Architecture of Ure2p Prion Filaments
2003; Elsevier BV; Volume: 278; Issue: 44 Linguagem: Inglês
10.1074/jbc.m306004200
ISSN1083-351X
AutoresUlrich Baxa, Kimberly L. Taylor, Joseph S. Wall, Martha N. Simon, Naiqian Cheng, Reed B. Wickner, Alasdair C. Steven,
Tópico(s)Trace Elements in Health
ResumoThe [URE3] prion is an inactive, self-propagating, filamentous form of the Ure2 protein, a regulator of nitrogen catabolism in yeast. The N-terminal "prion" domain of Ure2p determines its in vivo prion properties and in vitro amyloid-forming ability. Here we determined the overall structures of Ure2p filaments and related polymers of the prion domain fused to other globular proteins. Protease digestion of 25-nm diameter Ure2p filaments trimmed them to 4-nm filaments, which mass spectrometry showed to be composed of prion domain fragments, primarily residues ∼1–70. Fusion protein filaments with diameters of 14–25 nm were also reduced to 4-nm filaments by proteolysis. The prion domain transforms from the most to the least protease-sensitive part upon filament formation in each case, implying that it undergoes a conformational change. Intact filaments imaged by cryo-electron microscopy or after vanadate staining by scanning transmission electron microscopy (STEM) revealed a central 4-nm core with attached globular appendages. STEM mass per unit length measurements of unstained filaments yielded 1 monomer per 0.45 nm in each case. These observations strongly support a unifying model whereby subunits in Ure2p filaments, as well as in fusion protein filaments, are connected by interactions between their prion domains, which form a 4-nm amyloid filament backbone, surrounded by the corresponding C-terminal moieties. The [URE3] prion is an inactive, self-propagating, filamentous form of the Ure2 protein, a regulator of nitrogen catabolism in yeast. The N-terminal "prion" domain of Ure2p determines its in vivo prion properties and in vitro amyloid-forming ability. Here we determined the overall structures of Ure2p filaments and related polymers of the prion domain fused to other globular proteins. Protease digestion of 25-nm diameter Ure2p filaments trimmed them to 4-nm filaments, which mass spectrometry showed to be composed of prion domain fragments, primarily residues ∼1–70. Fusion protein filaments with diameters of 14–25 nm were also reduced to 4-nm filaments by proteolysis. The prion domain transforms from the most to the least protease-sensitive part upon filament formation in each case, implying that it undergoes a conformational change. Intact filaments imaged by cryo-electron microscopy or after vanadate staining by scanning transmission electron microscopy (STEM) revealed a central 4-nm core with attached globular appendages. STEM mass per unit length measurements of unstained filaments yielded 1 monomer per 0.45 nm in each case. These observations strongly support a unifying model whereby subunits in Ure2p filaments, as well as in fusion protein filaments, are connected by interactions between their prion domains, which form a 4-nm amyloid filament backbone, surrounded by the corresponding C-terminal moieties. Ure2p is a cytoplasmic homodimeric protein (2 × 40 kDa) (1Taylor K.L. Cheng N. Williams R.W. Steven A.C. Wickner R.B. Science. 1999; 283: 1339-1343Crossref PubMed Scopus (264) Google Scholar, 2Perrett S. Freeman S.J. Butler P.J. Fersht A.R. J. Mol. Biol. 1999; 290: 331-345Crossref PubMed Scopus (79) Google Scholar) involved in regulation of nitrogen catabolism in Saccharomyces cerevisiae (3Drillien R. Aigle M. Lacroute F. Biochem. Biophys. Res. Commun. 1973; 53: 367-372Crossref PubMed Scopus (61) Google Scholar, 4Coschigano P.W. Magasanik B. Mol. Cell. Biol. 1991; 11: 822-832Crossref PubMed Scopus (222) Google Scholar). An inactive aggregated prion form of Ure2p has been shown to provide the molecular explanation (5Wickner R.B. Science. 1994; 264: 566-569Crossref PubMed Scopus (1079) Google Scholar, 6Masison D.C. Wickner R.B. Science. 1995; 270: 93-95Crossref PubMed Scopus (330) Google Scholar, 7Edskes H.K. Gray V.T. Wickner R.B. Proc. Natl. Acad. Sci. U. S. A. 1999; 96: 1498-1503Crossref PubMed Scopus (173) Google Scholar) for the nonchromosomal genetic element [URE3] (8Lacroute F. J. Bacteriol. 1971; 106: 519-522Crossref PubMed Google Scholar). The term "prion" (infectious protein) was first introduced for the mammalian PrP protein (9Prusiner S.B. Science. 1982; 216: 136-144Crossref PubMed Scopus (4073) Google Scholar), but is now used more generally (5Wickner R.B. Science. 1994; 264: 566-569Crossref PubMed Scopus (1079) Google Scholar). Upon prion conversion, Ure2p is inactivated, apparently by entering an aggregated filamentous state (6Masison D.C. Wickner R.B. Science. 1995; 270: 93-95Crossref PubMed Scopus (330) Google Scholar, 7Edskes H.K. Gray V.T. Wickner R.B. Proc. Natl. Acad. Sci. U. S. A. 1999; 96: 1498-1503Crossref PubMed Scopus (173) Google Scholar, 10Speransky V.V. Taylor K.L. Edskes H.K. Wickner R.B. Steven A.C. J. Cell Biol. 2001; 153: 1327-1336Crossref PubMed Scopus (74) Google Scholar). The filaments partition between daughter cells when an infected yeast cell divides, explaining the non-Mendelian genetics; filaments are introduced to a cell lacking [URE3] by cytoplasmic mixing when it fuses with a [URE3] cell, thus explaining the infectivity. Knowledge of the filament structure is essential for gaining further understanding of prionogenesis in this system. Ure2p consists of a C-terminal functional domain and an N-terminal prion domain (6Masison D.C. Wickner R.B. Science. 1995; 270: 93-95Crossref PubMed Scopus (330) Google Scholar), a pattern repeated in another yeast protein, Sup35p, which can become the [PSI] prion (11Liebman S.W. Derkatch I.L. J. Biol. Chem. 1999; 274: 1181-1184Abstract Full Text Full Text PDF PubMed Scopus (87) Google Scholar). The Ure2p functional domain binds to the transcription factor Gln3p, preventing its migration to the nucleus (12Blinder D. Coschigano P.W. Magasanik B. J. Bacteriol. 1996; 178: 4734-4736Crossref PubMed Google Scholar, 13Beck T. Hall M.N. Nature. 1999; 402: 689-692Crossref PubMed Scopus (793) Google Scholar, 14Kulkarni A.A. Abul-Hamd A.T. Rai R. El Berry H. Cooper T.G. J. Biol. Chem. 2001; 276: 32136-32144Abstract Full Text Full Text PDF PubMed Scopus (71) Google Scholar). This interaction is prevented in the prion state, probably by steric blocking of the binding site of Ure2p for Gln3p (15Baxa U. Speransky V. Steven A.C. Wickner R.B. Proc. Natl. Acad. Sci. U. S. A. 2002; 99: 5253-5260Crossref PubMed Scopus (148) Google Scholar). Residues 90–354 suffice to regulate Gln3p activity (16Maddelein M.L. Wickner R.B. Mol. Cell. Biol. 1999; 19: 4516-4524Crossref PubMed Scopus (86) Google Scholar), and residues 100–354 are conserved among other yeasts (17Edskes H.K. Wickner R.B. Proc. Natl. Acad. Sci. U. S. A. 2002; 99: 16384-16391Crossref PubMed Scopus (76) Google Scholar). This domain has a similar fold to glutathione S-transferase (GST) 1The abbreviations used are: GST, glutathione S-transferase; CTEM, conventional transmission electron microscopy; STEM, scanning transmission electron microscopy; GFP, green fluorescent protein; CA, carbonic anhydrase; GdnHCl, guanidine hydrochloride; Tricine, N-[2-hydroxy-1,1-bis(hydroxymethyl)ethyl]glycine; LC-MS, liquid chromatography-mass spectrometry; EM, electron microscopy.1The abbreviations used are: GST, glutathione S-transferase; CTEM, conventional transmission electron microscopy; STEM, scanning transmission electron microscopy; GFP, green fluorescent protein; CA, carbonic anhydrase; GdnHCl, guanidine hydrochloride; Tricine, N-[2-hydroxy-1,1-bis(hydroxymethyl)ethyl]glycine; LC-MS, liquid chromatography-mass spectrometry; EM, electron microscopy. (18Bousset L. Belrhali H. Janin J. Melki R. Morera S. Structure. 2001; 9: 39-46Abstract Full Text Full Text PDF PubMed Scopus (111) Google Scholar, 19Umland T.C. Taylor K.L. Rhee S. Wickner R.B. Davies D.R. Proc. Natl. Acad. Sci. U. S. A. 2001; 98: 1459-1464Crossref PubMed Scopus (95) Google Scholar), as predicted from sequence similarity (4Coschigano P.W. Magasanik B. Mol. Cell. Biol. 1991; 11: 822-832Crossref PubMed Scopus (222) Google Scholar). The prion domain is necessary for prion-based inactivation of Ure2p and induces the de novo appearance of [URE3] (6Masison D.C. Wickner R.B. Science. 1995; 270: 93-95Crossref PubMed Scopus (330) Google Scholar, 16Maddelein M.L. Wickner R.B. Mol. Cell. Biol. 1999; 19: 4516-4524Crossref PubMed Scopus (86) Google Scholar, 20Masison D.C. Maddelein M.L. Wickner R.B. Proc. Natl. Acad. Sci. U. S. A. 1997; 94: 12503-12508Crossref PubMed Scopus (146) Google Scholar). This domain has an unusual amino acid composition, consisting of ∼45% Asn and Gln and ∼20% Ser and Thr. The Asn-rich tract extends to residue 89, but residues 1–65 suffice for prion-forming ability (although 1–80 are more efficient). In native Ure2p, the N-terminal region is highly protease-sensitive and thought to be unstable or even unfolded (2Perrett S. Freeman S.J. Butler P.J. Fersht A.R. J. Mol. Biol. 1999; 290: 331-345Crossref PubMed Scopus (79) Google Scholar, 21Thual C. Bousset L. Komar A.A. Walter S. Buchner J. Cullin C. Melki R. Biochemistry. 2001; 40: 1764-1773Crossref PubMed Scopus (75) Google Scholar). Ure2p forms amyloid filaments in vitro (1Taylor K.L. Cheng N. Williams R.W. Steven A.C. Wickner R.B. Science. 1999; 283: 1339-1343Crossref PubMed Scopus (264) Google Scholar), and several findings indicate that these filaments are similar to the aggregated prion form found in vivo: (i) The N-terminal domain is necessary and sufficient for prion behavior and aggregation in vivo (6Masison D.C. Wickner R.B. Science. 1995; 270: 93-95Crossref PubMed Scopus (330) Google Scholar, 7Edskes H.K. Gray V.T. Wickner R.B. Proc. Natl. Acad. Sci. U. S. A. 1999; 96: 1498-1503Crossref PubMed Scopus (173) Google Scholar, 20Masison D.C. Maddelein M.L. Wickner R.B. Proc. Natl. Acad. Sci. U. S. A. 1997; 94: 12503-12508Crossref PubMed Scopus (146) Google Scholar) and for filament formation in vitro (1Taylor K.L. Cheng N. Williams R.W. Steven A.C. Wickner R.B. Science. 1999; 283: 1339-1343Crossref PubMed Scopus (264) Google Scholar). (ii) Filaments of Ure2p could be detected in [URE3] cells in vivo but not in [ure-o] cells (10Speransky V.V. Taylor K.L. Edskes H.K. Wickner R.B. Steven A.C. J. Cell Biol. 2001; 153: 1327-1336Crossref PubMed Scopus (74) Google Scholar). These filaments are similar in thickness (∼25 nm) to those assembled in vitro. (iii) The proteinase K digestion pattern of Ure2p in extracts of [URE3] cells is very similar to that of in vitro formed filaments (1Taylor K.L. Cheng N. Williams R.W. Steven A.C. Wickner R.B. Science. 1999; 283: 1339-1343Crossref PubMed Scopus (264) Google Scholar, 6Masison D.C. Wickner R.B. Science. 1995; 270: 93-95Crossref PubMed Scopus (330) Google Scholar). Although considerable effort has been expended on examining Ure2p filaments, a consensus has not been reached as to their overall architecture and conformation (i.e. presence of amyloid 2Amyloid is defined as a filamentous, protease-resistant, conformational state of a protein that is rich in β-structure and binds the dye Congo Red, producing apple-green birefringence. The term "amyloid filament" tends to be used not only for filaments consisting entirely of amyloid but also for any filament that has a significant amyloid content.2Amyloid is defined as a filamentous, protease-resistant, conformational state of a protein that is rich in β-structure and binds the dye Congo Red, producing apple-green birefringence. The term "amyloid filament" tends to be used not only for filaments consisting entirely of amyloid but also for any filament that has a significant amyloid content.). We have proposed that the prion domains stack to form a core fiber of amyloid (1Taylor K.L. Cheng N. Williams R.W. Steven A.C. Wickner R.B. Science. 1999; 283: 1339-1343Crossref PubMed Scopus (264) Google Scholar, 10Speransky V.V. Taylor K.L. Edskes H.K. Wickner R.B. Steven A.C. J. Cell Biol. 2001; 153: 1327-1336Crossref PubMed Scopus (74) Google Scholar, 22Wickner R.B. Annu. Rev. Genet. 1996; 30: 109-139Crossref PubMed Scopus (67) Google Scholar), which is surrounded by the C-terminal domains. Backbones of polymerized prion domains have also been envisaged for Sup35p filaments (23Glover J.R. Kowal A.S. Schirmer E.C. Patino M.M. Liu J.J. Lindquist S. Cell. 1997; 89: 811-819Abstract Full Text Full Text PDF PubMed Scopus (543) Google Scholar) and HET-s filaments (24Balguerie A. Dos Reis S. Ritter C. Chaignepain S. Coulary-Salin B. Forge V. Bathany K. Lascu I. Schmitter J.M. Riek R. Saupe S.J. EMBO J. 2003; 22: 2071-2081Crossref PubMed Scopus (170) Google Scholar, 25Nazabal A. Dos Reis S. Bonneu M. Saupe S.J. Schmitter J.M. Biochemistry. 2003; 42: 8852-8861Crossref PubMed Scopus (41) Google Scholar). On the other hand, an alternative model has recently been proposed whereby Ure2p molecules are connected to each other by N-terminal to C-terminal interactions, and the filaments do not contain amyloid (26Bousset L. Thomson N.H. Radford S.E. Melki R. EMBO J. 2002; 21: 2903-2911Crossref PubMed Scopus (129) Google Scholar, 27Bousset L. Briki F. Doucet J. Melki R. J. Struct. Biol. 2003; 141: 132-142Crossref PubMed Scopus (68) Google Scholar). In this study, we have investigated the overall architecture of in vitro assembled Ure2p filaments and fusion protein filaments by a combination of biophysical and biochemical analyses. After controlled proteolytic digestion, the reaction products were analyzed by Western blotting, mass spectrometry, and conventional negative staining EM. We also visualized intact filaments by a high resolution negative staining technique in the scanning transmission electron microscope (STEM) (28Muller S.A. Engel A. Micron. 2001; 32: 21-31Crossref PubMed Scopus (51) Google Scholar) and by cryo-electron microscopy of vitrified specimens, and we measured their values of mass per unit length from dark-field STEM images of freeze-dried filaments (28Muller S.A. Engel A. Micron. 2001; 32: 21-31Crossref PubMed Scopus (51) Google Scholar, 29Thomas D. Schultz P. Steven A.C. Wall J.S. Biol. Cell. 1994; 80: 181-192Crossref PubMed Scopus (52) Google Scholar, 30Wall J.S. Hainfeld J.F. Simon M.N. Methods Cell Biol. 1998; 53: 139-164Crossref PubMed Google Scholar). The results of these experiments substantiate our amyloid core model of the filament structure and show that the prion domain residues, ∼1–70 form this core. They further indicate residues ∼71–90, although part of the prion domain by amino acid composition and prion-inducing ability are not in the central core but link it to C-terminal appendages. URE2 was cloned into a pFLAG vector (Kodak) by PCR using the primers 5′-ggaactcatatgcatcaccatcaccatcacatgtatccacgtgggaatatgatgaa-taacaacggc-3′ and 5′-ggaactgtcgacgaattctgtggttggggtaac-3′ and the NdeI and SalI restriction sites. To improve expression levels in Escherichia coli, the AGA codons 253 and 254 for arginine were replaced with CGT using site-directed PCR mutagenesis with the primers 5′-cggatgaggttcgtcgtgtttacggtgtag-3′ and 5′-ctacaccgtaaacacgacgaacctcatccg-3′ as described (31Komar A.A. Guillemet E. Reiss C. Cullin C. Biol. Chem. 1998; 379: 1295-1300PubMed Google Scholar), yielding plasmid pKT41-1. The N-terminal primer encodes a His6 tag and a thrombin cleavage site, and therefore pKT41-1 codes for a protein with the sequence: MH6MYPRGNUre2p1–354. Overexpression was performed using E. coli BL21 in LB medium containing 0.1 mg/ml ampicillin at 37 °C for 4 h after induction with 1 mm isopropyl β-d-thiogalactosidase at an A 550 of ∼1.0. After harvesting, the cells were resuspended in 50 mm sodium phosphate, 300 mm NaCl, pH 8.0, containing protease inhibitors (Complete EDTA-free, Roche Applied Science), and lysed by high pressure. Insoluble material was removed by centrifugation (40,000 × g, 1 h). The expressed protein was recovered in one step, using a nickel-nitrilotriacetic acid Superflow column (Qiagen). The protein was bound to the column in 50 mm sodium phosphate, 300 mm NaCl, pH 8.0, washed extensively with 50 mm sodium phosphate, 300 mm NaCl, 20 mm imidazole, pH 8.0, and eluted with 50 mm sodium phosphate, 300 mm NaCl, 250 mm imidazole, pH 8.0. This protein was either used directly in experiments (His6-Ure2p) or digested with thrombin (Roche Applied Science) at room temperature for 20 h to remove the His6 tag. The sequences of Ure2p1–65, Ure2p1–80, and Ure2p1–89 were cloned in the pFLAG vector (Kodak) using PCR and the NdeI and SalI restriction sites, producing plasmids pKT53, pKT54, and pKT55, respectively. For all, the N-terminal primer included a His6 tag, however, without the thrombin cleavage site. The sequences of the proteins were MH6-Ure2p1–65, MH6-Ure2p1–80, and MH6-Ure2p1–89. Synthesis and purification of Ure2p1–65 was as described previously (1Taylor K.L. Cheng N. Williams R.W. Steven A.C. Wickner R.B. Science. 1999; 283: 1339-1343Crossref PubMed Scopus (264) Google Scholar). Ure2p1–65-barnase, Ure2p1–65-CA, and Ure2p1–65-GST were prepared as described (15Baxa U. Speransky V. Steven A.C. Wickner R.B. Proc. Natl. Acad. Sci. U. S. A. 2002; 99: 5253-5260Crossref PubMed Scopus (148) Google Scholar). Plasmid pH469 coding for Ure2p1–80-SDDDDKGGRGFP-H6 was kindly supplied by H. Edskes. Ure2p1–65-GFP was cloned by digestion and religation of pH469 with NotI, exploiting the two NotI sites of pH469 after position 65 of Ure2p and in front of GFP, leading to pUB7. Expression from pUB7 yielded Ure2p1–65-GFP-H6. Expression and purification was performed as described for Ure2p1–80-GFP (15Baxa U. Speransky V. Steven A.C. Wickner R.B. Proc. Natl. Acad. Sci. U. S. A. 2002; 99: 5253-5260Crossref PubMed Scopus (148) Google Scholar). To keep numbering of amino acids in the different constructs clear, all proteins and fusion constructs were numbered with 1 being the first Met of Ure2p. Additional sequences at the N terminus (His6 tag and thrombin site) for some constructs were assigned negative numbers. All proteins were frozen in liquid nitrogen shortly after purification and stored at -80 °C to prevent filament formation. Filaments were assembled by incubating protein solutions (at 1–5 mg/ml) at 4 °C, usually in 20 mm Tris/HCl, 200 mm NaCl, pH 8.0. For Ure2p1–65-barnase the buffer was 20 mm Tris/HCl, 200 mm KCl, pH 7.5. For some fusion proteins, filament formation was accelerated by continual agitation of the solution. Filaments were purified by centrifugation (25,000 × g, 30 min) and washed at least two times with buffer before use. Rabbit antiserum Ure2-#2 has been described previously (1Taylor K.L. Cheng N. Williams R.W. Steven A.C. Wickner R.B. Science. 1999; 283: 1339-1343Crossref PubMed Scopus (264) Google Scholar, 5Wickner R.B. Science. 1994; 264: 566-569Crossref PubMed Scopus (1079) Google Scholar). Although raised against Ure2p, it is very specific for the N-terminal domain (approximately residues 1–80) (Supplementary Fig. S2). Rabbit antiserum Ure2–3C specific for the C-terminal domain of Ure2p (Ure2p64–354) was affinity-purified using recombinant Ure2p (10Speransky V.V. Taylor K.L. Edskes H.K. Wickner R.B. Steven A.C. J. Cell Biol. 2001; 153: 1327-1336Crossref PubMed Scopus (74) Google Scholar). To refine the specificity to Ure2p90–354, antibodies recognizing the region 64–89 were removed by passage over a 1.0-ml nickel-nitrilotriacetic acid Superflow column (Qiagen) previously loaded with His6-Ure2p1–89 in 6 m GdnHCl (to prevent filament formation) and then changed back to 20 mm Tris/HCl, 200 mm NaCl, pH 8.0. 0.5 ml of a protein solution at 4–8 mg/ml in 20 mm Tris/HCl, 200 mm NaCl, pH 8.0, was incubated with proteinase K (Promega) (0.1 mg/ml, 30 units/mg) for 16–20 h at 37 °C. The insoluble material was collected by centrifugation at 25,000 × g for 30 min at 4 °C and washed by resuspension in 20 mm Tris/HCl, 200 mm NaCl, pH 8.0, followed by centrifugation as above, three times. The pellet was then dissolved in either 6 m GdnHCl for mass spectrometry or in 10 m urea, 2% SDS with 5 min boiling for SDS-PAGE. For less rigorous digestion, the same protein sample was digested with 10 μg/ml proteinase K and incubated at 37 °C. Small aliquots were taken at various times and analyzed directly on SDS-PAGE gels. Samples were dissolved in 10 m urea, 2% SDS, and the urea concentration was kept as high as possible throughout the subsequent procedures. SDS-PAGE was performed on 10–20% gradient gels (ICN Biomedicals) with Tris-Tricine buffer. Gels were stained with Colloidal Coomassie (ICN Biomedicals) or used in Western blots to transfer proteins to a polyvinylidene difluoride membrane (Millipore). The membrane was treated with 0.2 m NaOH for 30 min before being treated with the standard procedure. This step proved to be necessary for reproducible staining of the proteinase K digests on Western blots (32Schlumpberger M. Wille H. Baldwin M.A. Butler D.A. Herskowitz I. Prusiner S.B. Protein Sci. 2000; 9: 440-451Crossref PubMed Scopus (55) Google Scholar). LC-MS was performed on an HP1100 LC-MSD system (Agilent Technologies, Palo Alto, CA), consisting of a binary pump, a degasser, an autosampler, an absorbance detector, an HP1100 MSD, and a Chem-Station for data acquisition and processing. Mass spectra were acquired in positive-ion mode, scanning from 300 to 1700 m/z every 4 s. A Zorbax SB-C3 reversed phase column (150 mm × 2.1 mm) equipped with a guard column was used for chromatographic separation. The flow rate was 0.2 ml/min. Separation was done with a gradient of 5% acetic acid in water to acetonitrile at 40 °C. Proteins and peptides were detected using absorption at 280 nm and MS signal in the positive mode. However, peptides from the N-terminal domain of Ure2p did not show a signal at 280 nm, because they contain neither tryptophan nor tyrosine. Apomyoglobin was used as a calibration standard. CTEM—For negative staining, samples were adsorbed onto freshly glow-discharged carbon-coated grids, rinsed with water, and stained with 1% uranyl acetate. Micrographs were recorded on a Zeiss EM 902 (Leica, Deerfield, IL). For cryo-EM, drops were adsorbed to holey carbon films, blotted, vitrified, and imaged on a CM200-FEG microscope (FEI, Mahwah, NJ) (33Cheng N. Conway J.F. Watts N.R. Hainfeld J.F. Joshi V. Powell R.D. Stahl S.J. Wingfield P.E. Steven A.C. J. Struct. Biol. 1999; 127: 169-176Crossref PubMed Scopus (50) Google Scholar). STEM—Dark-field micrographs of freeze-dried specimen were recorded at 1.0 or 2.0 nm/pixel at the Brookhaven STEM resource. Mass measurements were performed by interactive analysis of the micrographs using PIC programs (34Trus B.L. Kocsis E. Conway J.F. Steven A.C. J. Struct. Biol. 1996; 116: 61-67Crossref PubMed Scopus (43) Google Scholar). Tobacco mosaic virus particles served as a mass standard (131.4 kDa/nm) (35Wall J.S. Hainfeld J.F. Annu. Rev. Biophys. Biophys. Chem. 1986; 15: 355-376Crossref PubMed Scopus (189) Google Scholar). Some samples were stained with 2% methylamine vanadate (Nanoprobes, Stony Brook, NY). Filament Formation Requires a Covalently Attached Prion Domain—Ure2p, Ure2p1–89, and the fusion proteins Ure2p1–65-barnase, Ure2p1–65-carbonic anhydrase (CA), Ure2p1–65-glutathione S-transferase (GST), Ure2p1–65-green fluorescent protein (GFP), and Ure2p1–80-GFP were expressed in E. coli and purified. Under the same conditions of incubation, all constructs formed filaments as detected by negative staining EM (Fig. 1). The filaments were either single (Type A) or paired (Type B) (15Baxa U. Speransky V. Steven A.C. Wickner R.B. Proc. Natl. Acad. Sci. U. S. A. 2002; 99: 5253-5260Crossref PubMed Scopus (148) Google Scholar). The width of single filaments increased with subunit molecular mass from ∼14 nm for Ure2p1–65-barnase (19.6 kDa) to ∼25 nm for Ure2p (41.8 kDa) (Fig. 1, bottom right panel). In no case did the C-terminal moiety, i.e. the appended protein without the Ure2p prion domain, form filaments under these conditions. It has been shown that the prion domain Ure2p1–65 promotes the polymerization of full-length Ure2p to form cofilaments (1Taylor K.L. Cheng N. Williams R.W. Steven A.C. Wickner R.B. Science. 1999; 283: 1339-1343Crossref PubMed Scopus (264) Google Scholar). Accordingly, we tested whether it has the same effect on the C-terminal domain alone. This experiment was performed with several prion domain constructs (Ure2p1–65, Ure2p1–80, and Ure2p1–89) mixed with C-terminal domain constructs Ure2p81–354 or Ure2p65–354, or with Ure2p as a positive control. By SDS-PAGE, Ure2p was detected in the insoluble fraction with each prion domain construct, and filaments were observed by negative staining. However, neither Ure2p65–354 nor Ure2p81–354 was detected in the insoluble fraction for any prion domain construct. These data show that a covalently attached prion domain is needed for the protein to enter filaments. Filaments of Ure2p and Fusion Constructs All Have Protease-resistant Core Fibers—All filament preparations were subjected to extensive digestion with proteinase K (100 μg/ml at 37 °C for 16 h). In each case, the reaction left sedimentable material consisting of much thinner filaments (Fig. 1). These filaments had the same diameter (4 nm) and tendency to aggregate into bundles (Fig. 1), whether they were derived from Ure2p or any of the fusion protein filaments; in these properties, they resemble polymers of Ure2p1–65 (1Taylor K.L. Cheng N. Williams R.W. Steven A.C. Wickner R.B. Science. 1999; 283: 1339-1343Crossref PubMed Scopus (264) Google Scholar) and Ure2p1–89 (Fig. 1). Similarly, thin filaments were observed in earlier work after proteinase K digestion of cofilaments of Ure2p and Ure2p1–65 (1Taylor K.L. Cheng N. Williams R.W. Steven A.C. Wickner R.B. Science. 1999; 283: 1339-1343Crossref PubMed Scopus (264) Google Scholar), with the qualification that those digestions were less rigorous and some of the resulting filaments had patches of greater width between 4-nm segments. 3K. L. Taylor, N. Cheng, R. B. Wickner, and A. C. Steven, unpublished results. Congo Red staining of the digested material produced apple-green birefringence, indicating that they are amyloid (Supplementary Fig. S1). Again, this is a property that they share with Ure2p1–65 filaments (1Taylor K.L. Cheng N. Williams R.W. Steven A.C. Wickner R.B. Science. 1999; 283: 1339-1343Crossref PubMed Scopus (264) Google Scholar, 32Schlumpberger M. Wille H. Baldwin M.A. Butler D.A. Herskowitz I. Prusiner S.B. Protein Sci. 2000; 9: 440-451Crossref PubMed Scopus (55) Google Scholar). The Protease-resistant Core Fibers Consist of N-terminal Peptides—To determine which fragments of Ure2p make up the thin filaments, this material was analyzed by SDS-PAGE (Fig. 2). With Coomassie Blue staining, all constructs yielded a fuzzy band at ∼10 kDa (Fig. 2). The specific affinity of these bands for the dye is about 2-fold lower than that of intact Ure2p, as determined by quantitative gel analysis. All of these bands reacted positively with a polyclonal antiserum that specifically recognizes Ure2p1–80; conversely, an antibody specific for Ure2p90–354 did not detect this material (Fig. 2). In contrast, digestions of soluble Ure2p with proteinase K for 16 h produced very little insoluble material, and no filaments were detected by EM. Moreover, no band was visible on Coomassie-stained gels or Western blots using either the C- or N-terminal antibody (data not shown). If the proteinase K digestion was less rigorous (10 μg/ml at 37 °C for up to 1 h) and soluble material was not separated from insoluble material, then Ure2p filaments yielded similar digestion patterns to those obtained from soluble Ure2p or C-terminal domain (Ure2p65–354) (Fig. 3). The main bands revealed by Coomassie staining correspond to fragments of the C-terminal domain, as identified by Western blots (Fig. 3). The N-terminal fragments are inconspicuous, because fragments of the C-terminal domain run in the same region of the gel (7–10 kDa) and stain more strongly with Coomassie Blue. We further characterized these earlier stages of proteinase K digestion of Ure2p filaments by mass spectrometry and gel analysis (Fig. 3). The first cut is between the N-terminal and C-terminal domains around residue 95 and produces a C-terminal fragment of ∼30 kDa. This fragment is not stable, however, and a second cut follows around residue 285 in the so-called cap region, a flexible loop that is unique to Ure2p among GST-like proteins (18Bousset L. Belrhali H. Janin J. Melki R. Morera S. Structure. 2001; 9: 39-46Abstract Full Text Full Text PDF PubMed Scopus (111) Google Scholar), producing relatively stable fragments of ∼20 and ∼8 kDa (Fig. 3). These fragments are subject to fraying at the ends as detected by mass spectrometry and by slight shifts to lower masses on SDS gels. Fraying is observed from residue 95 up to 120 and in the cap region from residue 285 down to 272 and up to 294. Further cleavage then takes place at sites that appear to be protected initially but become exposed as a result of prior cuts and fraying: cutting around residue 150 generates another relatively stable fragment at ∼14 kDa and cutting around residue 230 generates a fragment at ∼9 kDa. When fusion protein filaments were subjected to similar, less rigorous, proteinase K digestion (10 μg/ml at 37 °C for up to 1 h) without separation of insoluble material, each construct gave a characteristic pattern on SDS gels with Coomassie staining, indistinguishable from that obtained from unpolymerized protein (data not shown). Again, because of the weak staining of Ure2p N-domain, these patterns were dominated by fragments of the C-terminal moieties, in this case the appended proteins. GST and CA were digested slightly more slowly than Ure2p but were completely degraded after 16 h. However, GFP and barnase were rather resistant to proteinase K. Most of the GFP was still native and functional but detached from the filaments and found in the soluble fraction (Fig. 2, bottom panel). Some barnase was also found in native and active form in the post-digestion supernatant. These observations indicate that
Referência(s)