Genetic Interactions of DST1 in Saccharomyces cerevisiae Suggest a Role of TFIIS in the Initiation-Elongation Transition

Corresponding author: Jeffrey N. Strathern, Bldg. 539, Rm. 151, P.O. Box B, Frederick, MD 21702-1201., strather{at}ncifcrf.gov (E-mail)

Communicating editor: F. WINSTON

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

TFIIS promotes the intrinsic ability of RNA polymerase II to cleave the 3'-end of the newly synthesized RNA. This stimulatory activity of TFIIS, which is dependent upon Rpb9, facilitates the resumption of transcription elongation when the polymerase stalls or arrests. While TFIIS has a pronounced effect on transcription elongation in vitro, the deletion of DST1 has no major effect on cell viability. In this work we used a genetic approach to increase our knowledge of the role of TFIIS in vivo. We showed that: (1) dst1 and rpb9 mutants have a synthetic growth defective phenotype when combined with fyv4, gim5, htz1, yal011w, ybr231c, soh1, vps71, and vps72 mutants that is exacerbated during germination or at high salt concentrations; (2) TFIIS and Rpb9 are essential when the cells are challenged with microtubule-destabilizing drugs; (3) among the SDO (synthetic with Dst one), SOH1 shows the strongest genetic interaction with DST1; (4) the presence of multiple copies of TAF14, SUA7, GAL11, RTS1, and TYS1 alleviate the growth phenotype of dst1 soh1 mutants; and (5) SRB5 and SIN4 genetically interact with DST1. We propose that TFIIS is required under stress conditions and that TFIIS is important for the transition between initiation and elongation in vivo.

PREMESSENGER RNA transcription in eukaryotes is driven by the RNA polymerase II (RNApol II) and can be divided into initiation, elongation, and termination (BENTLEY 1995 ; GREENBLATT 1997 ). The association of a number of factors with RNApol II occurs via the carboxy terminal domain (CTD) of the Rpb1/Rpo21 subunit of the polymerase (ALLISON and INGLES 1989 ; SCAFE et al. 1990 ) and thus the CTD plays a major role in the transition and regulation of all the steps, along with RNA processing (NEUGEBAUER and ROTH 1997 ; BENTLEY 1999 ). The RNApol II is preassembled with the mediator and other polypeptides into a large complex called the holoenzyme (KIM et al. 1994 ). The mediator integrates and transmits information to the polymerase via CTD, mediating both activation and repression of transcription initiation (FLANAGAN et al. 1991 ; KOLESKE and YOUNG 1995 ; MYERS and KORNBERG 2000 ; DAVIS et al. 2002 ), and at some promoters can be recruited before the loading of the core polymerase (BHOITE et al. 2001 ; COSMA et al. 2001 ; PARK et al. 2001 ). RNApol II usually is paused after polymerization of the first nucleotides. These initial pauses are overcome by the action of transcriptional activators (YANKULOV et al. 1994 ; BENTLEY 1995 ; MASON and LIS 1997 ; TANG et al. 2000 ) concomitantly with an increase of the phosphorylation level of the CTD and a dissociation of the mediator (CADENA and DAHMUS 1987 ; WEEKS et al. 1993 ; O'BRIEN et al. 1994 ; OTERO et al. 1999 ).

The RNA polymerization rate during transcription elongation is dependent on the DNA context and on the action of transcription elongation factors (UPTAIN et al. 1997 ; KROGAN et al. 2002 ; SVEJSTRUP 2002 ; SHILATIFARD et al. 2003 ). While the mediator and other initiation factors are dissociated from the RNApol II during promoter clearance, elongation factors are associated with the promoters and travel along the open reading frames (ORFs) during elongation (ZAWEL et al. 1995 ; SVEJSTRUP et al. 1997 ; POKHOLOK et al. 2002 ). The mechanism of action of those factors is heterogeneous (SHILATIFARD 1998 ). Some factors help the elongation complex to overcome nucleosome barriers (ORPHANIDES et al. 1999 ), speed the polymerization rate up and/or down (BENGAL et al. 1991 ; CHAFIN et al. 1991 ; HARTZOG et al. 1998 ; WADA et al. 1998 ), or reactivate arrested polymerases concomitantly to transcript cleavage (WIND and REINES 2000 ).

TFIIS is an archetypical transcription elongation factor that is highly conserved among eukaryotes and is a functional homolog of the GreA and GreB factors from eubacteria (LABHART and MORGAN 1998 ; FISH and KANE 2002 ). TFIIS (also called factor SII and P37) was biochemically isolated as an RNApol II transcription stimulatory factor from mice (SEKIMIZU et al. 1979 ; UENO et al. 1979 ), humans (REINBERG and ROEDER 1987 ), cow (RAPPAPORT et al. 1987 ), and flies (PRICE et al. 1987 ). In yeast, it was originally isolated as an RNApol II transcription elongation factor named P37 (SAWADOGO et al. 1980 , SAWADOGO et al. 1981 ). TFIIS promotes the reactivation of the RNApol II at arrest sites and was suggested to not have a major role in initiation (RAPPAPORT et al. 1987 ; REINBERG and ROEDER 1987 ; REINES et al. 1989 ). RNApol II pauses are classified as arrests if TFIIS is essential for reactivation of elongation and as stalls if TFIIS is not required. TFIIS induces endonucleolytic cleavage, usually releasing dinucleotides if the polymerase is stalled and four or more nucleotides if arrested (GU et al. 1993 ; IZBAN and LUSE 1993A , IZBAN and LUSE 1993B ). The arrested RNA polymerases are formed after backtracking and extrusion of the 3'-end of the RNA from the catalytic center (REEDER and HAWLEY 1996 ; KOMISSAROVA and KASHLEV 1997 ; KIREEVA et al. 2000B ) and the reactivation of RNApol II involves the stimulation by TFIIS of the intrinsic RNA cleavage activity of the polymerase (RUDD et al. 1994 ; KETTENBERGER et al. 2003 ; WEILBAECHER et al. 2003 ). Blockages of RNA polymerases by failures in elongation not only alter gene expression but also can result in genome instability (REINES et al. 1999 ; AGUILERA 2002 ).

In Saccharomyces cerevisiae TFIIS is encoded by the DST1 gene, also known as PPR2 (HUBERT et al. 1983 ; CLARK et al. 1991 ; ARCHAMBAULT et al. 1992 ; NAKANISHI et al. 1992 ). With the cloning of the genes or cDNAs it became evident that the factors SII, P37, IIS, and, surprisingly, a factor called DNA strand transfer (DST; SUGINO et al. 1988 ) were the same. The C terminus of the protein contains a zinc ribbon that defines the TFIIS domain (QIAN et al. 1993A , QIAN et al. 1993B ), also present in the Rpa12, Rpb9, and Rpc11 subunits of RNApol I, II, and III, respectively. TFIIS physically interacts with Rpb1 (WU et al. 1996 ; ARCHAMBAULT et al. 1998 ), and mutations in DST1 show genetic interactions with genes involved in transcription elongation such as CTK1 (JONA et al. 2001 ), RTF1 (COSTA and ARNDT 2000 ), SPT5 (LINDSTROM and HARTZOG 2001 ), SPT16 (ORPHANIDES et al. 1999 ), or RPB1 (ARCHAMBAULT et al. 1992 ).

In spite of all the evidence pointing to the importance of TFIIS in transcription, a dst1 knockout is viable. Moreover, aside from sensitivity to nucleotide-depleting drugs, dst1 cells have no apparent growth defect under several different conditions (EXINGER and LACROUTE 1992 ; NAKANISHI et al. 1992 ). In this study we have undertaken a genetic approach to gain further insight into the role of TFIIS in vivo in the yeast S. cerevisiae. First, to find factors required for cell growth in the absence of TFIIS, we performed a synthetic genetic array (SGA) analysis of a dst1 knockout against the nonessential gene yeast knockout collection. This approach is complementary to a synthetic lethal screen carried out previously (DAVIE and KANE 2000 ). Second, we focused on the characterization of the strongest synthetic interaction uncovered by the SGA.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Media and growth conditions:
Standard media such as rich medium YEPD or synthetic complete (SC) medium with bases and amino acids omitted as specified and sporulation medium were prepared according to standard procedures (SHERMAN et al. 1986 ). AA synthetic media are similar to SC but contain all 20 amino acids, myo-inositol, para-aminobenzoic acid, adenine, and uracil at a final concentration of 85 mg/liter (170 mg/liter for leucine; 17 mg/liter for para-aminobenzoic acid). Unless specified, all the chemicals were purchased from Sigma (St. Louis). Nourserothricin (Werner BioAgents), geneticin/G418, and hygromycin B were added to rich medium at concentrations of 100, 200, and 300 µg/ml, respectively (WACH et al. 1994 ; GOLDSTEIN and MCCUSKER 1999 ). Benomyl, nocodazole, and thiabendazole (TBZ) were added at final concentrations of 25, 7, and 75 µg/ml, respectively, from stock solutions at 10 mg/ml in dimethyl sulfoxide (benomyl and nocodazole) and 25 mg/ml in N,N'-dimethylformamide (TBZ). Doxycycline was added at a final concentration of 20 µg/ml from stock solutions at 5 mg/ml in 50% ethanol. L-Canavanine sulfate and 5-fluoroorotic acid (5-FOA) were added to synthetic medium at concentrations of 100 and 500 µg/ml, respectively. Plates of SC + FOA medium were prepared by using 1 g/liter proline as the nitrogen source. Plates of 6-azauracil (6-AU) medium were prepared by using freshly made 100 mg/ml stock solutions in H₂O added to AA medium lacking uracil at concentrations of 10 or 3 mg/ml. Mycophenolic acid was added to AA complete medium at 50 µg/ml. Formamide was added to the media to final concentration of 3%. All yeast strains were grown at 30°.

Genetic analysis, manipulations, and strains:
Genetic analysis and manipulations were performed according to published procedures (ORR-WEAVER et al. 1981 ; ITO et al. 1983 ; SHERMAN et al. 1986 ; SCHIESTL and GIETZ 1989 ; ROTHSTEIN 1991 ). Tetrads were dissected on YEPD and incubated at 30° following the growth phenotype of the colonies every day for 3–8 days.

All yeast strains used belong to or are derivatives of the S. cerevisiae 288C BY series of the yeast knockout collection (GIAEVER et al. 2002 ) and were purchased from either EUROSCARF (Frankfurt, Germany) or ResGen. All the mutations relevant to this study are replacements of the corresponding ORFs by the kanMX4, hphMX4, or natMX4 cassettes. All the strains and details of any particular strain used in this work will be provided upon request. Relevant strains are shown in Table 1.

To facilitate the analysis of synthetic interactions, we routinely changed the markers of the deletions. The kanMX4, hphMX4, and natMX4 cassettes were replaced in vivo in yeast by transforming with PCR fragments obtained using the oligos Cassette-i-A (GTCACCCGGCCAGCGACATGGA) and Cassette-i-B (CGAATCGACAGCAGTATAGCGACCAGCATT) and by selecting for the corresponding drug resistance. We amplify the cassettes with 358 bp 5' to the ATG and 214 bp 3' to the end of the ORF. We took advantage of the fact that the cassettes were made using the same promoter and terminator providing homology regions for the PCR and for homologous recombination in yeast. As templates for the PCR we used the plasmids pFA6a-kanMX4 (containing the kanMX4 cassette), pAG32 (containing the hphMX4 cassette), and pAG25 (containing the natMX4 cassette) (WACH et al. 1994 ; GOLDSTEIN and MCCUSKER 1999 ).

We constructed GRY3001 to allow us to follow the 6-AU phenotype of the different mutants. Yeast strains carrying the trp1::hisG-URA3-hisG deletion, in which 546 bp of the TRP1 ORF plus 139 bp of the region upstream of the ATG was removed, were obtained by transforming the yeast strain GRY3000 with a TRP1 blaster plasmid (D. GOTTE and J. N. STRATHERN, unpublished results) containing the hisG-URA3-hisG direct repeats (ALANI et al. 1987 ) and selecting for Ura+ Trp– transformants. The proper integration was demonstrated genetically by cosegregation of Ura+ Trp– in tetrad analysis, papillation on FOA, and linkage to the TRP1 locus. The loss of the URA3 marker by direct repeat recombination can be selected by growth on FOA. Other strains carrying trp1::hisG-URA3-hisG were made by genetic crosses with GRY3001 or derivatives.

We constructed a conditional allele of DST1 by replacing its natural promoter from position –138 to –1 (both included) by the kanMX4 cassette, the doxycycline-sensitive tTA activator gene plus the bacterial tetO2 promoter. This allele of DST1 (PtetDST1) is expressed constitutively and can be repressed by the addition of doxycycline to the media. GRY3003 was constructed by transforming GRY3002 with a 3.9-kb PCR fragment obtained using pCM224 (BELLI et al. 1998 ) and the oligos dst1tet-A (TTGATATATAATATCCAATTTCATTATAGGGAAATTTTCACAGCTGAAGCTTCGTACGCT) and dst1tet-B (CTAGATTCTTAACATGTACCAGTACTTCCTTACTATCCATACTAGTGGATCTGATAAACG) and by selecting for G418R transformants. The sequences in boldface type correspond to the region of homology with the DST1 locus. The proper integration and the absence of mutations was monitored by PCR using oligos tet-A (AAGCGAATTTCTTATGATTT) and dst1-Bintern (CGGACTCACCTACAGTTGTC) and by subsequent sequencing of the PCR fragment. The kanMX4 cassette was replaced by the hphMX4 as described above, creating the strain GRY3004.

SGA analysis:
SGA analysis was carried out as described (TONG et al. 2001 ; JORGENSEN et al. 2002 ). The starting strain, Y3803, which contains a marked deletion of DST1, was constructed by crossing the Y3656 strain with the MATa dst1::kanMX4 strain from the yeast BY4741 knockout collection, replacing the kanMX4 cassette with the natMX4 cassette, transforming the diploid with the p4339 plasmid cut with EcoRI and selecting for the desired markers after sporulation. The MAT dst1::natMX4 starting strain was mated by 4700 individual MATa xxx::kanMX4 haploid deletion strains.

Oligonucleotides, plasmids, and genomic libraries:
Oligonucleotides were purchased from Invitrogen (San Diego). The oligos, presented always in 5'–3' orientation with artificial sequences introduced for cloning purposes underlined, are described below.

• YEplac181 is a 2µ multicopy plasmid carrying the LEU2 gene for selection in yeast (GIETZ and SUGINO 1988 ).

• pRS415 is a centromeric plasmid carrying the LEU2 gene for selection in yeast (SIKORSKI and HIETER 1989 ).

• pRS415-SOH1 was constructed by cloning a 1.1-kb XbaI-XhoI SOH1-containing fragment obtained by PCR into pRS415 opened with XbaI and XhoI. PCR amplification was made using the oligos SOH1-5'XbaI (AAATCCACTATTCCATCTAGACCACACTGTCAGTATGGCAT), SOH1-3'XhoI (GATTTGGATGGATTTCTCGAGATCTTCTAATGTCTTGCGAG), and S288C genomic DNA as template.

• pRS415-soh1-nsiI* was constructed by opening pRS415-SOH1 with NsiI, subsequently treating with Klenow, and religating the plasmid. The original NsiI site lies in the position +213 of SOH1 ORF.

• YEplac181-SOH1 was constructed by cloning a 1.1-kb XhoI-XbaI SOH1 fragment obtained by PCR using the oligos SOH1-5'XbaI and SOH1-3'XhoI (described above) of plasmid pRS415-SOH1 into YEplac181 opened with SmaI and XbaI. The XhoI site of the insert was made blunt with Klenow prior to ligation.

• YEplac181-DST1 was constructed by cloning a 1.8-kb BamHI-SalI DST1 fragment obtained by PCR using the oligos dst1-5'bam (AATACGACTATTCCAGGATCCCGTATGGTATAGAACCCAGAT) and dst1-3'xho1 (GATTTGGATGGATTTCTCGAGATCTTAAATTGTATTTCTTTA) into YEplac181 opened with BamHI and SalI.

• YEplac181-TAF14 was constructed by cloning a 1.6-kb EcoRI-HpaI TAF14 fragment obtained by cutting a 1.8-kb PCR using the oligos TAF14-A (TTCAGACGTCACAGGAATTCATATTGTTAATA) and TAF14-B (AAGAGGATTCACATGGGCAAAAT) into YEplac181 opened with EcoRI and SmaI.

• YEplac181-RTS1 was constructed by cloning a 3.4-kb RTS1 fragment obtained by PCR using the oligos RTS1-A (TTTCACGACTTGACTGTGAG) and RTS1-B (CGAAGATATATTTGGAGAAA) into YEplac181 opened with SmaI. The insert was made blunt with Klenow prior to ligation.

• YEplac181-SUA7 was constructed by cloning a 1.9-kb SUA7 fragment obtained by cutting with NsiI after treatment with Klenow a 2.7-kb PCR using the oligos SUA7-A (AATGATCCGTTTTATTGG) and SUA7-B (GTTCTCGCCTAAGTCATT) into YEplac181 opened with PstI and SmaI.

• YEplac181-GAL11 was constructed by cloning a 4.4-kb GAL11 fragment obtained by PCR using the oligos GAL11-A (GCAAAAGAAGCGGCGAGG) and GAL11-B (CGGCCTCATCAAAACATT) into YEplac181 opened with SmaI. The insert was made blunt with Klenow prior to ligation.

• YEplac181-TYS1 was constructed by cloning a 2.2-kb PstI-SacI TYS1 fragment obtained by PCR using the oligos TYS1-A (TTCAGACGTCACAGCTGCAGCATCGGCCTCGACAC) and TYS1-B (CTGTGACGTCTGAAGAGCTCGGTGGTAAAAAAACT) into YEplac181 opened with PstI and SacI.

• p4339 was made by cloning into pCR2.1-TOPO (Invitrogen) a PCR fragment obtained using the oligos MX4-Forward (ACATGGAGGCCCAGAATACCC) and MX4-Reverse (CAGTATGCGACCAGCATTCAC), which provided 55 bp of homology on both sides of the cassette ORFs and using pAG25 as template.

We used two multicopy genomic libraries to isolate multicopy suppressors. The MW90 library is a YEp351-based episomic LEU2 gene bank (WALDHERR et al. 1993 ). The second library is a 2µ URA3 library based on pHR81 (NEHLIN et al. 1989 ). MW90 candidates were sequenced using oligos YEp351-A (GCGGATAACAATTTCACACAGGAAA) and YEp351-B (ATTAAGTTGGGTAACGCCAGGGTTT). pHR81 candidates were sequenced using oligos pHR81-A (AAAGGGGGATGTGCTGCAAG) and pHR81-B (CTGCCACTCCTCAATTGGATTAGTC).

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

SGA analysis reveals synthetic growth phenotype of dst1 with deletions of YAL011w, YBR231c, VPS72, SOH1, FYV4, VPS71, GIM5, and HTZ1:
To identify genes with functions related to TFIIS, we crossed a dst1 mutant with the arrayed collection of nonessential gene knockout mutants. The resulting diploids were scored for the ability to give rise by meiosis to the corresponding double mutants by the protocol described previously (TONG et al. 2001 ). The SGA showed a reproducible synthetic growth phenotype (SGP) of a dst1::natMX4 allele with deletions of the following 30 ORF/genes: YAL011w, YBR231c, YDL066w/IDP1, YDL068w, YDR485c/VPS72, YGL066w/SGF73, YGL124c/MON1, YGL127c/SOH1, YGL215w/CLG1, YHR013c/ARD1, YHR059w/FYV4, YHR108w/GGA2, YJL115w/ASF1, YJR145c/RPS4A, YJL211c, YLR085c/ARP6, YLR087c/CSF1, YLR114c, YLR174w/IDP2, YLR268w/SEC22, YLR384c/IKI3, YLR410w/VIP1, YML032c/RAD52, YML041c/VPS71, YML094w/GIM5, YMR035w/IMP2, YMR038c/LYS7, YNL297c/MON2, YOL012c/HTZ1, and YPR024w/YME1.

All candidates were retested after dissection of at least 15 tetrads of each cross onto YEPD medium by visualizing the SGP. Even though some synthetic interactions with dst1 were evident on minimal medium, such as rad52, or under different stress conditions, only those that showed a clear SGP in rich medium were selected for further characterization. This reduced the number of positive candidates for genetic interaction with DST1 to eight SDO (synthetic growth with Dst one) genes: YAL011w, YBR231c, VPS72, SOH1, FYV4, VPS71, GIM5, and HTZ1 (see Fig 1A). Among the SDO genes the synthetic growth phenotype is significantly stronger in the case of soh1 where we were hardly able to see the formation of microcolonies 5 days after the dissection of the tetrads (Fig 1A).

View larger version (39K): In this window In a new window Download PPT slide   Figure 1. SGP and salt sensitivities. (A) SGP of dst1, rpb9, and soh1 with sdo mutants. (Top) Tetrads of the crosses of dst1 by sdo1, sdo2, and soh1, followed by (bottom) a representation of multiple synthetic phenotypes. The absence of SGP is represented by an open circle, the presence of SGP as a small open circle, and severe SGP (S-SGP) as an outlined circumference of a black circle. na, not applicable. (B) Sensitivities to salt of single sdo and double dst1 sdo mutants. Fiftyfold dilutions for each strain were placed on YPED and YPED + 1 M NaCl and pictures were taken after 2 and 3 days, respectively.

Deletion of both YML094w/GIM5 and the complementary ORF YML094c-A shows a SGP with dst1. Gim5 is a subunit of the prefolding chaperone or GIM complex (GEISSLER et al. 1998 ; VAINBERG et al. 1998 ). Some of the other subunits of this complex are Yke2/Gim1 and Gim3. To ensure that the SGP with yml094w mutant was due to the absence of Gim5 and not to the putative Yml094c protein, we analyzed the crosses dst1 by gim3 and dst1 by yke2. Both yke2 and gim3 showed a synthetic phenotype with dst1 (not shown).

Some mutants previously described to have a synthetic lethal phenotype with dst1 were not uncovered by the SGA approach. Among them are kex2, ctk1, snf2, and snf5 (DAVIE and KANE 2000 ; JONA et al. 2001 ). To evaluate the level of false negatives obtained by this method, we crossed dst1 by kex2, ctk1, snf2, and snf5 deletions. All the crosses showed a very clear and strong SGP but not lethality (F. MALAGON and J. N. STRATHERN, unpublished results). Since we followed the tetrads' behavior for 5 days or more, this discordance may be due to differences in criteria. In any case, the fact that kex2, ctk1, snf2, and snf5, as well as srb5 (see Synthetic growth phenotype of dst1 with deletions of SRB5 and SIN4) did not show up as positive candidates suggests that the scrutiny for nonessential synthetic lethal mutants with dst1 is still incomplete.

We tested the significance of the SGP found with dst1 by crossing the sdo mutants by an rpb9::hphMX4 deletion strain. We chose this gene because Rpb9 is an RNApol II subunit and acts coordinately with TFIIS. Rpb9 is essential for TFIIS to stimulate the readthrough and transcript cleavage activity of the RNApol II (AWREY et al. 1997 ) and rpb9 mutants are epistatic to dst1 (VAN MULLEM et al. 2002 ). The double-mutant dst1 rpb9 did not show an SGP by our criteria. The results of the crosses by rpb9, summarized in Fig 1A, are very similar to those obtained with dst1 except that rpb9 has a stronger synthetic phenotype with gim5 and htz1. The synthetic interactions of RPB9 with YAL011w, YBR231c, VPS72, SOH1, FYV4, VPS71, GIM5, and HTZ1 reinforce the SGA results and also the importance of Rpb9 for TFIIS activity.

Finally, we tested the SGP during vegetative growth. We did that by plating for single colonies on rich medium of the single dst1, sdoX (any of the single sdo mutants), and double dst1 sdoX mutants and comparing the growth. All the double dst1 sdoX mutants showed a slight but consistent retardation of colony growth with respect to the parental strains (not shown). A remarkable exception was the dst1 soh1 mutant, which clearly grew more slowly than the single dst1 or soh1 strains (see Fig 1B). It has been reported that TFIIS is recruited to the transcribed genes under stress but not under optimal growth conditions (POKHOLOK et al. 2002 ). This led us to test the SGP phenotype at 37° in minimal media and in high salt concentration. We found no differences for growth at 37° or in minimal media but a clear and generalized accentuation of the SGP on 1 M NaCl (Fig 1B). Neither dst1 nor sdoX, with the exception of gim5, is sensitive to 1 M NaCl. This may indicate that TFIIS is required during germination and at high salt concentrations in the absence of the SDO genes and in the absence of SOH1 during optimal growth conditions.

dst1 mutants are sensitive to microtubule-destabilizing drugs:
To understand why the SDO genes show the SGP with dst1, we tried to find common features among them. According to the Saccharomyces Genome Database (http://www.yeastgenome.org/), only GIM5 and HTZ1 have a molecular function assigned (summarized in Fig 2A). Gim5 has tubulin-binding activity and is involved in the folding of actin and tubulin (GEISSLER et al. 1998 ; VAINBERG et al. 1998 ) and Htz1 is a histone variant involved in silencing and regulation of transcription from RNApol II promoters (JACKSON and GOROVSKY 2000 ; SANTISTEBAN et al. 2000 ; ADAM et al. 2001 ; MENEGHINI et al. 2003 ). SOH1 is involved in genome stability and genetically interacts with components of the RNApol II holoenzyme (FAN and KLEIN 1994 ; FAN et al. 1996 ). VPS71 and VPS72 were isolated in a screening for protein-vacuolar targeting mutants (BONANGELINO et al. 2002 ) and FYV4 was isolated as sensitive to K1 killer toxin (PAGE et al. 2003 ).

View larger version (43K): In this window In a new window Download PPT slide   Figure 2. Sensitivities to microtubule-destabilizing drugs and formamide of dst1, rpb9, and sdo mutants. (A) Molecular functions and biological processes assigned to the SDO genes (http://www.yeastgenome.org). (B) Sensitivity of dst1, rpb9, and sdo mutants to nocodazole, benomyl, TBZ, and formamide. Tenfold dilutions for each strain were placed on YPED or YEPD plus the correspondent chemical, and pictures were taken after 3 days for YEPD and YEPD + formamide and after 4 days for YPD + nocodazole, benomyl, and TBZ.

The fact that the absence of Gim5 and other components of the GIM complex increases the requirement of TFIIS for cell growth may indicate that improper folding of actin and tubulin interferes with transcription. The correct microtubule and microfilament network architecture in the cell are important for a large number of different biological processes (YARM et al. 2001 ; BARR 2002 ; SCHOTT et al. 2002 ). A coordinated input-output of tubulin subunits is essential for the biological function of the microtubules for such processes as chromosome disjunction in mitosis. Actin cables have a major role in organelle transport, and actin patches function in endo- and exocytosis. Additionally, a direct interaction has been shown between the microtubules/microfilaments and mRNA distribution (NASMYTH and JANSEN 1997 ; TAKIZAWA et al. 1997 ; OLEYNIKOV and SINGER 1998 ; BEACH et al. 1999 ; JANSEN 1999 ) and also a direct role of nuclear actin in transcription elongation (PERCIPALLE et al. 2003 ). We reasoned that the phenotypes of some of the sdo mutants might also be explained by a deficiency in actin/tubulin organization. This is particularly likely in the cases of GIM5, VPS71, and VPS72. To test this hypothesis, we checked the sensitivity of all of them to the microtubule-destabilizing drug nocodazole. Six of the sdo mutants were highly sensitive to nocodazole, while the soh1 and fyv4 mutants were not (Fig 2B). Furthermore, dst1 and rpb9 mutants showed a strong hypersensitivity to nocodazole. The fact that rpb9 is also sensitive to nocodazole, along with the described biochemical activities of TFIIS, suggests that this new phenotype for dst1 is directly related to its role in transcription rather than to a direct structural function in microtubule organization. Similar results were obtained with other microtubule-destabilizing drugs such as benomyl and TBZ (see Fig 2B), with the only exception that soh1 and fyv4 mutants are slightly more resistant than the wild type to benomyl.

It has been reported that htz1 mutants show a strong sensitivity to formamide (JACKSON and GOROVSKY 2000 ). The molecular mechanism that renders cells sensitive to formamide is unknown. We wondered whether this phenotype could be generalized to the rest of the sdo mutants and to dst1. As indicated in Fig 2B, all the sdo mutants except gim5 show a strong sensitivity to formamide. Interestingly, we found discordance in the behavior of the dst1 (resistant) and rpb9 (sensitive) mutants for this phenotype. We used the formamide sensitivity as an experimental tool for dissecting genetic interactions with dst1 and soh1 (see below).

Characterization of the genetic interaction of soh1 and dst1:
The SOH1 gene is highly conserved in evolution from yeast to humans. It was originally isolated as a suppressor of the thermosensitivity of hpr1 mutants in yeast (FAN and KLEIN 1994 ) and has a role in genome integrity. In mammals, the role of Soh1 in transcription is well defined as an integral component of the mediator complexes (GU et al. 1999 ). In Saccharomyces, SOH1 is synthetically lethal with mutations in genes encoding two subunits of the RNApol II, Rpo21/Rpb1 and Rpb2, the transcription initiation factor II B (Sua7/TFIIB; FAN et al. 1996 ), and with the histone H3 methyl transferase factor Set2 (KROGAN et al. 2003 ). We focused specifically on the interaction DST1-SOH1 for three reasons: (1) it showed the strongest SGP; (2) soh1 mutants are not hypersensitive to nocodazole, decreasing the probability of an indirect SGP mediated by a microtubule effect; and (3) the published data on SOH1 suggest a direct role in RNApol II transcription.

First, we asked if soh1 mutants show synthetic interaction with the other sdo genes. soh1 has a severe SGP with yal011w, ybr231c, vps71, vps72, and fyv4 and no synthetic phenotype with gim5 and htz1 (Fig 1A). Moreover, among the other mutants originally pulled out in the SGA with dst1, soh1 has a SGP with ard1, arp6, and yme1 (not shown). The common synthetic behavior of soh1 and dst1 with other mutants reinforces the genetic interaction between SOH1 and DST1.

Second, we tested the sensitivity of soh1 and the other sdo mutants to the nucleotide-depleting drug 6-azauracil and/or to mycophenolic acid. Transcription elongation mutants usually show sensitivity to one or both of those drugs due to a deficient induction of IMD2/PUR5 (EXINGER and LACROUTE 1992 ; SHAW and REINES 2000 ). Neither soh1 nor the rest of the sdo mutants are sensitive to 6-azauracil by our criteria and only htz1 is sensitive to mycophenolic acid (not shown). These results argue against a direct role of soh1 in transcription elongation.

Finally, we constructed a conditional soh1 dst1 double mutant for subsequent genetic approaches (see next section). For this purpose, we replaced the endogenous DST1 promoter with an artificial promoter that can be partially repressed by the drug doxycycline (see MATERIALS AND METHODS). As expected, the addition of doxycycline to a Ptet-DST1 strain has no noticeable growth phenotype, while repression of Ptet-DST1 in a soh1 background causes a severe growth phenotype (Fig 3). This phenotype can be rescued by the introduction of the SOH1-carrying plasmid pRS415-SOH1, but not with pRS415-soh1-nsiI* carrying a 4-bp deletion of SOH1 (not shown).

View larger version (24K): In this window In a new window Download PPT slide   Figure 3. Effect of the repression of DST1 in a soh1 mutant. Diagram of the PtetDST1 conditional allele and effect of repression of the tet promoter with doxicyclin in a PtetDST1 and a PtetDST1 soh1 strain. The behavior of wild type, dst1, and soh1 strains is similar to that of the single PtetDST1 (not shown). Tenfold dilutions for each strain were placed on YPED or YEPD + doxycyclin and pictures were taken after 3 days.

Isolation and identification of multicopy suppressors of the synthetic growth defect of dst1 soh1:
To understand the function of DST1 and SOH1 in transcription and the mechanism by which those genes are required for cell viability, we searched for genes that suppressed the synthetic growth phenotype of a dst1 soh1 double mutant by overexpression. Such genes might have functions partially related to DST1, SOH1, or both. To do so, we isolated genes that suppressed the inability of a PtetDST1 soh1 strain to grow under repression of the tet promoter. We used two different multicopy libraries (see MATERIALS AND METHODS). After isolation and retransformation of the candidates, we selected those with a clear suppression phenotype. We also obtained several other candidates with a reproducible but very weak phenotype that we did not characterize. We sequenced the selected candidates and subcloned the genes into YEplac181 to have a single gene insert and to homogenize the multicopy vector harboring them. The genes that are able to suppress the SGP in the PtetDST1 soh1 strain are DST1, SOH1, TAF14, RTS1, SUA7, GAL11, and TYS1 (see Fig 4A). To eliminate any possible artifact due to an effect on the artificial Ptet promoter, we tested the ability of all of them to suppress the SGP in a soh1 dst1 strain. As shown in Fig 4B, all can complement the double-deletion strain.

View larger version (67K): In this window In a new window Download PPT slide   Figure 4. Suppression of growth phenotypes of dst1, soh1, and dst1 soh1 by the presence of multiple copies of TAF14, RTS1, SUA7, GAL11, or TYS1. (A) Growth phenotype of PtetDST1 soh1, under expression or repression of the tet promoter, and double dst1 soh1 strains transformed with YEplac181 (LEU2 containing multicopy vector) or the corresponding gene cloned in the same vector. Tenfold dilutions for each strain were placed on AA-Leu or AA-Leu + doxicyclin and pictures were taken after 3 days. (B) Sensitivity to formamide of a soh1 strain transformed with the plasmids described in A. The spots contained 105 cells plated on AA-Leu + formamide and the pictures were taken after 4 days. (C) Sensitivity to nocodazole of a dst1 strain transformed with the plasmids described in A. The spots contained 5 x 103 cells plated on YEPD + formamide and the pictures were taken after 4 days.

Sua7, Taf14, and Gal11 proteins are integral components of the RNApol II holoenzyme (KOLESKE and YOUNG 1995 ; MYER and YOUNG 1998 ). SUA7 (suppressor of upstream ATG; HAMPSEY et al. 1991 ) is an essential gene and functions in site selection for transcriptional initiation, and mutants in SUA7 shift start-site selection downstream of normal (PINTO et al. 1992 ). SUA7 was also isolated as a suppressor of hpr1 (an alias of SUA7 is SOH4) and point mutants of this gene are synthetic lethal with soh1 (FAN and KLEIN 1994 ; FAN et al. 1996 ). Taf14/Taf30 is a promiscuous protein that physically interacts with or belongs to several components of the RNApol II or accessories complexes such as TFIID, TFIIF, Swi/Snf, and the mediator (GAVIN et al. 2002 ; SANDERS et al. 2002 ). Gal11, also found associated with several complexes of the RNApol II, has a major role in the recruitment of the holoenzyme to the promoters and is an integral component of the mediator (MYER and YOUNG 1998 ; LEE et al. 1999 ; SANDERS et al. 2002 ).

Rts1 is a protein serine/threonine phosphatase 2A regulatory subunit (SHU et al. 1997 ; EVANS and HEMMINGS 2000 ) involved in stress-related responses (EVANGELISTA et al. 1996 ; ZABROCKI et al. 2002 ). While Rts1 is not directly associated with the holoenzyme, it should be pointed out that RTS1 (ROX three suppressor) is a multicopy suppressor of another component of the mediator (EVANGELISTA et al. 1996 ; GUSTAFSSON et al. 1997 ). There is no obvious explanation for the isolation of TYS1, an essential gene that encodes a tyrosyl-tRNA synthetase with a role in tRNA export and cell wall formation (SARKAR et al. 1999 ; AZAD et al. 2001 ; DAGKESSAMANSKAIA et al. 2001 ).

We used the differential sensitivity of dst1 and soh1 mutants to nocodazole and formamide to define the multicopy suppressors as specific to any of the two genes or to the double-mutant interaction. Among them SOH1, TAF14, RTS1, SUA7, and TYS1 were able to suppress, at least partially, the formamide sensitivity of a soh1 single mutant (Fig 4B). Only DST1 suppressed the sensitivities to 6-azauracil or mycophenolic acid of a single dst1 mutant (not shown). Nevertheless, GAL11 suppressed the sensitivity to nocodazole of a dst1 mutant in spite of the toxicity of overexpressing GAL11 (see Fig 4B).

Synthetic growth phenotype of dst1 with deletions of SRB5 and SIN4:
The deletion of DST1 can be partially suppressed by overexpression of a well-defined subunit of the mediator, Gal11, and shows synthetic growth phenotype with soh1 mutants lacking a putative subunit of the mediator. To determine further interactions with mediator subunits, we crossed a dst1 mutant with strains carrying deletions of GAL11, HRS1/PGD1, SIN4, MED1, SRB8, SRB9, ROX3, SRB2, and SRB5. Among them, dst1 showed a severe synthetic growth phenotype with srb5 (see Fig 5A) and a synthetic phenotype with sin4 at 37° (see Fig 5B). The rest of the mutants did not show a synthetic phenotype with dst1 in the conditions tested. Srb5 belongs to the Srb4 subcomplex of the mediator involved in basal transcription (THOMPSON et al. 1993 ; KIM et al. 1994 ; KANG et al. 2001 ), and Sin4 belongs to the Gal11 module of the Rgr1 subcomplex of the mediator (LI et al. 1995 ; KANG et al. 2001 ). These results extend the genetic interactions of DST1 to other genes encoding initiation factors.

View larger version (36K): In this window In a new window Download PPT slide   Figure 5. Genetic interaction of DST1 with SRB5 and SIN4. (A) SGP of dst1 with srb5 mutants. We show one tetratype tetrad of the cross of dst1 by srb5 as an example. The picture was taken after 3 days. (B) Growth phenotype of single dst1 and sin4 and double dst1 sin4 mutant. Fiftyfold dilutions for each strain were placed on YPD at 30° and 37°. The pictures were taken after 2 days.

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Function of the SDO genes in transcription:
We identified new synthetic interactions of DST1 with genes involved in transcription, vesicular transport, folding of actin and tubulin, resistance to killer toxin, and two ORFs of unknown function. Among the genes uncovered using the SGA approach, HTZ1 and SOH1 participate in RNApol II transcription, GIM5 encodes for a chaperone of actin and tubulin, and the rest are very poorly characterized. The fact that all the SDO genes also show synthetic interaction with RPB9, encoding a subunit of the RNApol II, may suggest that some other genes revealed by the SGA may also have a transcription-related function.

Even though the precise molecular functions of Htz1 and Soh1 are not clear, both proteins seem to have a major role in transcription initiation. Htz1 performs redundant functions with the Swi/Snf complex (SANTISTEBAN et al. 2000 ), localizes mainly in the promoters, and has been suggested to act as a transcription activator (LAROCHELLE and GAUDREAU 2003 ). Several lines of evidence suggest that Soh1 belongs to the RNApol II holoenzyme. First, SOH1 was isolated as a suppressor of hpr1 (FAN and KLEIN 1994 ). Other genes isolated as suppressors of mutations of HPR1 are components of the holoenzyme such as SOH2/RPB2, SOH4/SUA7, HRS1/PGD1/MED3, and HRS2/SRB2 (SANTOS-ROSA and AGUILERA 1995 ; FAN et al. 1996 ; PIRUAT and AGUILERA 1996 ; SANTOS-ROSA et al. 1996 ; PIRUAT et al. 1997 ). Second, soh1 mutants are synthetically lethal with mutations in RPB1, RPB2, and SUA7 (FAN et al. 1996 ). And third, SOH1 is a highly conserved gene found to be an integral part of the mediator in humans and in Drosophila (BOUBE et al. 2000 ; GU et al. 2002 ). The high conservation of the mediator subunits in evolution suggests that Soh1 is also a component of the mediator in yeast (ASTURIAS et al. 1999 ; DOTSON et al. 2000 ; BOUBE et al. 2002 ). Moreover, the Srb5 and Sin4 subunits of the mediator are required for cell growth in the absence of TFIIS at 30° and 37°, respectively.

Sensitivity to microtubule-destabilizing drugs and salt:
The SGA and tetrad analysis of the sdo mutants helped us to define a new phenotype for dst1 mutants. We found a connection with microtubule metabolism for TFIIS and showed that dst1 and rpb9 mutants are sensitive to nocodazole, benomyl, and TBZ. The simplest hypothesis to explain the nocodazole hypersensitivity phenotype of dst1 is that TFIIS is required for expression of tubulin-related genes. We cannot exclude other possibilities such as a direct interaction of TFIIS with proteins controlling the modulation of microtubules/microfilaments. In this sense, a Ca²⁺-dependent association of TFIIS with Cmd1/calmodulin has been reported (STIRLING et al. 1994 ).

The same type of hypothesis can be formulated to explain the salt sensitivity of the double mutants dst1 sdoX. The simplest hypothesis is that TFIIS regulates the expression of some genes involved in salt tolerance. In agreement with this idea, it has been reported that mRNA levels in general and specifically of ENA1, a gene required for salt tolerance, are severely affected in a dst1 background when combined with the rpb2-10 allele (LENNON et al. 1998 ; WIND-ROTOLO and REINES 2001 ).

Function of the multicopy suppressors of soh1 dst1:
To further characterize the strong genetic interaction between DST1 and SOH1, we isolated multicopy suppressors of the SGP of dst1 soh1 mutants and found five genes that were strong suppressors without fully complementing the wild-type growth level: GAL11, TAF14, SUA7, RTS1, and TYS1. The isolation of GAL11, TAF14, and SUA7, encoding components of the RNApol II holoenzyme, as multicopy suppressors of the SGP of dst1 soh1 mutants clearly points out that the growth defect of the double mutant is most likely due to a defect in initiation of transcription. Especially relevant is the isolation of GAL11, a subunit of the mediator, as a specific suppressor of dst1. Genetic interactions, such as suppression of gal11 by overexpression of the CTD kinase domain of Sgv1/Bur1, suggest a role for Gal11 specifically in the initiation-elongation transition (BADI and BARBERIS 2001 ).

Does TFIIS participate in transcription initiation?
Our genetic characterization of dst1 mutants indicates an interaction of TFIIS and functions involved in the initiation of transcription. We and others (DAVIE and KANE 2000 ) have shown that the absence of Htz1, Soh1, or the Swi/Snf complex increases the requirement of TFIIS for cell viability. Surprisingly, neither DAVIE and KANE 2000 nor we were able to pull out transcription elongation genes as synthetic with dst1. On the other hand, overexpression of Gal11 can partially rescue phenotypes associated with the absence of TFIIS, and the growth defect of the dst1 soh1 double mutant is suppressed by high copy number of SUA7 or TAF14. Could this indicate that TFIIS has a role in initiation? Some reports show that the functional homologs of TFIIS in eubacteria, GreA and GreB, have a role in initiation of transcription (HSU et al. 1995 ; SEN et al. 2001 ). There is, to our knowledge, no molecular evidence for a role of TFIIS in initiation. Nevertheless, some data, like the synthetic interaction of dst1 with kin28 point mutants (LINDSTROM and HARTZOG 2001 ), suggest that TFIIS may contact the initiation complex.

What features of initiation might require the action of TFIIS? After the abortive initiation mode is finished, there are promoter-proximal blocks in the first 50 nucleotides in the absence of gene-specific activators that may facilitate the mediator-elongator exchange (OTERO et al. 1999 ). This kind of arrest or halt of the RNApol II after the synthesis of 20–40 nucleotides is a well-established mechanism of control of transcription at certain genes like c-myc (STROBL and EICK 1992 ; KRUMM et al. 1995 ; PLET et al. 1995 ; SCHNEIDER et al. 1999 ), c-fos (FIVAZ et al. 2000 ), and DHRF and ACTG1 (CHENG and SHARP 2003 ) in humans or hsp70 in Drosophila (LEE et al. 1992 ; RASMUSSEN and LIS 1995 ; LI et al. 1996 ; TANG et al. 2000 ). Certainly the pausing of the RNApol II during initiation is an important regulatory point in transcription of specific genes and, consequently, antiarrest activities may play a role in initiation.

A model for a role of TFIIS in the initiation-elongation transition:
We imagine two mechanisms for the transition from initiation to elongation, one dependent on initiation factors like the mediator complex and another dependent upon TFIIS. In addition to a possible role in transcription elongation in stressed cells (POKHOLOK et al. 2002 ), we propose that TFIIS provides a mechanism to restart arrested initiation complexes in soh1 or srb5 mutants. We envision the arrest of the RNApol II shortly after polymerization of the first ribonucleotides as the cause of the SGP of soh1 dst1 mutants. The overexpression of initiation factors such as TFIIB/Sua7 or Gal11 may alter the interactions in the initiation complex promoting the advance of the RNApol II toward the elongation phase of transcription. Soh1 and Srb5 may act to prevent the arrest of the polymerase by either enhancing promoter escape or inhibiting the backtracking of the polymerase.

The TFIIS activity facilitating the cleavage and readthrough of the nascent RNA may be directly required to overcome RNApol II initiation-elongation arrests. The hypothesis for this role of TFIIS in vivo is strongly supported by in vitro studies with RNApol II carried out by D. S. Luse's lab. Researchers in this lab have shown that the RNApol II is arrest prone during promoter clearance (PAL et al. 2001 ) and that this phenomenon can be recreated in promoter-distal locations by shortening the nascent RNA or by hybridizing oligonucleotides to the transcript between 30 and 45 bp upstream of the 3'-end (UJVARI et al. 2002 ). Those arrests are accompanied by upstream translocation of the RNA polymerase as shown by the increase in the proportion of long TFIIS cleavage products (UJVARI et al. 2002 ). These in vitro data clearly show that TFIIS promotes the resumption of arrested polymerases in the interphase between initiation and elongation of transcription. One appealing possibility is that TFIIS acts not just to reactivate arrested complexes, but also to prevent the pausing of the polymerase during initiation by avoiding the hybridization of the 5'-end of the nascent RNA with the transcribing DNA strand. The inhibition of transcription by an overextended RNA-DNA hybrid in vitro has been previously reported (KIREEVA et al. 2000A ). In fact, TFIIS/DST has a DNA strand annealing activity that theoretically can avoid the formation of the overextended RNA-DNA hybrid (SUGINO et al. 1988 ; CLARK et al. 1991 ; KIPLING and KEARSEY 1991 ).

Note added in proof:
While this article was in press, two articles showed that the proteins encoded by VPS72, YAL011w, YBR231c, and VPS71 (renamed as Swc2, Swc3, Swc5, and Swc6, respectively) belong to a protein complex involved in the deposition of Htz1 at specific chromosome locations in vivo (N. J. KROGAN, M. C. KEOGH, N. DATTA, C. SAWA, O. W. RYAN et al., 2003, A Snf2 family ATPase complex required for recruitment of the histone H2A variant Htzl. Mol. Cell 12: 1565–1576; G. MIZUGUCHI, X. SHEN, J. LANDRY, W. H. WU, S. SEN et al., 2004, ATP-driven exchange of histone H2AZ variant catalyzed by SWR1 chromatin remodeling complex. Science 303: 343–348).

	ACKNOWLEDGMENTS

We thank S. Chavez, M. Kashlev, K. Christie, M. Kireeva, and A. Rattray for the critical reading of the manuscript. We also thank A. Aguilera and D. Garfinkel for providing the genomic libraries, D. Gotte for sequencing assistance, and M. Grau and M. Mills for editorial and administrative help. This work was sponsored by the National Cancer Institute, Department of Health and Human Services. The contents of this publication do not necessarily reflect the views or policies of the Department of Health and Human Services, nor does mention of trade names, commercial products, or organization imply endorsement by the U.S. government.

Manuscript received August 5, 2003; Accepted for publication December 17, 2003.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

ADAM, M., F. ROBERT, M. LAROCHELLE, and L. GAUDREAU, 2001  H2A.Z is required for global chromatin integrity and for recruitment of RNA polymerase II under specific conditions. Mol. Cell. Biol. 21:6270-6279.[Abstract/Free Full Text]

AGUILERA, A., 2002  The connection between transcription and genomic instability. EMBO J. 21:195-201.[CrossRef][Medline]

ALANI, E., L. CAO, and N. KLECKNER, 1987  A method for gene disruption that allows repeated use of URA3 selection in the construction of multiply disrupted yeast strains. Genetics 116:541-545.[Abstract/Free Full Text]

ALLISON, L. A. and C. J. INGLES, 1989  Mutations in RNA polymerase II enhance or suppress mutations in GAL4.. Proc. Natl. Acad. Sci. USA 86:2794-2798.[Abstract/Free Full Text]

ARCHAMBAULT, J., F. LACROUTE, A. RUET, and J. D. FRIESEN, 1992  Genetic interaction between transcription elongation factor TFIIS and RNA polymerase II. Mol. Cell. Biol. 12:4142-4152.[Abstract/Free Full Text]

ARCHAMBAULT, J., D. B. JANSMA, J. H. KAWASOE, K. T. ARNDT, and J. GREENBLATT et al., 1998  Stimulation of transcription by mutations affecting conserved regions of RNA polymerase II. J. Bacteriol. 180:2590-2598.[Abstract/Free Full Text]

ASTURIAS, F. J., Y. W. JIANG, L. C. MYERS, C. M. GUSTAFSSON, and R. D. KORNBERG, 1999  Conserved structures of mediator and RNA polymerase II holoenzyme. Science 283:985-987.[Abstract/Free Full Text]

AWREY, D. E., R. G. WEILBAECHER, S. A. HEMMING, S. M. ORLICKY, and C. M. KANE et al., 1997  Transcription elongation through DNA arrest sites. A multistep process involving both RNA polymerase II subunit Rpb9 and TFIIS. J. Biol. Chem. 272:14747-14754.[Abstract/Free Full Text]

AZAD, A. K., D. R. STANFORD, S. SARKAR, and A. K. HOPPER, 2001  Role of nuclear pools of aminoacyl-tRNA synthetases in tRNA nuclear export. Mol. Biol. Cell 12:1381-1392.[Abstract/Free Full Text]

BADI, L. and A. BARBERIS, 2001  Proteins that genetically interact with the Saccharomyces cerevisiae transcription factor Gal11p emphasize its role in the initiation-elongation transition. Mol. Genet. Genomics 265:1076-1086.[CrossRef][Medline]

BARR, F. A., 2002  Inheritance of the endoplasmic reticulum and Golgi apparatus. Curr. Opin. Cell. Biol. 14:496-499.[CrossRef][Medline]

BEACH, D. L., E. D. SALMON, and K. BLOOM, 1999  Localization and anchoring of mRNA in budding yeast. Curr. Biol. 9:569-578.[CrossRef][Medline]

BELLI, G., E. GARI, M. ALDEA, and E. HERRERO, 1998  Functional analysis of yeast essential genes using a promoter-substitution cassette and the tetracycline-regulatable dual expression system. Yeast 14:1127-1138.[CrossRef][Medline]

BENGAL, E., O. FLORES, A. KRAUSKOPF, D. REINBERG, and Y. ALONI, 1991  Role of the mammalian transcription factors IIF, IIS, and IIX during elongation by RNA polymerase II. Mol. Cell. Biol. 11:1195-1206.[Abstract/Free Full Text]

BENTLEY, D. L., 1995  Regulation of transcriptional elongation by RNA polymerase II. Curr. Opin. Genet. Dev. 5:210-216.[CrossRef][Medline]

BENTLEY, D., 1999  Coupling RNA polymerase II transcription with pre-mRNA processing. Curr. Opin. Cell Biol. 11:347-351.[CrossRef][Medline]

BHOITE, L. T., Y. YU, and D. J. STILLMAN, 2001  The Swi5 activator recruits the Mediator complex to the HO promoter without RNA polymerase II. Genes Dev. 15:2457-2469.[Abstract/Free Full Text]

BONANGELINO, C. J., E. M. CHAVEZ, and J. S. BONIFACINO, 2002  Genomic screen for vacuolar protein sorting genes in Saccharomyces cerevisiae.. Mol. Biol. Cell 13:2486-2501.[Abstract/Free Full Text]

BOUBE, M., C. FAUCHER, L. JOULIA, D. L. CRIBBS, and H. M. BOURBON, 2000  Drosophila homologs of transcriptional mediator complex subunits are required for adult cell and segment identity specification. Genes Dev. 14:2906-2917.[Abstract/Free Full Text]

BOUBE, M., L. JOULIA, D. L. CRIBBS, and H. M. BOURBON, 2002  Evidence for a mediator of RNA polymerase II transcriptional regulation conserved from yeast to man. Cell 110:143-151.[CrossRef][Medline]

BRACHMANN, C. B., A. DAVIES, G. J. COST, E. CAPUTO, and J. LI et al., 1998  Designer deletion strains derived from Saccharomyces cerevisiae S288C: a useful set of strains and plasmids for PCR-mediated gene disruption and other applications. Yeast 14:115-132.[CrossRef][Medline]

CADENA, D. L. and M. E. DAHMUS, 1987  Messenger RNA synthesis in mammalian cells is