Abel, A.A., Wade,
P.A., Burton, Z.F., and Jaehning, J.A. (1997). Cdc73p and Paf1p are
found in a novel RNA polymerase II-containing complex distinct
from the Srbp-containing holoenzyme. Mol. Cell. Biol. 17, 1160–
Sikorski, R.S., and Boeke, J.D. (1991). In vitro mutagenesis and
plasmid shuffling: from cloned gene to mutant yeast. Methods En-
zymol. 194, 302–318.
Simic, R., Lindstrom, D.L., Tran, H.G., Roinick, K.L., Costa, P.J.,
Johnson, A.D., Hartzog, G.A., and Arndt, K.M. (2003). Chromatin
remodeling protein Chd1 interacts with transcription elongation
factors and localizes to transcribed genes. EMBO J. 22, 1846–1856.
Squazzo, S.L., Costa, P.J., Lindstrom, D.L., Kumer, K.E., Simic, R.,
Jennings, J.L., Link, A.J., Arndt, K.M., and Hartzog, G.A. (2002). The
Paf1 complex physically and functionally associates with transcrip-
tion elongation factors in vivo. EMBO J. 21, 1764–1774.
Strasser, K., Masuda, S., Mason, P., Pfannstiel, J., Oppizzi, M., Ro-
driguez-Navarro, S., Rondon, A.G., Aguilera, A., Struhl, K., Reed,
R., and Hurt, E. (2002). TREX is a conserved complex coupling tran-
scription with messenger RNA export. Nature 417, 304–308.
Wade, P.A., Werel, W., Fentzke, R.C., Thompson, N.E., Leykam, J.F.,
Burgess, R.R., Jaehning, J.A., and Burton, Z.F. (1996). A novel col-
lection of accessory factors associated with yeast RNA polymerase
II. Protein Expr. Purif. 8, 85–90.
Wang, Y., Liu, C.L., Storey, J.D., Tibshirani, R.J., Herschlag, D., and
Brown, P.O. (2002). Precision and functional specificity in mRNA
decay. Proc. Natl. Acad. Sci. USA 99, 5860–5865.
Zenklusen, D., Vinciguerra, P., Wyss, J.C., and Stutz, F. (2002). Sta-
Normal Poly(A) Site Utilization Requires Paf1
ble mRNP formation and export require cotranscriptional recruit-
ment of the mRNA export factors Yra1p and Sub2p by Hpr1p. Mol.
Cell. Biol. 22, 8241–8253.
Zhu, B., Mandal, S.S., Pham, A.D., Zheng, Y., Erdjument-Bromage,
H., Batra, S.K., Tempst, P., and Reinberg, D. (2005). The human PAF
complex coordinates transcription with events downstream of RNA
synthesis. Genes Dev. 19, 1668–1673.
Zorio, D.A., and Bentley, D.L. (2004). The link between mRNA pro-
cessing and transcription: communication works both ways. Exp.
Cell Res. 296, 91–97.
The data are available at http://www.ncbi.nlm.nih.gov/projects/
geo/index.cgi accession number GSE3200.