Transcript Elongation Factor TFIIS Is Involved in Arabidopsis Seed Dormancy

doi:10.1016/j.jmb.2008.12.066 J. Mol. Biol. (2009) 386, 598–611

Available online at www.sciencedirect.com

Transcript Elongation Factor TFIIS Is Involved inArabidopsis Seed Dormancy

Marion Grasser1, Caroline M. Kane2, Thomas Merkle3, Michael Melzer4,Jeppe Emmersen5 and Klaus D. Grasser1⁎
1Department of Life Sciences,Aalborg University,Sohngaardsholmsvej 49,DK-9000 Aalborg, Denmark2Molecular and Cell Biology,University of California,408 Barker Hall, Berkeley,CA 94720-3202, USA3Genome Research, Faculty ofBiology, University of Bielefeld,Universitätsstr. 25,D-33594 Bielefeld, Germany4Leibniz Institute of PlantGenetics and Crop PlantResearch, Corrensstr. 3,D-06466 Gatersleben, Germany5Department of Health Scienceand Technology, AalborgUniversity, Fredrik Bajersvej 7,DK-9220 Aalborg, Denmark
Received 16 October 2008;received in revised form13 December 2008;accepted 22 December 2008Available online3 January 2009

*Corresponding author. E-mail addrAbbreviations used: RNAPII, RN

6-AU, 6-azauridine; daf, days after freverse-transcribed polymerase chaifluorescent protein; GPD, glyceraldedehydrogenase; GO, gene ontology.

0022-2836/$ - see front matter © 2008 E

Transcript elongation factor TFIIS promotes efficient transcription by RNApolymerase II, since it assists in bypassing blocks during mRNA synthesis.While yeast cells lacking TFIIS are viable, inactivation of mouse TFIIScauses embryonic lethality. Here, we have identified a protein encoded inthe Arabidopsis genome that displays a marked sequence similarity to TFIISof other organisms, primarily within domains II and III in the C-terminalpart of the protein. TFIIS is widely expressed in Arabidopsis, and a greenfluorescent protein–TFIIS fusion protein localises specifically to the cellnucleus. When expressed in yeast cells lacking the endogenous TFIIS,Arabidopsis TFIIS partially complements the sensitivity of mutant cells tothe nucleotide analog 6-azauridine, which is a typical characteristic oftranscript elongation factors. We have characterised Arabidopsis linesharbouring T-DNA insertions in the coding sequence of TFIIS. Plantshomozygous for T-DNA insertions are viable, and genomewide transcriptprofiling revealed that compared to control plants, a relatively smallnumber of genes are differentially expressed in mutant plants. TFIIS−/−

plants display essentially normal development, but they flower slightlyearlier than control plants and show clearly reduced seed dormancy.Plants with RNAi-mediated knockdown of TFIIS expression also areaffected in seed dormancy. Therefore, TFIIS plays a critical role inArabidopsis seed development.

© 2008 Elsevier Ltd. All rights reserved.

Keywords: transcript elongation; RNA polymerase II; transcriptome; yeastcomplementation; nuclear localisation
Edited by J. Karn
Introduction

Transcription by RNA polymerase II (RNAPII)proceeds throughmultiple stages, including promoterrecruitment, initiation, elongation, and termination.The vast majority of studies investigating the regula-tion of eukaryotic transcription have focused on the

ess: [email protected] polymerase II;lowering; rtPCR,n reaction; GFP, greenhyde-3-phosphate

lsevier Ltd. All rights reserve

early stages of the transcription cycle (i.e., preinitiationand initiation). In the past few years, it has becomeapparent that mRNA synthesis is a dynamic andhighly regulated step in gene expression that requiresthe concerted action of a variety of transcript elonga-tion factors.1–3 One of the factors regulating transcriptelongation is TFIIS, which was identified early as anactivity that facilitated transcript elongation in vitro.4

The ∼35-kDa TFIIS of yeast and mammals is a three-domain protein. Domain I has the least sequenceconservation and comprises the N-terminal part of theprotein that forms a four-helix bundle and is dis-pensable for transcription stimulation in vitro. Thecentral domain II contains a well-defined three-helixbundle and (together with the linker region between

d.

599Arabidopsis TFIIS

domains II and III) is involved in the interaction withRNAPII. The 45-amino-acid-residue C-terminaldomain III consists of a zinc ribbon fold and representsthe most highly conserved part of the protein.5,6 Incomplex with RNAPII, TFIIS extends from the poly-merase surface via a pore to the internal active site ofthe enzyme. From domain III protrudes a β-hairpinthat complements the RNAPII active site with twoinvariant and functionally critical acidic residues.7

Fig. 1. Amino acid sequence alignment of TFIIS from varA. thaliana (At; At2g38560), Homo sapiens (Hs; GenBank GenID 34883), Caenorhabditis elegans (Ce; GenBank Gene ID 1744S. cerevisiae (Sc; GenBank Gene ID 852839) were aligned usincase of human TFIIS, the abundant ubiquitously expressed isoare indicated by black bars below the sequences. Asterisksequences). The four cysteine residues in domain III that cohairpin residues D/E are underlined. The residues delineatinhuman TFIIS25 are indicated by (▶,◀) in the N-terminal partDNA insertions (cf. Fig. 5a) disrupting the Arabidopsis TFIIS cois indicated by (▾). Amino acid positions within the individu

Generally, synthesis of mRNA is processive, buttranscript elongation can be blocked in variousways, including certain DNA sequences that desta-bilise the DNA–RNA hybrid, causing reversetranslocation (backtracking) of RNAPII. This resultsin displacement of the extendable 3′-end of the RNAfrom the polymerase active site. Backtracking by afew nucleotides leads to (temporary) pausing, whilemore extensive backtracking causes (irreversible)

ious organisms. The amino acid sequences of TFIIS frome ID 6917), Drosophila melanogaster (Dm; GenBank Gene50), Oryza sativa (Os; Os03g60130 and Os07g12630), andg ClustalW (http://align.genome.jp/sit-bin/clustalw). Inform TFIIS.o24 is shown. The conserved domains II and IIIs indicate conserved residues (in at least six of sevenordinate a zinc ion are in boldface, and invariant acidicg the LW motif found to mediate nuclear localisation ofabove the sequences. The relative position of the two T-

ding sequence that were analysed in this study (see below)al sequences are given to the right.

Fig. 2. Amino acid sequence similarity of TFIIS fromdifferent organisms. The protein sequences (cf. Fig. 1) werealigned by multiple sequence alignment (ClustalW) andused to construct a neighbour-joining tree (http://align.genome.jp/sit-bin/clustalw) representing the degree ofsequence similarity between the proteins. The plant TFIISgroup (Arabidopsis and rice) is indicated by a circle.

600 Arabidopsis TFIIS

transcriptional arrest.8 TFIIS facilitates RNAPIIread-through of various blocks to transcript elonga-tion, including arrest sites,9,10 by strongly enhancingthe intrinsic RNA nuclease activity of RNAPII.11,12The endonucleolytic cleavage of the nascent tran-script results in the creation of a new 3′-end of thetranscript located in the RNAPII active site that iscorrectly base-paired to the DNA template and canbe extended. In the cleavage reaction, the conservedβ-hairpin of domain III plays a central role bybringing the two acidic residues close to the RNA, sothat they complement the active site of RNAPIIassisting in hydrolytic RNA cleavage. Moreover,TFIIS induces extensive structural changes in RNA-PII and facilitates realignment of the RNA in theactive site for catalysis.13,14 The activity of TFIIS isalso critical for efficient release of RNAPII frompromoter-proximal stall sites.15 In addition to thewell-established function in transcript elongation,TFIIS can play a role in transcript initiation.16–18The single-copy genes encoding TFIIS in Saccharo-

myces cerevisiae and Saccharomyces pombe are not essen-tial for viability; however, typical of transcriptelongationdefects, TFIISdeficiency results in increasedsensitivity to the drug 6-azauridine (6-AU).19–21

Inactivation of TFIIS in mice caused embryonic letha-lity due to defects in definitive hematopoiesis.22 Inplants, TFIIS has not been studied so far. However, azinc ribbon protein with similarity to the C-terminalpart of TFIIS, termed ET1, has been characterisedfrom maize, and related proteins are encoded inthe genomes of a wide variety of plant species. The163-amino-acid ET1 protein is significantly smallerthan TFIIS and is localised in plastids.23 In kernelsof mutant maize plants lacking ET1, starch syn-thesis is affected, resulting in defects in endospermdevelopment, while mutant seedlings exhibit avirescent phenotype. These defects are due toaberrant plastid development caused by theabsence of ET1.23 Here we report the identificationof a nuclear protein with striking sequence simila-rity to TFIIS that is encoded in Arabidopsis, and weexamine mutant plants lacking this protein incomparison to control plants.

Results

The Arabidopsis genome encodes a protein withstriking similarity to TFIIS

In view of the conservation of TFIIS in organismsas different as yeast and mammals,5,6 and the identi-fication of a small TFIIS-like protein in plantplastids,23 we wondered whether there is also aplant nuclear protein related to TFIIS. Therefore, wehave performed a BLAST search of the Arabidopsisthaliana database† using the amino acid sequence of

†http://www.arabidopsis.org/

the 35-kDa S. cerevisiae TFIIS as query. The searchrevealed one striking match with a 42-kDa proteinencoded in the Arabidopsis genome (AGI locusAt2g38560) termed in the following TFIIS. Thissequence was aligned with TFIIS sequences ofother organisms, including two related sequencesidentified in the rice genome database‡ (Fig. 1).Compared to TFIIS sequences of other organisms,the plant sequences contain an insertion in the N-terminal part (∼35 residues in rice; ∼50 residues inArabidopsis). This region also corresponds to thepart of the protein that is most variable (both inlength and in sequence) among the three vertebrateisoforms of TFIIS24 and that is remarkably exten-ded in some vertebrate TFIIS variants.26 The align-ment also reveals that the plant TFIIS sequencesdisplay significant similarity to typical TFIISproteins (Fig. 1), primarily in the C-terminal part(domains II and III). Thus, Arabidopsis TFIIS sharesan ∼25% overall amino acid sequence identity withS. cerevisiae TFIIS and an ∼60% identity withindomain III. The four cysteine residues that coordi-nate a zinc ion in domain III,13 as well as the acidichairpin residues D359 and E360 in ArabidopsisTFIIS, are conserved (Fig. 1). A multiple aminoacid sequence alignment was used to construct aneighbour-joining tree of the different TFIISsequences, illustrating the relation of TFIIS fromdifferent eukaryotes (Fig. 2). The monocot anddicot plant sequences form a subgroup in the TFIISprotein family.

‡http://rice.tigr.org/


Arabidopsis TFIIS is ubiquitously expressed,and a green fluorescent protein–TFIIS fusionlocalises to the cell nucleus

To examine the expression of TFIIS in Arabidopsis,RNA was isolated from different tissues and, afterreverse transcription, used for detection of theTFIIS transcript by PCR. Using TFIIS-specificprimers, amplification products of the expectedsize (159 bp) were obtained from all RNA samples(Fig. 3a), which were derived from the leaves,roots, and flower buds of 4-week-old plants; greenparts of 10-day-old seedlings; seeds harvested15 days after flowering (daf); and suspension cul-tured cells. Amplification of 16S RNA served asreference. The reverse-transcribed polymerasechain reaction (rtPCR) experiment indicated thatTFIIS is ubiquitously expressed in Arabidopsis and

Fig. 3. Arabidopsis TFIIS is ubiquitously expressed, anda GFP–TFIIS fusion localises to the cell nucleus. (a) rtPCRanalysis of total RNA isolated from the leaves, roots, andflowers of 4-week-old plants; green parts of 10-day-oldseedlings; seeds harvested 15 daf; and suspensioncultured cells. PCR was performed with primers specificfor TFIIS (top) and 16S RNA (bottom). BY-2 protoplastswere transformed with the plasmids driving the expres-sion of GFP/DsRed fusion proteins, and the localisation ofthe fusion proteins was analysed by confocal laserscanning microscopy. (b) GFP alone, (c) NLS-CHS-DsRed, (c) GFP–TFIIS, and (e) the transmission image ofthe same protoplast as in (d). The size bar in (e)corresponds to 10 μm.

that the expression levels differ. For instance, thereis a lower amount of the TFIIS transcript in 15-dafseeds. In vertebrates, there are three conservedisoforms of TFIIS; one of them is ubiquitouslyexpressed, while the others display a more tissue-specific expression.24,27

Since TFIIS typically is considered a nuclearprotein although the related maize ET1 localises toplastids,23 we have examined the subcellular locali-sation of Arabidopsis TFIIS. We have constructedplasmids suitable for the expression of green fluores-cent protein (GFP) fusion proteins in plant proto-plasts. In transient transformation assays performedwith tobacco BY-2 suspension cell protoplasts, theexpression of the GFP fusion proteins was driven bythe CaMV 35S promoter. Transformed protoplastswere analysed by confocal laser scanning micro-scopy. We have used constructs that served ascontrols28,29 in this experiment: GFP that is found inboth cytosol and nucleus (Fig. 3b), and NLS-CHS-DsRed that localises to the nucleus (Fig. 3c). As theC-terminal part of yeast TFIIS is deeply inserted intoRNAPII,13,14 we have constructed an N-terminalGFP fusion of TFIIS. In line with predictions usingthe software pTARGET§ and pSORTǁ, a GFP–TFIISfusion clearly accumulated in the nucleus (Fig. 3d).Colocalisation of nucleus and green fluorescence isalso evident from the comparison of the nuclearGFP–TFIIS fluorescence and the transmission imageof the same protoplast (Fig. 3d and e). Similar tohuman TFIIS,25 Arabidopsis TFIIS is underrepre-sented in the nucleolus (Fig. 3d). Residues withinthe so-called LW motif that were shown to becritical for nuclear targeting of human TFIIS25 areessentially conserved in Arabidopsis TFIIS (residuesL38-K84; cf. Fig. 1).

Arabidopsis TFIIS partially complements the6-AU sensitivity of yeast cells lacking yTFIIS

Yeast cells lacking TFIIS become sensitive to anucleotide analog 6-AU, a drug that reduces nucleo-tide pools of guanosine triphosphate and uridinetriphosphate in yeast.30 Yeast cell mutants in avariety of elongation factors are sensitive to thisdrug, which is considered characteristic ofmutationsin TFIIS.6 To determine whether or not ArabidopsisTFIIS had functional similarity to bona fide TFIIS fromother organisms, the open reading frame of theArabidopsis gene was cloned behind the yeastglyceraldehyde-3-phosphate dehydrogenase (GPD)gene promoter in p425GPD and transformed intoyeast strain CMKy7 in which the gene encodingTFIIS has been disrupted by URA3. This strain issensitive to 6-AU.31 The growth of CMKy7 trans-formed with the empty vector was compared to thegrowth of CMKy7 transformed with the plasmidencoding either S. cerevisiae TFIIS or ArabidopsisTFIIS on media with and without 6-AU (Fig. 4). In

§http://bioapps.rit.albany.edu/pTARGET/ǁhttp://psort.ims.u-tokyo.ac.jp/form.html

Fig. 4. Arabidopsis TFIIS partially complements the6-AU sensitivity of the yeast disruption of TFIIS. CMKy7cells transformed with either p425GPD, p425GPD-yTFIIS,or p425GPD-TFIIS were grown to late-log phase anddiluted to 0.1 A600 units with water. Serial 5-fold dilutionswere made, and 5 μl of each dilution was plated onto theindicated media. Cells were grown at 30 °C for the indi-cated number of days. 6-AU was present when indicatedat 75 μg/ml.


the absence of 6-AU, yeast with the vector alonegrew faster than yeast with either the native TFIISor the Arabidopsis TFIIS, and the latter grew theslowest. On 6-AU, the growth was distinctly diffe-rent. Cells containing p425GPD-yTFIIS grewrobustly, as expected. Cells expressing ArabidopsisTFIIS grew better than those containing the emptyvector on 6-AU; not only were the cells visible morequickly, but there were also more colonies visible ateach dilution. These results were obtained in fourindependent experiments. The comparison forgrowth on solid media with and without 6-AU isthus quite compelling for partial complementationby Arabidopsis TFIIS for yeast TFIIS.In an attempt to be more quantitative, the

growth rate of each culture was followed in liquidSC-LeuUra medium lacking 6-AU. Cells with thevector alone had a doubling time of 2.5 h, cells withyeast TFIIS had a doubling time of 2.8 h, and cellswith the Arabidopsis TFIIS had a doubling time of3.3 h. These results are consistent with those seen onsolid SC-LeuUra medium. Several attempts to deter-

mine growth rates in the same media containing6-AU were performed. In each case, the growthrates would suddenly accelerate for one or anotherstrain, suggesting mutations in drug uptake path-ways. Previous work looking at dosage suppressorsfor yeast TFIIS in the presence of 6-AU found manygenes associated with drug uptake.32 The ability ofyeast to adapt to and interfere with drug uptake inmedia containing 6-AU prevents any simple growthrate determination in the presence of drug. Thus,we have presented the data for growth on solidmedia containing 6-AU to document the comple-mentation seen in yeast expressing the Arabidopsisgene. However, the fact that cells with the vectoronly grow fastest on media lacking 6-AU but growslowest on media with 6-AU documents the growthadvantage conferred by Arabidopsis TFIIS on 6-AU.That the complementation is not robust is not

surprising, given the previously reported speciesspecificity of TFIIS (reviewed by Fish and Kane5).Clearly expressing this plant protein in these yeastcells complements the drug sensitivity of CMKy7,and this result provides strong evidence that thisgene from Arabidopsis encodes the plant transcrip-tion elongation factor TFIIS. Further biochemicalstudies will be necessary to validate this geneticobservation through transcription studies in vitro.

Molecular characterisation of Arabidopsisplants lacking TFIIS

The transcribed region of the Arabidopsis gene(AGI locus At2g38560) encoding TFIIS consists oftwo exons (Fig. 5a). To examine the role of TFIISin Arabidopsis, two T-DNA insertion lines(SALK_056755 and SALK_027259; ecotype Col-0)from the SALK collection33 were analysed. The T-DNA insertions are located in the second exon ofTFIIS, disrupting the coding sequence of the gene(Fig. 5a). The insertion sites were confirmed by PCRanalysis of genomic DNA (Fig. 5b), in combinationwith DNA sequencing of PCR fragments spanningthe left borders of the T-DNA (obtained withprimers P2+P3). These analyses revealed that (i)we were able to isolate plants homozygous for theT-DNA insertions (TFIIS−/−), termed tfIIs-1 andtfIIs-2, and (ii) the insertions are placed ∼333 and∼518 bp downstream of the translational startcodon in tfIIs-1 and tfIIs-2, respectively. Therefore,the insertions are situated in the sequence encodingthe N-terminal part of TFIIS (cf. Fig. 1). RNAisolated from tfIIs-1 and tfIIs-2 plants (and, forcomparison, from Col-0) was examined by North-ern blot analysis using a radiolabelled probespecific for TFIIS (Fig. 5c). While a band of theexpected size (∼1600 nt) hybridised with the probeusing Col-0 RNA, no corresponding hybridisationsignal was detected with the tfIIs-1 and tfIIs-2 RNA.Reduced amounts of abnormal transcripts of ∼900and ∼4000 nt were detected in the mutant samples(Fig. 5c), as often observed in Arabidopsis T-DNAinsertion lines.34–36 The short transcript found intfIIs-1 appears to terminate within the T-DNA,

Fig. 5. Characterisation of Arabidopsis TFIIS T-DNAinsertion lines. (a) Schematic representation of the TFIISgene. Boxes represent exons, while lines indicate introns(not drawn to scale). 5′-Untranslated region and 3′-untranslated region, as well as the ATG start codon andthe TGA stop codon, are indicated. Triangles indicate theT-DNAs (∼4.5 kb) inserted into exon 2, and oligonucleo-tide primers used for molecular characterisation of thelocus are indicated by arrows. (b) PCR analysis of genomicDNA isolated from seedlings documenting wild-typeplants by amplifying a 834-bp fragment of the TFIISgene using primers P1+P2, as well as plants homozygous(TFIIS−/−; termed tfIIs-1 and tfIIs-2 in the following) for theT-DNA insertions by amplifying 766- and 580-bp frag-ments, respectively, cover the left border of the T-DNAsobtained with primers P3+P2. (c) Northern blot analysisof RNA isolated from Col-0, tfIIs-1, and tfIIs-2 plants usinga radiolabelled TFIIS-specific probe (spanning the regionfrom 145 to 925 bp, relative to the ATG start codon, of theTFIIS coding sequence). The hybridising band of ∼1600 ntcorresponding to the TFIIS transcript is detected in Col-0(indicated by an arrow on the left), but not in the mutants.In tfIIs-1 and tfIIs-2, hybridising bands of ∼900 and∼4000 nt (indicated by arrows on the right) are detected.A control experiment with the same blot hybridised with aprobe specific for the Actin 2 gene is shown below.


while the long transcript observed in tfIIs-2 mostlikely is initiated from a promoter in the T-DNAconstruct.34–36 By DNA sequence analysis of thetfIIs-2 T-DNA border region, in-frame stop codonswere detected. Therefore, in both mutants, nofunctional TFIIS synthesis is expected, consistentwith the common phenotype observed for the twoT-DNA insertion lines (see below).

Comparative transcript profiling of tfIIs-1 andCol-0

Since TFIIS is involved in the regulation of tran-scription,5,6 we have performed a genomewide geneexpression analysis of 10-day-old tfIIs-1 and Col-0seedlings using microarray hybridisation. Evalua-tion of the data revealed that 32,354 spots gaveconsistently above-background signals. Statisticalanalysis identified that 733 gene models (∼2.3%)exhibited a significantly differential expression(adjusted Pb0.01) between tfIIs-1 and Col-0: 512gene models were N2-fold down-regulated in tfIIs-1,while 221 gene models were N2-fold up-regulated inthe mutant (in addition, 464 gene models weresignificantly differentially regulated with a b2-folddifference). The larger number of down-regulatedgenes is in line with the presumed positive effect ofTFIIS on transcription. According to the microarrayexperiment, the TFIIS gene is among the mostseverely down-regulated genes in tfIIs-1. To gaininsight into biological processes in which thedifferentially expressed genes are involved, thegenes that were significantly differentially expressed(N2-fold) were analysed for gene ontology (GO).37

The GO categories that were identified as mostsignificantly up-regulated comprise genes involvedin glycoside/glucosinolate biosynthesis/metabo-lism, while the most significantly down-regulatedGO category comprises genes involved in secondarymetabolism (including phenylpropanoids, aminoacids, and aromatic compounds) (Fig. 6). UsingrtPCR analyses, the transcript levels of 12 selectedgenes were examined in Col-0 and tfIIs seedlings.These genes comprised candidates that, according tothe microarray results, were up-regulated or down-regulated in tfIIs-1 relative to Col-0, as well as geneswhose expression appeared to be unaffected. Thisexperiment revealed that the relative gene expres-sion levels determined by microarray hybridisationand rtPCR are essentially consistent, verifying theobtained microarray data. In addition, the rtPCRexperiment demonstrated that the second T-DNAinsertion line tfIIs-2 displayed transcript levelscomparable to those of tfIIs-1.

Phenotypic analysis of plants lacking TFIIS

Several studies have demonstrated that certainelongation factors are essential for proper develop-ment in higher eukaryotes, including plants.38,39While TFIIS is not essential in S. cerevisiae and S.pombe,19–21 targeted disruption of TFIIS in miceresulted in embryonic lethality.22 We have examined


the growth and development of tfIIs-1 and tfIIs-2(lacking TFIIS) compared to Col-0.Mature seeds of allthree genotypes germinatednormally, andnomarkeddifferences were observed in vegetative seedlingdevelopment. Compared to Col-0 (∼23 days afterstratification), tfIIs-1 and tfIIs-2 bolted slightly earlier(∼20 days after stratification), as seen by theelongation of the primary inflorescence (Fig. 7a).Flowering and seed production of tfIIs-1 and tfIIs-2were similar to those of Col-0. Striking differencesbetween mutants and wild type were noticed,however, when the germination of seeds harvestedfrom green siliques was tested. Seeds (still green or

Fig. 6. GO analysis of genes down-regulated in tfIIs-1 relativthe BiNGO plug-in version. GO categories that were significgenes were identified. Yellow to orange circles correspond to th(adjusted P≤0.01). The size of the circles is proportional to th

just turning yellowish) were harvested 15 daf andsown for germination tests without prior storage.Seed germination was scored after 7 days of incu-bation in a plant growth chamber. While Col-0 seedshardly germinated, tfIIs-1 and tfIIs-2 seeds germi-nated efficiently (Fig. 7b and Table 1). Seeds (15 daf)were sectioned and analysed by light microscopy(Fig. 7c) to examine whether the embryos of tfIIsseeds are properly developed. Analysis of severalseeds of each genotype did not reveal morphologicaldifferences between the tfIIs-1, tfIIs-2 and Col-0embryos. When 15-daf seeds were dry-stored for2 weeks prior to the germination tests, the germi-

e to Col-0. The analysis was performed in Cytoscape usingantly overrepresented among the differentially expressede level of significance of the overrepresented GO categorye number of genes in each category.

Fig. 7. Plants lacking TFIIS bolt earlier than Col-0 and display reduced seed dormancy. (a) Early bolting of tfIIs-1 andtfIIs-2 compared to Col-0, as illustrated by 25-day-old plants grown under long-day conditions. (b) Compared to Col-0,tfIIs-1 and tfIIs-2 display reduced seed dormancy (Pb0.01). The 15-daf seeds were sown on wet paper and incubated for7 days in a growth chamber. tfIIs-1 and tfIIs-2 seeds germinate efficiently, while Col-0 seeds hardly germinate underthese conditions. (c) Normal embryo development in tfIIs-1 and tfIIs-2. Light microscopic analysis of sections of 15-dafseeds of Col-0 and tfIIs-1 and tfIIs-2. Several seeds from different siliques were analysed for each genotype, revealingcomparable embryos. Representative examples are shown. Radicles of the embryos are indicated by “r,” while “c”indicates cotyledons.


nation of the mutant lines and Col-0 were compa-rable (Table 1). To examine whether (partial) down-regulation of TFIIS expression levels also affects seedgermination, an RNAi construct (pFGC5941-TFIIS)targeting TFIIS was introduced into Arabidopsis byAgrobacterium-mediated transformation. Thirty-twoindependent transgenic plant lines identified byPCR-based genotyping were tested for TFIIS expres-sion using Northern blot analysis. Five third-generation lines that had TFIIS transcript levels of∼50% compared to Col-0 were analysed in germi-nation tests. The 15-daf seeds of the five RNAi linescompared to Col-0 showed increased germinationrates, but the rates are lower than those of the T-DNAinsertion lines (Table 1). Most likely due to thevariability of the germination trait, and possibly incombination with imprecise estimation of the TFIISexpression levels of the RNAi lines, no strict corre-lation between TFIIS expression levels and germina-tion rates is seen. However, the germination ratesobserved for tfIIs-1 and tfIIs-2 lacking the TFIIS tran-

script are clearly higher than those for Col-0, whilethe RNAi lines that have reduced TFIIS expressionshow “intermediate” germination rates. Theseexperiments demonstrate that only the germinationof freshly harvested green seeds is affected by TFIISexpression,whereas the germination ofmature seedsand 15-daf seeds after dry storage is indistinguish-able for the different tested genotypes. This suggeststhat the lack of TFIIS (or lower expression levels)reduces the dormant stage of the fully developedseed before it can resume germination. Therefore,TFIIS may be involved in the regulation of seeddormancy in Arabidopsis.

Discussion

We have identified a nuclear 378-amino-acid-residue TFIIS protein encoded in the Arabidopsisgenome. Plant TFIIS is slightly bigger than its coun-terparts from other organisms, primarily due to an

Table 1. Germination rates of 15-daf seeds of the differentplant lines

Plant line Typea

TFIISexpressionb

(%)

15-dafseedsc

(%) P

15-daf seedsafter 14-daystoraged (%)

Col-0 Wild type 100 5±3 98±1tfIIs-1 Insertion 0 63±19 b0.01 98±2tfIIs-2 Insertion 0 71±19 b0.01 99±1Col-0 Wild type 100 5±4 98±1RNAi-1 RNAi ∼50 28±9 b0.01 99±1RNAi-2 RNAi ∼40 32±9 b0.01 95±3RNAi-3 RNAi ∼60 48±15 b0.01 92±2RNAi-4 RNAi ∼40 38±12 b0.01 91±2RNAi-5 RNAi ∼40 30±7 b0.01 74±5

a Type of plant line used for germination tests. Col-0 served asreference for the T-DNA insertion lines and RNAi lines.

b TFIIS expression estimated by Northern blot analysis.c Germination rate of freshly sawn 15-daf seeds.d Germination rate of seeds harvested at 15 daf after 14-day dry

storage.


insertion in the N-terminal part of the protein; how-ever, domains II and III, which are essential for pro-moting transcription in vitro,5,6 are well conserved.We have characterised two independent Arabidopsislines homozygous for T-DNA insertions (tfIIs-1 andtfIIs-2) that disrupt the coding sequence of TFIIS;accordingly, no wild-type TFIIS transcript wasdetected in the mutant lines. Comparative genome-wide transcript profiling of tfIIs-1 and Col-0 revealedthat the expression of 0.7% of the genes is signifi-cantly up-regulated in the mutant, while 1.6% of thegenes are down-regulated. In view of the positiverole of TFIIS in transcription, facilitating read-through of various blocks to transcript elonga-tion,5,6,40 the transcript profiling data imply thatTFIIS is critically involved in the transcription of onlya subset of genes, or that there is functional overlapwith other proteins, as discussed below. Clearly, thevast majority of genes appear to be transcribedefficiently in the absence of TFIIS. In line with thisresult, a recent study in which TFIIS was ∼50%down-regulated by siRNAs in humanMCF cell linesrevealed that the reduced amount of TFIIS causeddifferential expression of only a relatively smallnumber of genes.41 However, these experimentsmayresult in an underestimation of the overall require-ment for TFIIS in gene transcription. In our transcriptprofiling experiment, for instance, the aerial parts of10-day-old seedlings were analysed and, therefore,many types of tissue (e.g., roots, floral organs, andfruits) are not represented. The requirement for TFIISmay depend on the type of tissue and/or develop-mental stage that clearly influences differentially thetranscription of many genes and their expressionlevels. As a consequence, it can be expected that alarger number of genes will require the presence ofTFIIS for proper transcription in certain tissues or atparticular stages of development.Plants lacking TFIIS develop similarly to Col-0,

except that they flower slightly earlier than thecontrol plants, suggesting that inArabidopsis, TFIIS isnot critical for vegetative development. While the

inactivation of mouse TFIIS causes embryonic letha-lity,22 in yeast, the gene encoding TFIIS is notessential, and deletion of the gene has no effect ongrowth on media without drugs.5 However, inacti-vation of TFIIS, in combination with either deletionsor mutations in chromatin remodelling factors42 andtranscription factors,1,43 is indeed lethal. Thus, thereare functional overlaps between TFIIS and a numberof other proteins involved in transcriptional regula-tion. Perhaps the same is true in Arabidopsis. A cleareffect of TFIIS is seen in Arabidopsis seeds. Thegermination of tfIIs seeds harvested frommature drysiliques is comparable to Col-0 seeds, but a markeddifference exists in the germination of 15-daf seedsharvested from elongated but still green siliques.Col-0 seeds sown without prior storage hardlygerminate, whereas tfIIs-1 and tfIIs-2 seeds of thesame condition germinate efficiently. When seedsfrom the RNAi lines that have reduced TFIIS tran-script levels were tested, they displayed intermedi-ate germination rates. Taken together, this suggeststhat TFIIS is involved in the regulation of seeddormancy in Arabidopsis. Seed dormancy, which isregulated by a range of phytohormones and envi-ronmental factors, is defined as a block to completegermination of an intact viable seed under favour-able conditions.44 A dry dormant seed can surviveextended periods of unfavourable conditions.Accordingly, the important biological role is thatseeds germinate after a period of dormancy, allo-wing germination to occur when conditions aresuitable for establishing a new plant generation, forinstance, germination in spring for summer annualsin temperate zones.45–47 Since it is expected thatTFIIS will not directly regulate dormancy, itprobably acts indirectly by modulating the tran-scription of genes that are critically involved in seeddormancy/germination. An analysis of our tran-script profiling data did not reveal known regula-tors of seed dormancy that are differentiallyexpressed in tfIIs-1 compared to Col-0, but ourmicroarray experiment was performed using RNAisolated from seedlings rather than seeds. There-fore, we have examined the expression of a varietyof Arabidopsis seed-dormancy-related genes, includ-ing SPT, PIL5, ATS2, ABI1-4, PER1, HUB1/2, andDOG1.48–52 However, according to rtPCR analysesof RNA isolated from 15-daf seeds of Col-0 andtfIIs lines, none of these genes displayed signifi-cantly differential expression (data not shown).This suggests that the effect of TFIIS on seeddormancy might be mediated by so far unknownfactors, for instance, encoded by loci that have beenmapped in various physiologically characterisedArabidopsis dormancy mutants.48 In order to iden-tify target genes involved in seed dormancy, futureexperiments ideally should systematically analysegene expression in fully developed green seeds ofthe different genotypes. For more detailed studieson the involvement of TFIIS in seed dormancy andgermination, it might be advantageous to useanother Arabidopsis ecotype (e.g., Cvi) that, com-pared to Col-0, has a more pronounced seed dor-


mancy.45 Interestingly, Arabidopsis Hub1, another“general” transcription regulator that modulateschromatin structure by ubiquitinating histone H2B,has been shown to be involved in seed dormancyand flowering, as well as in the growth of leaves androots.50,53,54

Several studies have shown that RNAPII tran-script elongation factors, including TFIIS, Elongin A,Spt5, Spt6, Paf, and Elongator, are essential forproper development inDrosophila, mouse, zebrafish,and Arabidopsis.22,55–60 This study demonstrates anonredundant role of TFIIS in Arabidopsis seeddormancy and, therefore, underscores the criticalrole of transcript elongation in plant development.38

Also, the partial complementation of the 6-AUsensitivity of yeast lacking TFIIS strongly suggeststhat the Arabidopsis gene encodes a transcription-elongation-regulatory factor. In addition to thenuclear TFIIS reported here, plants have a relatedplastid-localised zinc ribbon protein, as exemplifiedby maize ET1 that is required for proper plastiddevelopment.23 Since elongation regulation spansprokaryotic and eukaryotic organisms, these resultswith Arabidopsis and the effects of gene disruptionsduring development have much to show us aboutthe impact of elongation regulation.

Materials and Methods

Plasmid constructions

The coding sequence of TFIIS was amplified by PCRusing primers P4 and P5 (for primer sequences, seeSupplemental Table 1), DeepVent DNA polymerase(NEB), and an Arabidopsis cDNA library (derived fromthe green parts of seedlings of the ecotype Col-0) astemplate. The purified PCR fragment was digested usingterminal restriction enzyme recognition sites (BamHI/PstI) introduced through the PCR primers, and cloned bystandard methods into the plasmid pQE9cm61 digestedwith BamHI/PstI, giving pQE9cm-TFIIS. For constructionof a 5′-GFP fusion, pQE9cm-TFIIS was treated consecu-tively with PstI, Klenow enzyme, and EcoRI. The DNAfragment was inserted into EcoRI/Ecl136-digested plas-mid p5′GFP,28 giving p5′GFP-TFIIS. For construction of ayeast complementation construct, pQE9cm-TFIIS wasdigested with BamHI/PstI, and the TFIIS coding sequencewas inserted into the plasmid p425GPD,62 givingp425GPD-TFIIS. The sequence encoding S. cerevisiaeTFIIS was obtained by treating plasmid pET15bPPR263

consecutively with NdeI, Klenow enzyme, and BamHI.The DNA fragment was inserted into p425GPD treatedwith SpeI, Klenow enzyme, and BamHI, giving p425GPD-yTFIIS. For construction of an RNAi vector, a 780-bpregion of the TFIIS coding sequence was amplified by PCRusing primers P6/P7 and P8/P9. The two DNA fragmentswere digested with XhoI/NcoI and BamHI/XbaI, respec-tively, and inserted in two consecutive steps into plasmidpFGC5941¶ digested with the same enzymes, givingpFGC5941-TFIIS. All plasmid constructions were checkedby DNA sequencing.

¶ http://www.chromdb.org/rnai/vector_info.html

Subcellular localisation assays with GFP and DsRedfusion proteins

Protoplasts were prepared from dark-grown tobaccoBY-2 cells and transiently transformed with plasmidsencoding GFP and/or DsRed fusions by polyethylene-glycol-mediated transformation, as described pre-viously.28 Excitation of GFP was performed with standardUV light source and fluorescein isothiocyanate filters,while for the excitation of DsRed, tetramethyl-rhodamineisothiocyanate filters were used. For confocal laserscanning microscopy, samples were directly examinedunder oil with a 63× objective and a DM RE TCS4Dmicroscope (Leica) equipped with an argon–krypton laser(excitation, 488 nm; beam splitter, 500 nm; detection, 500–560 nm for GFP; excitation, 488/543 nm; double dichroic,488/543; detection, 500–530 nm; and 580–700 nm forsimultaneous excitation and detection of GFP and DsRed)using Leica Scanware. Analysis of the localisation of theGFP and/or DsRed fusion proteins was performed inthree independent experiments, representing approxi-mately 60–80 transformed protoplasts.

S. cerevisiae complementation

CMKy7 cells (MATa, ade2, ura3–52, his3–200, leu2–3, 112,lysΔ201, ppr2::hisG URA3 hisG)31 transformed with theempty p425GPD vector, with p425GPD-yTFIIS orp425GPD-TFIIS, were grown in SC-LeuUra liquid mediato late-log phase. Cells were diluted to an A600 of 0.1 inwater, and then serial 5-fold dilutions were made in water.Five microliters of each dilution was plated onto either SC-LeuUra or SC-LeuUra+75 μg/ml 6-AU, and cells wereincubated at 30 °C for several days, with comparisonsmade on days 2–7. Plates were scanned on each of thesedays, and the data for days 2 and 6 are presented. Longerincubations were necessary to detect growth on platescontaining 6-AU, especially for cells with p425GPD-TFIISor p425GPD vector.

Plant growth

A. thaliana (Col-0), T-DNA insertion lines (SALK_056755and SALK_027259) from the SALK collection33 kindlyprovided by the Nottingham Arabidopsis Stock Centrea,and transgenic plant lines were grown in soil in a phyto-chamber at 22 °C and 16 h of light (∼7000 lx) per day.64

After the seeds had been sown, they were stratified indarkness for 48 h at 4 °C prior to incubation in the plantgrowth chamber. For growth analysis, at least 30 plants ofeach genotype were grown individually in 7-cm pots. Forgermination tests, either seeds were harvested at differenttimes after flowering from still green siliques, or dry fullymature seeds were used. To ensure the age of the analysedseeds, flowers were labelled when the bud opened. Forgermination tests, seeds (freshly harvested or after2 weeks of dry storage) were placed on Whatman 3Mpaper soaked with water in Petri dishes and moved to aplant incubator (Percival Scientific; 16 h of light at 22 °C,∼7000 lx; 8 h of darkness at 19 °C). To compare germi-nation rates, seeds from different siliques were analysed,and at least three independent experiments (each examin-ing at least 3×80 seeds of each genotype) were performed.Seed germination was scored after 7 days of incubation.

a http://arabidopsis.info/


Data were statistically evaluated using Student's t test(Pb0.01, unless otherwise specified).

Agrobacterium-mediated Arabidopsis transformation

The Agrobacterium strain GV3101 was transformed withpFGC5941-TFIIS by electroporation, as describedpreviously.64 Arabidopsis plants (Col-0) were transformedby Agrobacterium-mediated transformation using thefloral dip method,65 and the seeds were harvested frommature plants. For selection, the seeds were sown on soil,and seedlings were selected by spraying them with asolution containing the herbicide BASTA, as previouslydescribed.66 Plants surviving the selection were furthercultivated and analysed.

PCR-based genotyping of plant lines

To distinguish among plants that are wild type,heterozygous for T-DNA insertions, or homozygous forT-DNA insertions, or to test the transgenic status oftransformants, genomic DNA was isolated from leaves.The genomic DNA was used for PCR analysis with TaqDNA polymerase (Amplicon) and primers specific for T-DNA, TFIIS, or the transgene (for primer sequences, seeSupplemental Table 1).

Northern blot analysis and rtPCR

Total RNA was extracted from ∼100 mg of frozen planttissue using the TRIzol method (Invitrogen). For Northernblot analyses, 10 μg of total RNA was separated byelectrophoresis in agarose–formaldehyde gels in 1× Mopsbuffer [40 mM 3-(N-morpholino)propanesulfonic acid,10 mM sodium acetate, and 1 mM ethylenediaminetetra-acetic acid, pH 7.2] and transferred onto Hybond Nmembrane (Amersham) by capillary blotting. Hybridisationwith 32P-labelled probes was carried out in hybridisationbuffer [7% SDS, 50 mM Na-phosphate (pH 7), 5× SSC(750 mM NaCl, 75 mM sodium citrate), 2% bovine serumalbumin, and 0.1% lauroylsarcosine] at 68 °C overnight. ForrtPCR, the RNA samples were treated with DNase (MBIFermentas). Reverse transcriptionwas performed using 1 μgof RNA and Revert Aid H minus M-MuLV reversetranscriptase (MBI Fermentas). The cDNA obtained wasused for PCR analyses using Taq DNA polymerase(Amplicon). For primer sequences, see Supplemental Table 1.

b http://www.microarrays.be

Microarray analyses

RNA was extracted from the green parts of 10-day-oldCol-0 and tfIIs-1 seedlings (stage 1.02 according to Boyeset al.67) using the RNeasy Plant Mini Kit (Qiagen). Fourreplicate RNA extractions of each genotype were per-formed using independent pools of plants. RNA extractswere controlled for integrity and purity using an AgilentBioanalyser and a Nanodrop spectrophotometer, respec-tively. Preparation of probes and hybridisation to dual-color Arabidopsis Whole Genome Oligo Microarray(Agilent) containing four arrays of forty-four thousand60-mer oligonucleotide probes represented around 40,000well-characterised transcripts. The arrays were hybri-dised in microarray hybridisation chambers (Agilent)overnight at 60 °C. After washing had been performed,the slides were scanned with a DNA microarray scanner(Agilent) using the “extended dynamic range” (scanner

control software v7 update), and images were processedwith the Feature Extraction Software version 9.5 (Agi-lent). The Limma modeling procedure68 with moderatedt-statistics was used to derive P values from the Agilent-processed signal values. To control the false discoveryrate, the Benjamini–Hochberg method was used to adjustP values for multiple testing.69 An adjusted P value ofb0.01 and an absolute fold change of N2 were used asstringent restriction criteria for assigning significantlydifferentially expressed genes. Microarray analyses wereperformed by the VIB MicroArrays Facility (Leuven,Belgium).b Analysis of overrepresented GO categorieswas performed in Cytoscape (version 2.51) with theBiNGO plug-in (version 2.0).37,70 An adjusted P value of0.01 was used as threshold for significance. Overrepre-sentation was computed from hypergeometric test, andautomatic annotations were excluded (IEA code). An M–A plot was generated by subdividing the intensity ratio(M) and average intensity (A) data for all genes into threesubsets: upregulated genes with an adjusted P value ofb0.01 and a logFC of N1; down-regulated genes with anadjusted P value of b0.01 and a logFC of b−1; andremaining genes with an adjusted P value of N0.01.Datasets were plotted in R (version 2.6.2).

Light microscopy of Arabidopsis siliques

For primary fixation, siliques were cut into 2- to 3-mmpieces and kept overnight at 4 °C in 50 mM cacodylatebuffer (pH 7.2) containing 2% (vol/vol) glutaraldehydeand 2% (vol/vol) formaldehyde, followed by one washwith buffer and two washes with distilled water. Forsecondary fixation, samples were transferred to a solutionof 1% (wt/vol) OsO4. After 1 h, samples were washedthree times in distilled water. Samples were dehydratedby increasing the concentration of ethanol stepwise [30%,40%, 50%, 60%, 75%, 90%, and 2× 100% (vol/vol) ethanolfor 1 h each]. After 1 h of dehydration with propyleneoxide, the samples were infiltrated with Spurr resin (PlanoGmbH, Marburg, Germany) as follows: 25%, 50%, and75% (vol/vol) Spurr in propylene oxide for 4 h each, andthen 100% resin overnight. Samples were transferred intoembedding moulds, kept there for 6 h in fresh resin, andpolymerised at 70 °C for 24 h. Semithin sections of 3 μmwere mounted on slides and stained for 2 min with a 1:1mixture of 1% Azur II and 1% methylene blue in 1%aqueous borax at 60 °C. Light microscopic analysis wascarried out with an AxioCam HRc CCD camera systemattached to a Zeiss Axivert 135 microscope.

Accession codes

The microarray data set has been deposited at GeneExpression Omnibus (National Center for BiotechnologyInformation) under accession number GSE10663.

Acknowledgements

We thank Paul Van Hummelen and his collea-gues at the VIB MicroArrays Facility for perform-ing the microarray analysis. We are grateful to the


Nottingham Arabidopsis Stock Centre for provi-ding the Arabidopsis T-DNA insertion lines. Thiswork was supported by grants from the DanishResearch Council, Det Obelske Familiefond, andCarlsbergfondet to K.D.G.

Supplementary Data

Supplementary data associated with this articlecan be found, in the online version, at doi:10.1016/j.jmb.2008.12.066

References

1. Arndt, K. M. & Kane, C. M. (2003). Running with RNApolymerase: eukaryotic transcript elongation. TrendsGenet. 19, 543–550.

2. Saunders, A., Core, L. J. & Lis, J. T. (2006). Breakingbarriers to transcription elongation. Nat. Rev. Mol.Cell Biol. 7, 557–567.

3. Sims, R. J., Belotserkovskaya, R. & Reinberg, D.(2004). Elongation by RNA polymerase II: the shortand the long of it. Genes Dev. 18, 2437–2468.

4. Sekimizu, K., Kobayashi, N., Mizuno, D. & Natori,S. (1976). Purification of a factor from Ehrlichascites tumor cells specifically stimulating RNApolymerase II. Biochemistry, 15, 5064–5070.

5. Fish, R. N. & Kane, C. M. (2002). Promotingelongation with transcript cleavage stimulatoryfactors. Biochim. Biophys. Acta, 1577, 287–307.

6. Wind, M. & Reines, D. (2000). Transcriptionelongation factor SII. BioEssays, 22, 327–336.

7. Cramer, P. (2004). RNA polymerase II structure:from core to functional complexes. Curr. Opin. Genet.Dev. 14, 218–226.

8. Shilatifard, A., Conaway, R. C. & Conaway, J. W.(2003). The RNA polymerase II elongation complex.Annu. Rev. Biochem. 72, 693–715.

9. Reines, D., Chamberlin, M. J. & Kane, C. M. (1989).Transcription elongation factor SII (TFIIS) enablesRNA polymerase II to elongate through a block totranscription in a human gene in vitro. J. Biol. Chem.264, 10799–10809.

10. Wiest, D. K., Wang, D. & Hawley, D. K. (1992).Mechanistic studies of transcription arrest at theadenovirus major late attenuation site. Comparisonof purified RNA polymerase II and washed elonga-tion complexes. J. Biol. Chem. 267, 7733–7744.

11. Izban, M. G. & Luse, D. S. (1992). The RNApolymerase II ternary complex cleaves the nascenttranscript in a 3′–5′ direction in the presence ofelongation factor SII. Genes Dev. 6, 1342–1356.

12. Reines, D. (1992). Elongation factor-dependenttranscript shortening by template-engaged RNApolymerase II. J. Biol. Chem. 267, 3795–3800.

13. Kettenberger, H., Armache, K. J. & Cramer, P. (2003).Architecture of the RNA polymerase II–TFIIS com-plex and implications for mRNA cleavage. Cell, 114,347–357.

14. Kettenberger, H., Armache, K. J. & Cramer, P. (2004).Complete RNA polymerase II elongation complexstructure and its interactions with NTP and TFIIS.Mol. Cell, 16, 955–965.

15. Adelman, K., Marr, M. T., Werner, J., Saunders, A.,Ni, Z., Andrulis, E. D. & Lis, J. T. (2005). Efficientrelease from promoter-proximal stall sites requires

transcript cleavage factor TFIIS. Mol. Cell, 17,103–112.

16. Guglielmi, B., Soutourina, J., Esnault, C. & Werner,M. (2007). TFIIS elongation factor and Mediator actin conjunction during transcription initiation in vivo.Proc. Natl Acad. Sci. USA, 104, 16062–16067.

17. Kim, B., Nesvizhskii, A. I., Rani, P. G., Hahn, S.,Aebersold, R. & Ranish, J. A. (2007). The transcriptionelongation factor TFIIS is a component of RNApolymerase II preinitiation complexes. Proc. NatlAcad. Sci. USA, 104, 16068–16073.

18. Prather, D. M., Larschan, E. & Winston, F. (2005).Evidence that the elongation factor TFIIS plays a rolein transcription initiation at GAL1 in Saccharomycescerevisiae. Mol. Cell. Biol. 25, 2650–2659.

19. Archambault, J., Lacroute, F., Ruet, A. & Friesen, J. D.(1992). Genetic interaction between transcriptionelongation factor TFIIS and RNA polymerase II. Mol.Cell. Biol. 12, 4142–4152.

20. Nakanishi, T., Nakano, A., Nomura, K., Sekimizu, K.& Natori, S. (1992). Purification, gene cloning, andgene disruption of the transcription elongation factorS-II in Saccharomyces cerevisiae. J. Biol. Chem. 267,13200–13204.

21. Williams, L. A. & Kane, C. M. (1996). Isolation andcharacterization of the Schizosaccharomyces pombegene encoding transcript elongation factor TFIIS.Yeast, 12, 227–236.

22. Ito, T., Arimitsu, N., Takeuchi, M., Kawamura, N.,Nagata, M., Saso, K. et al. (2006). Transcriptionelongation factor S-II is required for definitivehematopoiesis. Mol. Cell. Biol. 26, 3194–3203.

23. da Costa e Silva, O., Lorbiecke, R., Garg, P., Müller, L.,Waβmann, M., Lauert, P. et al. (2004). The Etched 1gene of Zea mays (L.) encodes a zinc ribbon proteinthat belongs to the transcriptionally active chromo-some (TAC) of plastids and is similar to thetranscription factor TFIIS. Plant J. 38, 923–939.

24. Labhart, P. & Morgan, G. T. (1998). Identification ofnovel genes encoding transcription elongation factorTFIIS (TCEA) in vertebrates: conservation of threedistinct TFIIS isoforms in frog, mouse, and human.Genomics, 278–288.

25. Ling, Y., Smith, A. J. & Morgan, G. T. (2006). Asequence motif conserved in diverse nuclear proteinsidentifies a protein interaction domain utilised fornuclear targeting by human TFIIS. Nucleic Acids Res.34, 2219–2229.

26. Taira, Y., Kubo, T. & Natori, S. (2000). Participationof transcription elongation factor XSII-K1 in meso-derm-derived tissue development in Xenopus laevis.J. Biol. Chem. 275, 32011–32015.

27. Weaver, Z. A. & Kane, C. M. (1997). Genomiccharacterization of a testis-specific TFIIS (TCEA2)gene. Genomics, 46, 516–519.

28. Haasen, D., Köhler, C., Neuhaus, G. & Merkle, T.(1999). Nuclear export of proteins in plants: AtXPO1 isthe export receptor for leucine-rich nuclear exportsignals in Arabidopsis thaliana. Plant J. 20, 695–706.

29. Kiilerich, B., Stemmer, C., Merkle, T., Launholt, D.,Gorr, G. & Grasser, K. D. (2008). Chromosomal highmobility group (HMG) proteins of the HMGB-typeoccurring in the moss Physcomitrella patens. Gene,407, 86–87.

30. Exinger, F. & Lacroute, F. (1992). 6-Azauracil inhibi-tion of GTP biosynthesis in Saccharomyces cerevisiae.Curr. Genet. 22, 9–11.

31. Weilbaecher, R. G. (1997). Characterization of an RNApolymerase II subunit, RPB9, and a transcript elongation


factor, TFIIS, from Saccharomyces cerevisiae. PhD thesis,University of California, Berkeley.

32. Stark, C. J. (2003). Genetic and genomic analyses ofthe Saccharomyces cerevisiae transcript elongationfactor, TFIIS. PhD thesis, University of California,Berkeley.

33. Alonso, J. M., Stepanova, A. N., Leisse, T. J., Kim,C. J., Chen, H., Shinn, P. et al. (2003). Genome-wideinsertional mutagenesis of Arabidopsis thaliana.Science, 301, 653–657.

34. Bertrand, C., Benhamed, M., Li, Y. F., Ayadi, M.,Lemonnier, G., Renou, J. P. et al. (2005). ArabidopsisHAF2 gene encoding TATA-binding protein (TBP)-associated factor TAF1, is required to integratelight signals to regulate gene expression andgrowth. J. Biol. Chem. 280, 1465–1473.

35. Ullah, H., Chen, J. G., Young, J. C., Im, K. H.,Sussman, M. R. & Jones, A. M. (2001). Modulation ofcell proliferation by heterotrimeric G protein inArabidopsis. Science, 292, 2066–2069.

36. Vlachonasios, K. E., Thomashow, M. F. & Triezenberg,S. J. (2003). Disruption mutations of ADA2b andGCN5 transcriptional adaptor genes dramaticallyaffect Arabidopsis growth, development, and geneexpression. Plant Cell, 15, 626–638.

37. Maere, S., Heymans, K. & Kuiper, M. (2005). BiNGO:a Cytoscape plugin to assess overrepresentation ofgene ontology categories in biological networks.Bioinformatics, 21, 3448–3449.

38. Grasser, K. D. (2005). Emerging role for transcriptelongation in plant development. Trends Plant Sci. 10,484–490.

39. Winston, F. (2001). Control of eukaryotic transcriptelongation. Genome Biol. 2, 1006.1–1006.3.

40. Svejstrup, J. Q. (2007). Contending with transcrip-tional arrest during RNAPII transcript elongation.Trends Biochem. Sci. 32, 165–171.

41. Hubbard, K., Catalano, J., Puri, R. K. & Gnatt, A.(2008). Knockdown of TFIIS by RNA silencing inhibitscancer cell proliferation and induces apoptosis. BMCCancer, 8, 133.

42. Davie, J. K. & Kane, C. M. (2000). Genetic interac-tions between TFIIS and the SWI/SNF chromatinremodelling complex. Mol. Cell. Biol. 20, 5960–5973.

43. Fish, R. N., Ammermann, M. L., Davie, J. K., Lu,B. F., Pham, C., Howe, L. A. et al. (2006). Geneticinteractions between TFIIF and TFIIS. Genetics, 173,1871–1884.

44. Bewley, J. D. (1997). Seed germination and dormancy.Plant Cell, 9, 1055–1066.

45. Bradford, K. & Nonogaki, H. E. (2007). Seed Develop-ment, Dormancy and Germination. Blackwell Publish-ing, Oxford, UK.

46. Finch-Savage, W. E. & Leubner-Metzger, G. (2006).Seed dormancy and the control of germination. NewPhytol. 171, 501–523.

47. Koornneef, M., Bentsink, L. &Hilhorst, H. (2002). Seeddormancy and germination. Curr. Opin. Plant Biol. 5,33–36.

48. Bentsink, L. & Koornneef, M. (2002). Seed dormancyand germination. The Arabidopsis Book. AmericanSociety of Plant Biologists, Rockville, MD.

49. Bentsink, L., Jowett, J., Hanhart, C. J. & Koornneef, M.(2006). Cloning of DOG1, a quantitative trait locuscontrolling seed dormancy in Arabidopsis. Proc. NatlAcad. Sci. USA, 103, 17042–17047.

50. Liu, Y., Koornneef, M. & Soppe, W. J. (2007). Theabsence of histone H2B monoubiquitination in theArabidopsis hub1 (rdo4) mutant reveals a role for

chromatin remodeling in seed dormancy. Plant Cell,19, 433–444.

51. Penfield, S., Josse, E. M., Kannangara, R., Gilday,A. D., Halliday, K. J. & Graham, I. A. (2005).Cold and light control seed germination throughthe bHLH transcription factor SPATULA. Curr.Biol. 15, 1998–2006.

52. Toorop, P. E., Barroco, R. M., Engler, G., Groot,S. P. C. & Hilhorst, H. W. M. (2005). Differentiallyexpressed genes associated with dormancy or germi-nation of Arabidopsis thaliana seeds. Planta, 221,637–647.

53. Cao, Y., Dai, Y., Cui, S. & Ma, S. (2008). Histone H2Bmonoubiquitination in the chromatin of FLOWERINGLOCUS C regulates flowering time in Arabidopsis.Plant Cell, 20, 2586–2602.

54. Fleury, D., Himanen, K., Cnops, G., Nelissen, H.,Boccardi, T. M., Maere, S. et al. (2007). The Arabidopsisthaliana homolog of yeast BRE1 has a function in cellcycle regulation during early leaf and root growth.Plant Cell, 19, 417–432.

55. Gerber, M., Eissenberg, J. C., Kong, S., Tenney, K.,Conaway, J. W., Conaway, R. C. & Shilatifard, A.(2004). In vivo requirement of the RNA polymeraseII elongation factor Elongin A for proper geneexpression and development. Mol. Cell. Biol. 24,9911–9919.

56. Guo, S., Yamaguchi, Y., Schilbach, S., Wada, T., Lee,J., Goddard, A. et al. (2000). A regulator oftranscriptional elongation controls vertebrate neuro-nal development. Nature, 408, 366–369.

57. He, Y., Doyle, M. R. & Amasino, R. M. (2004).PAF1-complex-mediated histone methylation ofFLOWERING LOCUS C chromatin is required forthe vernalization-responsive, winter-annual habit inArabidopsis. Genes Dev. 18, 2774–2784.

58. Kok, F. O., Oster, E., Mentzer, L., Hsieh, J. C.,Henry, C. A. & Sirotkin, H. I. (2007). The role of theSPT6 chromatin remodeling factor in zebrafishembryogenesis. Dev. Biol. 307, 214–226.

59. Nelissen, H., Fleury, D., Bruno, L., Robles, P., DeVeylder, L., Traas, J. et al. (2005). The elongata mutantsidentify a functional Elongator complex in plants witha role in cell proliferation during organ growth. Proc.Natl Acad. Sci. USA, 102, 7754–7759.

60. Oh, S., Zhang, H., Ludwig, P. & van Nocker, S. (2004).A mechanism related to the yeast transcriptionalregulator Paf1c is required for expression of theArabidopsis FLC/MAF MADS box gene family. PlantCell, 16, 2940–2953.

61. Grasser, K. D., Grimm, R. & Ritt, C. (1996). Maizechromosomal HMGc: two closely related structure-specific DNA-binding proteins specify a second typeof plant HMG-box protein. J. Biol. Chem. 271,32900–32906.

62. Mumberg, D., Müller, R. & Funk, M. (1995). Yeastvectors for the controlled expression of heterologousproteins in different genetic backgrounds. Gene, 156,119–122.

63. Christie, K. R., Awrey, D. E., Edwards, A. M. & Kane,C. M. (1994). Purified yeast RNA polymerase II readsthrough intrinsic blocks to elongation in response tothe yeast TFIIS analogue, P37. J. Biol. Chem. 269,936–943.

64. Launholt, D., Merkle, T., Houben, A., Schulz, A. &Grasser, K. D. (2006). Arabidopsis chromatin-asso-ciated HMGA and HMGB use different nucleartargeting signals and display highly dynamic locali-zation within the nucleus. Plant Cell, 18, 2904–2918.


65. Clough, S. J. & Bent, A. F. (1998). Floral dip: asimplified method for Agrobacterium-mediated trans-formation of Arabidopsis thaliana. Plant J. 16, 735–743.

66. Grønlund, J. T., Stemmer, C., Lichota, J., Merkle, T. &Grasser, K. D. (2007). Functionality of the beta/sixsite-specific recombination system in tobacco andArabidopsis: a novel tool for genetic engineering ofplant genomes. Plant Mol. Biol. 63, 545–556.

67. Boyes, D. C., Zayed, A. M., Ascenzi, R., McCaskill,A. J., Hoffman, N. E., Davis, K. R. & Görlach, J.(2001). Growth stage-based phenotypic analysis ofArabidopsis: a model for high throughput functionalgenomics in plants. Plant Cell, 13, 1499–1510.

68. Smyth, G. K. (2004). Linear models and empiricalbayes methods for assessing differential expression inmicroarray experiments. Stat. Appl. Genet. Mol. Biol. 3,Article 3, 1–26.

69. Benjamini, Y. & Hochberg, Y. (1995). Controlling thefalse discovery rate: a practical and powerfulapproach to multiple testing. J. R. Stat. Soc. B, 57,289–300.

70. Shannon, P., Markiel, A., Ozier, O., Baliga, N. S.,Wang, J. T., Ramage, D. et al. (2003). Cytoscape: asoftware environment for integrated models ofbiomolecular interaction networks. Genome Res. 13,2498–2504.

Transcript Elongation Factor TFIIS Is Involved in Arabidopsis Seed Dormancy

Documents

Transcript of Transcript Elongation Factor TFIIS Is Involved in Arabidopsis Seed Dormancy