Mutants in ABC10β, a Conserved Subunit Shared by All Three Yeast RNA Polymerases, Specifically Affect RNA Polymerase I Assembly
1999; Elsevier BV; Volume: 274; Issue: 13 Linguagem: Inglês
10.1074/jbc.274.13.8421
ISSN1083-351X
AutoresOlivier Gadal, George V. Shpakovski, Pierre Thuriaux,
Tópico(s)RNA modifications and cancer
ResumoABC10β, a small polypeptide common to the three yeast RNA polymerases, has close homology to the N subunit of the archaeal enzyme and is remotely related to the smallest subunit of vaccinial RNA polymerase. The eucaryotic, archaeal, and viral polypeptides share an invariant motif CX 2C… CC that is strictly essential for yeast growth, as shown by site-directed mutagenesis, whereas the rest of the ABC10β sequence is fairly tolerant to amino acid replacements. ABC10β has Zn2+ binding properties in vitro, and the CX 2C … CC motif may therefore define an atypical metal-chelating site. Hybrid subunits that derive most of their amino acids from the archaeal subunit are functional in yeast, indicating that the archaeal and eucaryotic polypeptides have a largely equivalent role in the organization of their respective transcription complexes. However, all eucaryotic forms of ABC10β harbor a HVDLIEK motif that, when mutated or replaced by its archaeal counterpart, leads to a polymerase I-specific lethal defect in vivo. This is accompanied by a specific lack in the largest subunit of RNA polymerase I (A190) in cell-free extracts, showing that the mutant enzyme is not properly assembled in vivo. ABC10β, a small polypeptide common to the three yeast RNA polymerases, has close homology to the N subunit of the archaeal enzyme and is remotely related to the smallest subunit of vaccinial RNA polymerase. The eucaryotic, archaeal, and viral polypeptides share an invariant motif CX 2C… CC that is strictly essential for yeast growth, as shown by site-directed mutagenesis, whereas the rest of the ABC10β sequence is fairly tolerant to amino acid replacements. ABC10β has Zn2+ binding properties in vitro, and the CX 2C … CC motif may therefore define an atypical metal-chelating site. Hybrid subunits that derive most of their amino acids from the archaeal subunit are functional in yeast, indicating that the archaeal and eucaryotic polypeptides have a largely equivalent role in the organization of their respective transcription complexes. However, all eucaryotic forms of ABC10β harbor a HVDLIEK motif that, when mutated or replaced by its archaeal counterpart, leads to a polymerase I-specific lethal defect in vivo. This is accompanied by a specific lack in the largest subunit of RNA polymerase I (A190) in cell-free extracts, showing that the mutant enzyme is not properly assembled in vivo. The nuclear genome of eucaryotes is transcribed by three heteromultimeric RNA polymerases that respectively contain 14, 12, and 17 distinct subunits in Saccharomyces cerevisiae. The two largest subunits are related to the β′ and β components of the α2ββ′ bacterial core enzyme and to the equivalent subunits of the archaeal and vaccinial RNA polymerases. Biochemical and genetic studies have established that they harbor the active site of the yeast (1Riva M. Carles C. Sentenac A. Grachev M.A. Mustaev A.A. Zaychikov E.F. J. Biol. Chem. 1990; 265: 16498-16503Abstract Full Text PDF PubMed Google Scholar, 2Dieci G. Hermann-Le Denmat S. Lukhtanov E. Thuriaux P. Werner M. Sentenac A. EMBO J. 1995; 14: 3766-3776Crossref PubMed Scopus (65) Google Scholar) and bacterial (3Zaychikov E. Martin E. Denissova L. Kozlov M. Heuman H. Nikiforov V. Goldfarb A. Mustaev A. Science. 1996; 273: 107-109Crossref PubMed Scopus (150) Google Scholar) enzymes. Homology to the bacterial α subunit, although less pregnant, was also observed with the eucaryotic and archaeal enzymes (4Martindale D.W. Nucleic Acids Res. 1990; 18: 2953-2959Crossref PubMed Scopus (43) Google Scholar, 5Langer D. Hain J. Thuriaux P. Zillig W. Proc. Natl. Acad. Sci. U. S. A. 1995; 92: 5768-5772Crossref PubMed Scopus (262) Google Scholar). The α2ββ′ core polymerase structure is therefore preserved in all eucaryotic, archaeal, bacterial, and viral forms of RNA polymerases. A number of additional subunits are structurally conserved or even strictly identical from one to another polymerases (6Thuriaux P. Sentenac A. Jones E.W. Pringle J.R. Broach J.R. The Molecular Biology of Yeast. 2. Cold Spring Harbor Laboratory, Cold Spring Harbor Laboratory, NY1992: 1-45Google Scholar, 7Woychik N.A. Young R.A. Conaway R.C. Conaway J.W. Transcription Mechanisms and Regulation. Raven Press, Ltd., New York1994: 227-242Google Scholar, 8Chédin S. Ferri M.L. Peyroche G. Andrau J.C. Jourdain S. Lefebvre O. Werner M. Carles C. Sentenac A. Cold Spring Harbor Symp. Quant. Biol. 1998; (in press)PubMed Google Scholar) and are functionally conserved from yeast to man (9Shpakovski G.V. Acker J. Wintzerith M. Lacroix J.F. Thuriaux P. Vigneron M. Mol. Cell. Biol. 1995; 15: 4702-4710Crossref PubMed Scopus (116) Google Scholar, 10McKune K. Moore P.A. Hull M.W. Woychik N.A. Mol. Cell. Biol. 1995; 15: 6895-6900Crossref PubMed Scopus (60) Google Scholar). Most of them are structurally related to known components of the archaeal polymerase, indicating the existence of an extended core enzyme form that is common to the archaeal and eucaryotic lineages (5Langer D. Hain J. Thuriaux P. Zillig W. Proc. Natl. Acad. Sci. U. S. A. 1995; 92: 5768-5772Crossref PubMed Scopus (262) Google Scholar). The present work deals with ABC10β, a small polypeptide of 70 amino acids that is shared by all three yeast RNA polymerases, is able to bind Zn2+ in vitro (11Carles C. Treich I. Bouet F. Riva M. Sentenac A. J. Biol. Chem. 1991; 266: 24092-24096Abstract Full Text PDF PubMed Google Scholar) and is essential for growth (12Woychik N.A. Young R.A. J. Biol. Chem. 1990; 265: 17816-17819Abstract Full Text PDF PubMed Google Scholar). We have previously determined the amino acid sequence of that polypeptide (13Lalo D. Carles C. Sentenac A. Thuriaux P. Proc. Natl. Acad. Sci. U. S. A. 1993; 90: 5524-5528Crossref PubMed Scopus (109) Google Scholar), which has a close homology to the N subunit of the archaeal enzyme (5Langer D. Hain J. Thuriaux P. Zillig W. Proc. Natl. Acad. Sci. U. S. A. 1995; 92: 5768-5772Crossref PubMed Scopus (262) Google Scholar, 10McKune K. Moore P.A. Hull M.W. Woychik N.A. Mol. Cell. Biol. 1995; 15: 6895-6900Crossref PubMed Scopus (60) Google Scholar, 14Lanzendörfer M. Langer D. Hain J. Klenk H.P. Holz I. Arnold-Ammer I. Zillig W. Syst. Appl. Microbiol. 1994; 16: 656-664Crossref Scopus (30) Google Scholar) and is remotely related to the smallest subunit of vaccinial RNA polymerase (15Amegadzie B.Y. Ahn B.Y. Moss B. J. Virol. 1992; 66: 3003-3010Crossref PubMed Google Scholar). We report here that the yeast subunit and its archaeal homolog are largely interchangeable in vivo and that the eucaryotic, archaeal, and viral polypeptides have in common an invariant metal binding domain (CX 2C … CC) critical for its biological activity in S. cerevisiae. Moreover, our data show that ABC10β discriminates between RNA polymerase I and the other two nuclear RNA polymerases via an invariant eucaryotic motif, HVDLIEK. Replacing this motif by its archaeal counterpart strongly interferes with the heteromultimeric assembly of yeast RNA polymerase I. Newly constructed strains and plasmids are listed in TableI. Plasmids pFL44L (16Bonneaud N. Ozier-Kalogeropoulos O. Li G. Labouesse M. Minvielle-Sebastia L. Lacroute F. Yeast. 1991; 7: 609-615Crossref PubMed Scopus (501) Google Scholar), pGEN and pGENSc-10β (9Shpakovski G.V. Acker J. Wintzerith M. Lacroix J.F. Thuriaux P. Vigneron M. Mol. Cell. Biol. 1995; 15: 4702-4710Crossref PubMed Scopus (116) Google Scholar), pFL44-RPB10e (13Lalo D. Carles C. Sentenac A. Thuriaux P. Proc. Natl. Acad. Sci. U. S. A. 1993; 90: 5524-5528Crossref PubMed Scopus (109) Google Scholar), pNOY102 (17Nogi Y. Yano R. Nomura M. Proc. Natl. Acad. Sci. U. S. A. 1991; 88: 3962-3966Crossref PubMed Scopus (142) Google Scholar), pASZ11 (18Stotz A. Linder P. Gene. 1990; 95: 91-98Crossref PubMed Scopus (197) Google Scholar) and YCPA43–12 (19Thuriaux P. Mariotte S. Buhler J.M. Sentenac A. Vu L. Lee B.S. Nomura M. J. Biol. Chem. 1995; 270: 24252-24257Abstract Full Text Full Text PDF PubMed Scopus (40) Google Scholar) were previously described. Plasmids bearing mutant or chimeric forms of RPB10 were expressed from the strong constitutive pPGK1 promoter of the TRP1 multicopy vector pGEN-B (Table I). The yeast (S. cerevisiae) and archaeal (Sulfolobus acidocaldarius) coding sequences were cloned between the EcoRI and XhoI sites of pGEN-B to generate pGEN-B/RPB10 and pGEN-B/RpoN. The latter plasmid contains a unique BamHI site in the RpoN coding sequence. The N-terminal and C-terminal halves of RPB10 andRpoN could be exchanged to form the F5 and F6 chimeric protein of Fig. 2 A. This required an internalBamHI site at the equivalent position of RPB10, made by a GAA → CAA change at the 32nd codon corresponding to the phenotypically silent rpb10-E32P allele of plasmid pGEN-B/RPB10(E32P). The remaining RPB10/RpoN chimerae (F2, F4, F7) were constructed by PCR 1The abbreviations used are:PCR, polymerase chain reaction; YPD, yeast-peptone-dextrose; pol, polymerase; HA, hemagglutinin. amplification with appropriate primers overlapping with either the EcoRI,BamHI, and XhoI site of pGEN-B/RPB10(E32P) and pGEN-B/RpoN. Plasmids bearing single-site mutations were generated by PCR-mediated mutagenesis of rpb10 (site-directed or random mutagenesis) on pGEN-Sc10β and pGVS102 (Table I). N- and C-terminal hemagglutinin-tagged forms of ABC10β were obtained by PCR amplification of RPB10 using primers bearing a single copy of the coding sequence of the hemagglutinin epitope.Table IPlasmids and yeast strainsNameYeast genetic markers/GenotypeOriginPlasmidspGEN-BORI(2μm)TRP1Derivate of pGEN (9Shpakovski G.V. Acker J. Wintzerith M. Lacroix J.F. Thuriaux P. Vigneron M. Mol. Cell. Biol. 1995; 15: 4702-4710Crossref PubMed Scopus (116) Google Scholar) obtained by replacing theBamHI site by AGATCCpGEN-Sc10β(K59E)ORI(2μm)TRP1 rpb10-K59EPCR-mediated mutagenesis of pGEN-Sc10β (9Shpakovski G.V. Acker J. Wintzerith M. Lacroix J.F. Thuriaux P. Vigneron M. Mol. Cell. Biol. 1995; 15: 4702-4710Crossref PubMed Scopus (116) Google Scholar)pGEN-Sc10β(157F)ORI(2μm) TRP1 rpb10–157FPCR-mediated mutagenesis of pGEN-Sc10βpGEN-Sc10β(ΔLEK)ORI(2μm) TRP1 rpb10-ΔLEKPCR-mediated mutagenesis of pGEN-Sc10βpGEN-RPB10-FxORI(2μm) TRP1 rpb10A set of mutant rpb10 temperature-sensitive, cloned in pGEN-B and bearing yeast/archaeal chimeras (see "Materials and Methods.")pRPONRpoNRpoN coding sequence (S. acidocaldarius) cloned in pUC18pGEN-B/RPB10ORI(2μm)TRP1Directional cloning of a 210-base pairEcoRI-XhoI fragment from pGEN-Sc10β in pGEN-BpGEN-B/RpoNORI(2μm) TRP1Directional cloning of a 210-base pair EcoRI-XhoI fragment from pRPON in pGEN-BpGEN-B/RPB10(E32P)ORI(2μm)TRP1Directional cloning of a 210-base pair EcoRI-XhoI fragment from pGEN-Sc10β in pGEN-BpYGVS102ORI(2μm) URA3 RPB10pFL44-RPB10e (13Lalo D. Carles C. Sentenac A. Thuriaux P. Proc. Natl. Acad. Sci. U. S. A. 1993; 90: 5524-5528Crossref PubMed Scopus (109) Google Scholar) deleted of a PvuII fragmentpRPB10–5CEN6 ARSxADE2 RPB10Directional cloning of the 1.5-kilobaseBamHI-KpnI fragment of pFL44-RPB10e in pASZ11 (18Stotz A. Linder P. Gene. 1990; 95: 91-98Crossref PubMed Scopus (197) Google Scholar)pRPB10-ADE3ORI(2μm) ADE3 URA3 RPB10Directional cloning of theBamHI-SalI fragment (bearing ADE3) in pGVS102pASZ11-RPA43CEN6 ARSx ADE2 RPA43Directional cloning of the 2.5-kilobaseBamHI-PstI genomic fragment bearingRPA43 (19Thuriaux P. Mariotte S. Buhler J.M. Sentenac A. Vu L. Lee B.S. Nomura M. J. Biol. Chem. 1995; 270: 24252-24257Abstract Full Text Full Text PDF PubMed Scopus (40) Google Scholar) in pASZ11 (18Stotz A. Linder P. Gene. 1990; 95: 91-98Crossref PubMed Scopus (197) Google Scholar)pFKB181ORI(2μm)URA3 SRP40Directional cloning of a 2.0-kilobaseClaI-NruI fragment (bearing SRP40) in pFL44L (30Ikonomova R. Sommer T. Képès F. DNA Cell Biol. 1997; 16: 1161-1173Crossref PubMed Scopus (7) Google Scholar)pGEN-RPB10WTHAORI(2μm) URA3 RPB10-HAPCR-mediated mutagenesis of pGEN-Sc10β to fuse the coding sequence with HA tagpGEN-RPB10F4HAORI(2μm)URA3 rpb10-F4HAPCR-mediated mutagenesis of pGEN-RPB10-F4 to fuse the coding sequence with HA tagYeast strainsYGVS017MATα rpb10-Δ::HIS3 leu2–3,112 ade2–1 ura3–52 lys2-Δ201 his3-Δ200 trp1-Δ1 (pRPB10–5)Offspring of YGVS013 (9Shpakovski G.V. Acker J. Wintzerith M. Lacroix J.F. Thuriaux P. Vigneron M. Mol. Cell. Biol. 1995; 15: 4702-4710Crossref PubMed Scopus (116) Google Scholar) × YPH499 (these strains are isogenic to the YPH500 strain stock Ref. 41Woychik N.A. McKune K. Lane W.S. Young R.A. Gene Expr. 1993; 3: 77-82PubMed Google Scholar)OG20MATα rpb10-Δ::HIS3 leu2–3,112 ade2–1 ura3–52 lys2-Δ201 his3-Δ200 trp1-Δ1 (pGEN-Sc10β(K59E))Offspring of YGVS013 (9Shpakovski G.V. Acker J. Wintzerith M. Lacroix J.F. Thuriaux P. Vigneron M. Mol. Cell. Biol. 1995; 15: 4702-4710Crossref PubMed Scopus (116) Google Scholar) × YPH499OG23MATα rpb10-Δ::HIS3 leu2–3,112 ade2–1 ura3–52 lys2-Δ201 his3-Δ200 trp1-Δ1 (pGEN-Sc10β(I57F))Offspring of YGVS013 (9Shpakovski G.V. Acker J. Wintzerith M. Lacroix J.F. Thuriaux P. Vigneron M. Mol. Cell. Biol. 1995; 15: 4702-4710Crossref PubMed Scopus (116) Google Scholar) × YPH499Y260MATa ura3–52 rpb1–126GF312–17cMATα ura3–52 leu2–3,112 trp1–289 lys2 ade2-Δ ade3-ΔG. Faye, unpublished dataD115–17cMATa ura3–52 leu2–3,112 trp1 his3-Δ200 lys2 ade2-Δ ade3-ΔOffspring of GF312–17c × YNN281OG10–6aMATa rpb10-Δ::HIS3 leu2–3,112 ade3Δ ade2 ura3–52 lys2 his3-Δ200 trp1(pRPB10-ADE3)Offspring of D115–17c × YGVS25 (9Shpakovski G.V. Acker J. Wintzerith M. Lacroix J.F. Thuriaux P. Vigneron M. Mol. Cell. Biol. 1995; 15: 4702-4710Crossref PubMed Scopus (116) Google Scholar)OG22–2bMATa rpb10-Δ::HIS3 rpa43-Δ::LEU2 leu2–3,112 ade2–1 ura3–52 lys2-Δ201 his3-Δ200 trp1-Δ1 (pRPB10–5) (pNOY102)Offspring of D128–1d (19Thuriaux P. Mariotte S. Buhler J.M. Sentenac A. Vu L. Lee B.S. Nomura M. J. Biol. Chem. 1995; 270: 24252-24257Abstract Full Text Full Text PDF PubMed Scopus (40) Google Scholar) × YGVS017OG22–2b-WTHAMATa rpb10-Δ::HIS3 rpa43-Δ::LEU2 leu2–3,112 ade2–1 ura3–52 lys2-Δ201 his3-Δ200 trp1-Δ1Offspring of D128–1d (19Thuriaux P. Mariotte S. Buhler J.M. Sentenac A. Vu L. Lee B.S. Nomura M. J. Biol. Chem. 1995; 270: 24252-24257Abstract Full Text Full Text PDF PubMed Scopus (40) Google Scholar) × YGVS017(pNOY102) (pASZ11-RPA43) (pGEN-RPB10-WTHA)OG22–2b-F4HAMATa rpb10-Δ::HIS3 rpa43-Δ::LEU2 leu2–3,112 ade2–1 ura3–52 lys2-Δ201 his3-Δ200 trp1-Δ1Offspring of D128–1d (19Thuriaux P. Mariotte S. Buhler J.M. Sentenac A. Vu L. Lee B.S. Nomura M. J. Biol. Chem. 1995; 270: 24252-24257Abstract Full Text Full Text PDF PubMed Scopus (40) Google Scholar) × YGVS017(pNOY102) (pASZ11-RPA43) (pGEN-RPB10-F4HA) Open table in a new tab Mutants and chimeric forms of rpb10 were tested for their ability to complement the null allelerpb10-Δ::HIS3 of strains YGVS017 or OG10–6a (Table I) on the rich medium YPD at 16, 30, and 37 °C, using a plasmid shuffle assay (9Shpakovski G.V. Acker J. Wintzerith M. Lacroix J.F. Thuriaux P. Vigneron M. Mol. Cell. Biol. 1995; 15: 4702-4710Crossref PubMed Scopus (116) Google Scholar). In vivo complementation of the null mutant was monitored by the formation of uracil auxotrophic subclones (selected in the presence of 5-fluoroorotic acid) in the case of OG10–6a or of red sectors on YPD in the case of YGVS017. Multicopy suppression was systematically monitored by testing several independent transformants by the relevant suppressor plasmid and by checking that spontaneous subclones lacking the suppressor plasmid (selected in the presence of 5-fluoroorotic acid) invariably lose the suppressor phenotype. [3H]Uracil labeling was done on strains made prototrophic for uracil by transformation with pFL44L. Cells were grown in casamino acids medium (30 °C) supplemented with adenine (20 μg/ml) to an optical density of 0.2 at 600 nm and shifted for 1, 4, or 8 h at 37 °C. 150 μCi of [5,6 3H]uracil (1mCi/ml) were added to 10 ml of culture for 10 min. RNA was prepared as described previously (20Hermann-Le Denmat S. Werner M. Sentenac A. Thuriaux P. Mol. Cell. Biol. 1994; 14: 2905-2913Crossref PubMed Scopus (88) Google Scholar). Total RNA were separated by electrophoresis on a 1.2% agarose gel after denaturation with glyoxal and dimethyl sulfoxide and transferred to a positively charged nylon membrane (Boehringer Mannheim) by vacuum blotting (2 h, 200 millibars) in a 785 Vacuum Blotter (Bio-Rad) in 10× SSC (1× SSC =0.15 m NaCl and 0.015 m sodium citrate). RNAs were cross-linked using a UV-Stratalinker apparatus (Stratagene) and hybridized againstDED1 and ACT1 probes that were PCR-amplified on yeast genomic DNA using TAACAACAACGGCGGCTACA and CCATCAAATCTCTGCCGTTG (DED1) or TTGAGAGTTGCCCCAGAAGAACACC and CACCATCACCGGAATCCAAAACAAT (ACT1) as primers. These probes were randomly labeled using a Megaprime DNA labeling kit (Amersham Pharmacia Biotech). Six-liter cultures of strains OG22–2b-WTHA (RPB10+) and F4HA (rpb10-F4) grown on minimal SD-Gal (supplemented with casamino extract but lacking adenine) were harvested at a absorbance of about 1.0. Cell-free extracts were prepared as previously described (21Riva M. Buhler J.M. Sentenac A. Fromageot P. Hawthorne D.C. J. Biol. Chem. 1982; 257: 4570-4577Abstract Full Text PDF PubMed Google Scholar) and cleared by centrifugation. 15 μg of the supernatant proteins was loaded on SDS-PAGE gels (11 and 8% polyacrylamide). These preparations were probed on nitrocellulose membranes by antibodies raised against purified pol I and by a mixture of antibodies raised against the two largest subunits of yeast RNA polymerase II, using a commercial detection system (Amersham Pharmacia Biotech). Fig.1 A presents the alignment of four eucaryotic, four archaeal, and three viral gene products related to ABC10β and documents the particularly strong conservation of the eucaryotic sequences with 41 identical amino acid positions in the fungal, plant, and human sequences. We have previously shown that the human and fission yeast polypeptides are functional in S. cerevisiae (9Shpakovski G.V. Acker J. Wintzerith M. Lacroix J.F. Thuriaux P. Vigneron M. Mol. Cell. Biol. 1995; 15: 4702-4710Crossref PubMed Scopus (116) Google Scholar, 10McKune K. Moore P.A. Hull M.W. Woychik N.A. Mol. Cell. Biol. 1995; 15: 6895-6900Crossref PubMed Scopus (60) Google Scholar, 22Shpakovski G.V. Lebedenko E.N. Thuriaux P. Bioorg. Khim. 1997; 23: 110-117PubMed Google Scholar, 23Mukerjee K. Chaterji D. Eur. J. Biochem. 1997; 247: 884-889Crossref PubMed Scopus (32) Google Scholar). The archaeal sequences (including the N polypeptide, a proven polymerase subunit of S. acidocaldarius (14Lanzendörfer M. Langer D. Hain J. Klenk H.P. Holz I. Arnold-Ammer I. Zillig W. Syst. Appl. Microbiol. 1994; 16: 656-664Crossref Scopus (30) Google Scholar)), are less stringently conserved but still clearly related to the eucaryotic subunit. Three cytoplasmic DNA viruses (vaccinia, African swine fever virus and chilo Iridovirus) encode products that are loosely related to each other and to ABC10β, but the vaccinial polypeptide copurifies with RNA polymerase and is thus in all likelihood a genuine subunit of that enzyme (15Amegadzie B.Y. Ahn B.Y. Moss B. J. Virol. 1992; 66: 3003-3010Crossref PubMed Google Scholar). In contrast, there is no ABC10β-related sequence in any of the bacterial genomes currently available in public data banks. In particular, ABC10β is unrelated to ω, a small polypeptide that copurifies with the holopolymerase of Escherichia coli ((23) and references therein). The N subunit of S. acidocaldarius and ofHaloarcula marismortui cannot replace ABC10β in vivo (9Shpakovski G.V. Acker J. Wintzerith M. Lacroix J.F. Thuriaux P. Vigneron M. Mol. Cell. Biol. 1995; 15: 4702-4710Crossref PubMed Scopus (116) Google Scholar, 24McKune K. Woychik N.A. J. Bacteriol. 1994; 176: 4754-4756Crossref PubMed Google Scholar). However, yeast/archaeal chimerae are largely interchangeable in vivo with the yeast subunit, because they support growth when introduced in a S. cerevisiae host strain that is deleted for RPB10, the gene encoding ABC10β (Fig. 2). The archaeal and eucaryotic polypeptides must therefore have largely equivalent functions in their respective transcription complexes. In particular, all constructions where the archaeal component extends up to the first 49 amino acids are viable. In contrast, a chimeric construction (rpb10-F8) extending the archaeal segment up to position His-53 was lethal, as was also the case for a rpb10-F4, swapping the C-end of ABC10β downstream His-53. Deleting the last five amino acids (downstream of Pro-65) had instead no growth defect (rpb10-ΔLEK). The amino acids falling between positions His-53 and Pro-65 are therefore likely to account for the functional incompatibility observed between full-size archaeal (S. acidocaldarius) and eucaryotic ABC10β in vivo. Interestingly, this domain is moderately conserved between archaeal and eucaryotic sequences but contains an invariant HVDLIEK motif in all eucaryotic sequences identified so far (Fig. 1 A). The S. acidocaldarius and S. cerevisiae subunits share 17 amino acids, of which 13 are identical in all eucaryotes and archaeal sequences (Fig.1 A). Amino acid replacements were generated at these 17 positions and at 11 other highly conserved amino acids, expressed from a strong promoter harbored on a multicopy plasmid (i.e.under conditions maximizing the expression of the mutant allele) and tested for their ability to support growth. Practically all the mutants tested were phenotypically silent or had but a partial growth defect, which in keeping with our domain swapping and heterospecific complementation data, indicates that the precise amino acid sequence of ABC10β is not very critical for its function (Fig. 1 B). However, conservative Cys to Ser or Thr substitutions at the CX 2C … CC motif are lethal at all temperatures tested, and mutants with partial growth defect also tend to cluster near this motif. This fits very well with the fact that the CX 2C … CC motif is the only sequence feature shared by all relatives of ABC10β, including the three viral gene products that hardly show any similarity outside this motif. In view of the striking similarity with the canonical CX 2C … CX 2C Zn2+ binding domain, and given that yeast ABC10β binds Zn2+ in vitro (25Treich I. Carles C. Riva M. Sentenac A. Gene Expr. 1992; 2: 31-37PubMed Google Scholar), we propose that the CX 2C … CC motif defines a new type of metal-coordinating domain, although Zn2+ need not be the true ligand chelated by that motif in vivo. As stated above, the growth properties of yeast/archaeal chimeric constructions suggest that the eucaryotic motif HVDLIEK is critical for the biological activity of ABC10β. This is confirmed by the properties of temperature-sensitive mutants isolated from a library of randomly mutagenizedRPB10 alleles. Only 2 of the 12 mutants thus obtained resulted from single amino acid substitutions (rpb10-K59E and rpb10-I57F), and both are in the HVDLIEK motif (rpb10-K59E actually replaces the invariant lysine by the glutamate of the archaeal sequence). These mutants are strongly temperature-sensitive on plates (see Fig. 4) but can still undergo two to three cycles of growth and division upon a shift to the restrictive temperature, as shown by their delayed growth arrest (Fig.3 A). We monitored the in vivo synthesis of rRNA and tRNA in the first 8 h after the shift (i.e. before growth recedes) by short-pulse labeling with tritiated uracil and followed the accumulation of specific mRNAs by Northern hybridization. The results are shown forrpb10-K59E, and similar data were obtained withrpb10-I57F (not shown). tRNA and 5 S rRNA synthesis (pol III) was not affected under these conditions, but there was a marked effect on the three rRNA species (25 S, 18 S, and 5.8 S) synthesized by pol I (Fig. 3B). In contrast to the rapid arrest observed for therpb1–1 mutant (26Nonet M. Scafe C. Sexton J. Young R. Mol. Cell. Biol. 1987; 7: 1602-1611Crossref PubMed Scopus (269) Google Scholar) specifically affecting the activity of pol II, mRNA synthesis goes on in the rpb10 mutants for at least 5 h after the temperature shift, although it progessively decreases thereafter (Fig. 3 C). Thus, both mutations primarily block the synthesis of pol I-dependent transcripts, although ABC10β is shared by all three polymerases.Figure 3Growth and transcription pattern in a rpb10-K59E mutant (OG20) and its isogenic wild type (WT) control (YGVS017). A, growth. Cell density was measured by turbidimetry. Double barsdenote parallel dilutions before overnight growth. B,RNA polymerase I- and III-dependent transcription. Cells were labeled for 10 min with 250 μCi of [3H]uracil (see "Material and Methods"). Lanes 1–4, strain OG20(rpb10-K59E). Lanes 5–8, wild type control YGVS017(RPB10 +). Cultures were shifted from 30 °C (lanes 1 and 5) to 37 °C for 1 h (lanes 2 and 6), 5 h (lanes 3 and7), or 8 h (lanes 4 and 8). Total RNA (5 mg as determined by absorbance at 260 nm) was separated on 6% acrylamide gel and exposed for 24 h to reveal the high molecular weight rRNA (25 S and 18 S). Low molecular weight RNAs (5.8 S rRNA, 5 S rRNA, and tRNAS) were revealed after a 7-day exposure. C, RNA polymerase II-dependent transcription. Total RNA was prepared from YGVS017(RPB10 +) and OG20(rpb10-K59E) and from Y260 (rpb1–1) (26Nonet M. Scafe C. Sexton J. Young R. Mol. Cell. Biol. 1987; 7: 1602-1611Crossref PubMed Scopus (269) Google Scholar). After separation on a 1% agarose gel, RNA was transferred to positively charged nylon membranes (Boehringer Mannheim) and hybridized with ACT1 and DED1 probes to determine the steady-state level of the corresponding mRNAs.View Large Image Figure ViewerDownload (PPT) To further explore the transcriptional specificity ofrpb10-K59E and rpb10-I57F mutants, we examined whether they could be rescued by making growth independent of pol I. The appropriate genetic context was created by introducing arpa43-Δ::LEU2 deletion that inactivates the essential pol I subunit A43 (19Thuriaux P. Mariotte S. Buhler J.M. Sentenac A. Vu L. Lee B.S. Nomura M. J. Biol. Chem. 1995; 270: 24252-24257Abstract Full Text Full Text PDF PubMed Scopus (40) Google Scholar) and is thus lethal, except in the presence of a plasmid that expresses the rDNA transcript from the galactose-inducible promoter pGAL7. This allows pol I-defective mutants to be rescued on galactose by pol II-dependent transcription (17Nogi Y. Yano R. Nomura M. Proc. Natl. Acad. Sci. U. S. A. 1991; 88: 3962-3966Crossref PubMed Scopus (142) Google Scholar, 19Thuriaux P. Mariotte S. Buhler J.M. Sentenac A. Vu L. Lee B.S. Nomura M. J. Biol. Chem. 1995; 270: 24252-24257Abstract Full Text Full Text PDF PubMed Scopus (40) Google Scholar). Growth at 37 °C was indeed restored inrpa43-Δ::LEU2 (pGAL7::rDNA) constructions harboring rpb10-K59E or rpb10-I57F, indicating that their temperature-sensitive phenotype is primarily because of a pol I defect. As a control, we reintroduced the wild typeRPA43 gene by genetic transformation and shifted the resulting transformants to a glucose medium, shutting off the RNA polymerase II-dependent synthesis of rRNA and thus restoring pol I dependence. Both mutants regained their temperature-sensitive phenotype under these conditions (Fig.4). This pol I-specific defect is an allele-specific property that was not observed in two other temperature-sensitive mutants (rpb10-H53Q andrpb10-M49L) nor in the rpc40-V78R mutant affecting AC40, the α-like subunit common to pol I and III. We next asked if the lethal phenotype of the rpb10-F4 andrpb10-F6 chimerae (where HVDLIEK is replaced by its archaeal counterpart) also results from a pol I-specific defect. Using the same experimental approach, we found again both chimeric constructions to be rescued on galactose, although they are fully lethal when pol II-dependent synthesis of rRNA is shut off in the presence of glucose (Fig. 5). However, growth was slower than in isogenic wild type controls, indicating that there is also a partial pol II and/or III defect. At this point, we began to wonder if ABC10β itself, although shared by all three polymerases, might be only essential for pol I-dependent transcription. Reassuringly, we found that the rpb10-Δ::HIS3deletion is lethal even when allowing pol II-dependent synthesis of rRNA (data not shown). Taking advantage of a previously described yeast genomic library constructed in the multicopy vector pFL44L (27Stettler S. Chiannilkulchai N. Hermann-Le Denmat S. Lalo D. Lacroute F. Sentenac A. Thuriaux P. Mol. Gen. Genet. 1993; 239: 169-176Crossref PubMed Scopus (80) Google Scholar), we looked for dosage-dependent suppressor genes able to partly correct the growth defect of the ts mutantsrpb10-I57F and rpb10-K59E at 37 °C. It has often been observed that conditional mutants in a given pol I, pol II, or pol III subunits are suppressed by increasing the gene dosage of another subunit of the same enzyme (13Lalo D. Carles C. Sentenac A. Thuriaux P. Proc. Natl. Acad. Sci. U. S. A. 1993; 90: 5524-5528Crossref PubMed Scopus (109) Google Scholar, 28Archambault J. Schappert K.T. Friesen J.D. Mol. Cell. Biol. 1990; 10: 6123-6131Crossref PubMed Scopus (45) Google Scholar, 29Thuriaux P. Werner M. Stettler S. Lalo D. Methods Mol. Genet. 1995; 6: 227-246Crossref Scopus (3) Google Scholar). None of the 13 other RNA polymerase I subunits had any suppressor effect, includingRPC19 and RPC40 (encoding the α-like subunits AC19 and AC40 of pol I and pol III), although ts alleles of these two genes are themselves suppressed by a high dosage of RPB10(13Lalo D. Carles C. Sentenac A. Thuriaux P. Proc. Natl. Acad. Sci. U. S. A. 1993; 90: 5524-5528Crossref PubMed Scopus (109) Google Scholar). However, an intriguing link between the α-like subunit AC40 and ABC10β is suggested by our observation that SRP40(encoding a putati
Referência(s)