The Composition, Function and Evolution of the tRNA Splicing Endonuclease Citation Trotta, Christopher Robert (1999) The Composition, Function and Evolution of the tRNA Splicing Endonuclease. Dissertation (Ph.D.), California Institute of Technology. doi:10.7907/n4y9-jj84. https://resolver.caltech.edu/CaltechTHESIS:08192025-203726793 Abstract The splicing of tRNA precursors is essential for the production of mature tRNA in organisms from all major phyla. In Bacteria, intron removal is through autocatalysis of group I introns. In Archaea and Eukaryotes, intron removal is dependent on protein-based enzymes. Recognition and cleavage of the splice sites is accomplished by the tRNA splicing endonuclease. In order to understand how the eukaryotic endonuclease accomplishes this task, the genes encoding all four subunits of the S. cerevisiae enzyme have been cloned. All four genes are essential. Two subunits, Sen2 and Sen34, contain a homologous domain of approximately 130 amino acids. Surprisingly this domain is found in the gene encoding the archaeal tRNA splicing endonuclease of H. volcanii and in other Archaea. Utilizing the alignment as a guide, a mutation was made in the yeast Sen34 subunit which demonstrated that the eukaryotic endonuclease contains two functionally independent active sites for cleavage of the 5' and 3' splice sites, encoded by the SEN2 and SEN34 genes, respectively. The homology to the archaeal enzymes suggests an ancient origin for the tRNA splicing reaction. Exploiting the smaller and simpler archaeal version of the endonuclease, a crystal structure of the Methanoccocus jannaschii enzyme was determined to a resolution of 2.3 angstroms. The structure indicates that the cleavage reaction is similar to that of ribonucleaseA. Insight gained from the architecture of this homotetrameric enzyme has allowed for a clearer understanding of the heterotetrameric splicing endonuclease of yeast. In particular, the eukaryotic enzyme has preserved the important structural features found in the archaeal enzyme which allow for precise spatial positioning of the two active sites for cleavage. Item Type: Thesis (Dissertation (Ph.D.)) Subject Keywords: (Molecular Biology and Biochemistry) Degree Grantor: California Institute of Technology Division: Biology Major Option: Molecular Biology and Biochemistry Thesis Availability: Public (worldwide access) Research Advisor(s): Abelson, John N. Thesis Committee: Abelson, John N. (chair) Attardi, Giuseppe Campbell, Judith L. Dunphy, William G. Defense Date: 17 June 1998 Record Number: CaltechTHESIS:08192025-203726793 Persistent URL: https://resolver.caltech.edu/CaltechTHESIS:08192025-203726793 DOI: 10.7907/n4y9-jj84 Default Usage Policy: No commercial reproduction, distribution, display or performance rights in this work are provided. ID Code: 17635 Collection: CaltechTHESIS Deposited By: Benjamin Perez Deposited On: 21 Aug 2025 22:55 Last Modified: 21 Aug 2025 23:04 Thesis Files PDF - Final Version See Usage Policy. 50MB Repository Staff Only: item control page

The Composition, Function and Evolution of the tRNA Splicing Endonuclease Thesis by Christopher R. Trotta In Partial Fulfillment of the Requirements For the Degree of Doctor of Philosophy California Institute of Technology Pasadena, California 1999 (Defended June 17, 1998) ii © 1999 Christopher R. Trotta All Rights Reserved iii Acknowledgements Ultimate thanks are due Dr. John Abelson for creating an incredible atmosphere within his laboratory which has allowed me to learn how great science is done. John has been a constant source of academic inspiration and enlightenment and I am very honored to have been able to complete my degree under his guidance. Thanks also to members of the Abelson lab without which I would have surely faultered. Specifically I owe a debt of gratitude to Drs. Peggy Saks, Jaime Arenas, John Wagner, Scott Stevens and Jeff Sampson who provided me with sage scientific advice as well as incredible personal support. Thanks also to Drs., Dan Ryan, Randy Story, Tracy Johnson, Imre Barta, Andrey Balakin, and Tau-Mu Yi who have all contributed to my success in one way or another over the years. Huge thanks to Dr. Hong Li who has provided an extremely fruitful collaboration. And of course thanks are due to my fellow Entruvidorian counterparts, Chang Hee Kim for being there during my darkest hour and Dr. Eric Arn for showing me the path to good science and good "rock 'n roll." Thanks to DJ for being there during all of the best times and making the bad ones seem bearable. Special thanks to Jon, Chris, Steve, Alan, Bill, Rick, Patrick, Trevor, Billy, Peter, Geoff and Tony, for providing the ultimate inspiration, the "keys" to my soul. Finally thanks to my family for their constant and unyielding support from afar. iv Abstract The splicing of tRNA precursors is essential for the production of mature tRNA in organisms from all major phyla. In Bacteria, intron removal is through autocatalysis of group I introns. In Archaea and Eukaryotes, intron removal is dependent on proteinbased enzymes. Recognition and cleavage of the splice sites is accomplished by the tRNA splicing endonuclease. In order to understand how the eukaryotic endonuclease accomplishes this task, the genes encoding all four subunits of the S. cerevisiae enzyme have been cloned. All four genes are essential. Two subunits, Sen2 and Sen34, contain a homologous domain of approximately 130 amino acids. Surprisingly this domain is found in the gene encoding the archaeal tRNA splicing endonuclease of H. volcanii and in other Archaea. Utilizing the alignment as a guide, a mutation was made in the yeast Sen34 subunit which demonstrated that the eukaryotic endonuclease contains two functionally independent active sites for cleavage of the 5' and 3' splice sites, encoded by the SEN2 and SEN34 genes, respectively. The homology to the archaeal enzymes suggests an ancient origin for the tRNA splicing reaction. Exploiting the smaller and simpler archaeal version of the endonuclease, a crystal structure of the Methanoccocus jannaschii enzyme was determined to a resolution of 2.3 angstroms. The structure indicates that the cleavage reaction is similar to that of ribonucleaseA. Insight gained from the architecture of this homotetrameric enzyme has allowed for a clearer understanding of the heterotetrameric splicing endonuclease of yeast. In particular, the eukaryotic enzyme has preserved the important structural features found in the archaeal enzyme which allow for precise spatial positioning of the two active sites for cleavage. V Table of Contents Chapter I - Review: tRNA Splicing - An RNA World Add-On or an Ancient Reaction? Introduction Intron Containing tRNA The Enzymatic Gymnastics of tRNA Splicing Characterization of the tRNA Splicing Enzymes Recognition of pre-tRNA by the tRNA Splicing Endo nuclease The Genes for the Endonuclease: A Bridge Between Kingdoms tRNA Splicing at the Atomic Level On the Evolution of the Endonucleases Origin of tRNA Intrans Concluding Remarks References Chapter II - The Yeast tRNA Splicing Endonuclease: A Two Active Site Tetrarneric Enzyme Homologous to the Archaeal tRNA Splicing Endonucleases Abstract Introduction Results Discussion Experimental Procedures Acknowledgments References 1 1 1 8 12 17 21 21 30 36 38 38 44 44 44 50 77 81 88 89 Chapter III - Three-Dimensional Structure and Evolution of a tRNA Splicing Enzyme Abstract Introduction Structure Determination A Two Domain Monomer The Homotetrarner: A Dimer of Two Dimers Subunit Interfaces Active Sites Implications for the Structure of Other Splicing Endonucleases References 94 Chapter IV - Investigations into the Endonuclease Active Site and Architecture Introduction Dissection of the Interaction Domains of the Archaeal Endonuclease 136 94 94 99 99 105 108 110 124 133 136 136 vi The Endonuclease Active Site The Molecular Ruler Materials and Methods Appendix I Appendix II Appendix III 140 142 145 147 152 163 I Chapter I - Review: tRNA Splicing: An RNA World Add-On or an Ancient Reaction? (to be published in RNA World II eds. Tom Cech, Ray Gesteland and John Atkins. Cold Spring Harbor Laboratory Press, 1998; a shorter version of this chapter written as a minireview was published in JBC 273: 12685-12688, 1998; Appendix I.) INTRODUCTION Introns are encoded in the genes for tRNA in organisms from all three kingdoms of life. Their removal is an essential step in the maturation of tRNA precursors. In Bacteria, introns are self-splicing and are removed by a Group I splicing mechanism (Biniszkiewicz et al., 1994; Kuhsel et al., 1990; Reinhold-Hurek and Shub, 1992). In Eukaryotes and Archaea, intron removal is mediated enzymatically by proteins. Recent progress in understanding both eukaryotic and archaeal tRNA splicing has revealed that the two processes, previously thought to be unrelated, are in fact similar. Insight gained from the comparison has provided a clearer understanding of intron recognition, the catalysis of intron removal and has given new insight into the evolution of the tRNA splicing process. INTRON-CONTAINING tRNA tRNA Introns in Eukaryotes Intervening sequences (introns) were discovered 20 years ago in the yeast genes for the tyrosine-inserting nonsense suppressor tRNA (Goodman et al.,1977) and for tRNAPhe (Valenzuela et al., 1978). With the completion of the yeast genome, it is known that of the 274 yeast tRNA genes 61 or 20% contain introns. Table I-1 lists the tRNAs which contain introns. PCR cloning of tRNAs from higher eukaryotes has revealed a similar distribution of intron-containing tRNA (Green et al., 1990; Schneider et al., 1993) Stange and Beier, 1986). The introns in all of the genes are small (14-61 bases) and they 2 Yeast tRNA Precursors Containing Intrans lntron Length (nucleotides) No. of Genes Ser 19 1 Ser 19 4 Lys 23 7 Pro 31,30,33 7,2,1 Trp 34 6 Phe 18, 19 3,7 Leu 32,33 8,2 lie 60 2 Leu 19 3 Tyr 14 8 tRNA tRNA CGA tRNA GCU tRNA uuu tRNA UGG tRNA CCA tRNA GAA tRNA CAA tRNA UAU tRNA UAG tRNA GTA Total= 61 Total tRNA Genes= 274 Table 1-1 3 FIGURE I-1. Eukaryotic precursor tRNA A Consensus of yeast pre-tRNA. 0 = nonconserved base in a region of conserved length or secondary structure; X= nonconserved base in a region of variable length or secondary structure. G,C,A,U = conserved base among the pre-tRNAs. Y and R =conserved pyrimidines and purines, respectively. The conserved anticodon-intron (A-I) base-pair is depicted. B. Model of the tertiary structure of pre-tRNAs based on the crystal structure of yeast tRNAPhe (from Lee and Knapp, 1985). Intron is indicated in bold, dashed lines. 4 A A-OH C C 0-0 0-0 0-0 0-0 0-0 0-0 O-O OOOOG y OA Ru 1 1 1 1 1 R GG CXOG xOOOOCTuc I I I I X XX XGXOY XX xx oo-ox xx x 0-0 0-0 0-0 / 0-00 A-I Basepalr 3' Splice Site v-FP0 U X 0-0 0-0 0-0 S' Splice Site - : \ xxx Yeast Splicing Substrate Consensus Tertiary Structure of Precursor tRNA Figure I-1 5 are all located in the same position, one base to the 3' side of the anticodon (Figure 1-lA). Structure probing revealed that the common "cloverleaf' tertiary structure seen in the crystal structure of tRNAPhe is maintained in intron-containing tRNAs and that the intron, including the splice sites, is the most exposed region of the molecule (Lee and Knapp, 1985; Swerdlow and Guthrie, 1984). This led to a proposed model for the tertiary structure of pre-tRNA shown in Figure I-lB. More recent experiments by TocchiniValentini and co-workers (Baldi et al., 1992) have demonstrated that a conserved base pair (the A-I base pair) between a base of the 5' -exon (position 32) immediately following the anticodon stem and a base in the single-stranded loop of the intron (position -3) is required for correct excision at the 3' splice site (Figure 1-lA). Thus the distinctive feature in eukaryotic intron-containing pre-tRNAs is the position of the intron and the presence of the A-I basepair. tRNA Introns in Archaea tRNA introns are found in every archaeal species which has been studied (LykkeAnderson and Garrett, 1997; Klenk et al., 1998; Kleman-Leyer et al., 1997; Thompson et al., 1989). Though they are often similar, archaeal introns are different from their eukaryotic counterparts. Comparative sequence analysis of a host of archaeal introns has revealed that the 5' and 3' splice sites are always located in a three nucleotide bulge separated by a four base-pair helix: the Bulge-Helix-Bulge (BHB) motif (Figure I-2) (Thompson et al., 1989). Intron length is variable from a short intron of 33 nucleotides found in the tRNA Met gene of Methanococcusjannaschii, to 105 nucleotides of the tRNATrp gene of Haloferax volcanii (Daniels et al., 1985), but the intron is required to 6 FIGURE I-2. Consensus archaeal substrate. Positions depicted as in Figure I-lA. The bulge-helix-bulge (BHB) motif is shaded in grey. 7 O-OH 0 0-0 0-0 G-C 3 Splice Site R-YA/ 1 R C-GA U-R C-G vf)-0 R /RR-Y 5 Splice Site R-Y 1 X X X X XX Archaeal Splicing Substrate Consensus Figure I-2 8 base-pair with the 5'exon to allow formation of the BHB motif. In general, the introns reside one base 3' to the anticodon, but introns have been found elsewhere in the tRNA molecule. One such example is that of the tRNAPro gene of Methanobacterium thermoautotrophicum where there are two introns in the anticodon stem that must be removed in an obligate order to produce the mature tRNA (Trotta, unpublished). Another repository for introns in Archaea is the rRNA genes. In fact, one of the earliest discovered archaeal introns, the 622 nucleotide intron of the 23S rRNA of Desulfurococcus mobilis, was shown through extensive biochemical characterization to be removed by the archaeal tRNA splicing system with dependence on the BHB motif (Kjems and Garrett, 1985; Kjems et al., 1989). THE ENZYMATIC GYMNASTICS OF tRNA SPLICING The Eukaryotic Pathway The accumulation of pre-tRNAs in the rnal-1 mutant (Hopper et al. 1978) provided a source of pre-tRNA for assay of tRNA splicing in vitro (Knapp et al. 1978). The system was employed to work out the pathway of tRNA splicing (Figure I-3) . The first step of the reaction is recognition and cleavage of the pre-tRNA substrate at the 5' and 3' splice sites by a tRNA splicing endonuclease. The cleavage by the endonuclease at the splice sites results in 2' ,3 ' cyclic phosphate and 5' -hydroxyl termini (Peebles et al. , 1983). The tRNA 5' - and 3'-exons are then the substrate for a series of reactions catalyzed by the tRNA splicing ligase (for a detailed review of the ligation mechanism see Arn and Abelson, 1998). Ligase is a multi-functional enzyme possessing phosphodiesterase, polynucleotide kinase, and RNA ligase activities. In the ligase reaction, phosphodiesterase opens the 2' ,3' cyclic phosphate to give a 2'-phosphate. Next the 5'-hydroxyl is phosphorylated by transfer of they-phosphate from an NTP cofactor. 9 F1GURE I-3 . The tRNA splicing pathway of yeast. Gene names are given in parentheses for the proteins of the pathway. CPDase, cyclic phosphdiesterase; ASTase, adenylyl synthetase. See text for details (Taken with permission from Abelson et al., 1998). IO pre-tRNA tRNA half-molecules Ligase (RLG1) CPDase Kinase +A,ppG Endonuclease (SEN54,SEN2 SEN34, SEN15) Ag:>p mature tRNA 2' Phosphotransferase (TPT1) • ~r~G ~ Ct'••, . L1gase Ligase ·7 ~ f C LigaseQA ~~' A-Iii \ . -CO2~~ NH, µ N H2 0o~.J" OH ~ • -CH2 OH OH ADP-ribose 1"-2" cyclic phosphate (Appr>P) OH OH NH, M," I ~-c•, oU) 0 OH OH NAO+ (AppN) Fi g ure I-3 :)f ASTase 11 Ligase is then adenylylated and the AMP moiety is transferred to the 5' -phosphate on the 3' -exon forming an activated phosphoanhydride. Finally ligase joins the two exons by catalyzing an attack of the 3' -hydroxyl on the activated donor phosphoanhydride to form a 3' ,5' phosphodiester, 2' -monophosphoester bond with the release of AMP. This results in a pre-tRNA molecule with a 2' -phosphate at the splice junction. Removal of the 2' phosphate is accomplished by the activity of a 2' -phosphotransferase that catalyzes the transfer of the phosphate to NAD, releasing nicotinamide and yielding ADP-ribose 1"-2" cyclic phosphate and mature tRNA (Appr>P; Figure I-3). The Archaeal Pathway Much less is known about the enzymes involved in the processing of archaeal pretRNA due to the difficulty in working with extracts from these organisms. As in the eukaryotic pathway, removal of introns requires the function of both an endonuclease and a ligase. The archaeal endonuclease recognizes substrates containing the consensus BHB motif and cleaves at the 5' and 3' splice sites to yield 2',3' cyclic phosphate and 5'hydroxyl termini (Thompson et al. , 1989). Thus the cleavage mechanism is identical to the eukaryotic mechanism providing the earliest clue to the relatedness of the two enzymes (see below). Upon cleavage by the endonuclease, the tRNA half molecules must be ligated to produce a mature tRNA, but the ligase activity has yet to be defined. It is clear that the yeast mechanism does not pertain since they-phosphate of ATP is not incorporated into spliced tRNA as seen in the yeast mechanism (Kjems and Garrett, 1985; Gomes and Gupta, 1997). 12 CHARACTERIZATION OF THE tRNA SPLICING ENZYMES The Eukaryotic tRNA Endonuclease (described in detail in Chapter II) The first step of tRNA splicing is the recognition and cleavage of the splice sites by the tRNA splicing endonuclease. Utilizing the in vitro system for tRNA splicing, the endonuclease was shown to be nuclear membrane-associated (Peebles et al., 1983). The detergent, Triton X-100, at 0.9% is required to release the activity from the membrane fraction. In a particularly difficult purification, the endonuclease was purified one-millionfold from a yeast nuclear membrane preparation to homogeneity and appeared to be a a~y trimer of 54, 44, and 34 kD subunits (Rauhut et al., 1990). Later a fourth subunit of 15kD was discovered (Trotta et al., 1997). Thus the enzyme is a a~yo heterotetramer. Cloning of the genes encoding each of these protein subunits was accomplished by both genetic and protein chemistry. Ho et al. (1990) screened a bank of temperature sensitive mutants and isolated a yeast mutant that accumulated 2/3 tRNA precursor molecules that were not cleaved at the 5' splice site. The gene, SEN2 (~plicing endonuclease (Winey and Culbertson, 1988), was cloned by complementation and antibodies to the product were shown to recognize the 44kD subunit of the purified splicing endonuclease. Development of a simpler affinity purification utilizing a tagged SEN2 gene allowed preparation of sufficient material to identify the other three gene products, SEN54, SEN34 and SEN15 (Trotta et al. , 1997). Two of the subunits, SEN2 and SEN34, contain a homologous domain of approximately 130 amino acids. SEN2 contains the only transmembrane sequence and presumably mediates membrane association of the enzyme. All four subunits contain a nuclear localization sequence. Two-hybrid interactions were detected between Sen54Sen2 and Sen34-Sen15. Since the enzyme is a stable tetramer, interaction must occur 13 between these two presumed heterodirners although they are not detected in the twohybrid assay (Trotta et al., 1997). The Eukaryotic tRNA Ligase The purification of the soluble tRNA ligase from yeast was simpler and quickly led to the isolation of a 95kD polypeptide and the cloning of its essential gene, RLG 1 (Phizicky et al., 1986; Phizicky et al., 1992). This single polypeptide carries out all three of the steps in the ligase mechanism (Figure I-3). Interestingly, partial proteolysis of the protein resulted in three fragments that were shown to contain distinct activities supporting a domain-like structure of the enzyme (Xu et al., 1990). The amino-terminal fragment is adenylylated at lysine 114 and sequence comparisons have shown this lysine residue to be equivalent to the active lysine of the T4 RNA ligase (Apostol et al. , 1991; Xu et al. , 1990). The carboxy-terrninal fragment contains the cyclic phosphodiesterase activity. The central domain then would be the likely locus of the kinase activity, however, an enzyme deleted in this domain paradoxically retains ligase activity. The enzyme has two nucleotide-binding sites, for GTP and for ATP (Belford et al., 1993). GTP is employed in the kinase step and ATP is used for the formation of the activated adenyly-RNA intermediate. The in vivo function of such a complex NTP requirement is unclear, but it has been suggested that such a requirement could couple the splicing reaction to transcription and/or translation (Belford et al. , 1993). The complexity of the domain structure of ligase raises interesting questions as to the origin of this protein. The yeast tRNA ligase is mechanistically and structurally related to phage T4 RNA ligase and polynucleotide kinase (Apostol et al., 1991). The phage enzymes together contain the three activities present in the tRNA ligase but in addition T4 polynucleotide kinase contains a phosphatase which removes the 2' -phosphate (Walker et 14 al., 1975). Thus T4 RNA ligase and T4 polynucleotide kinase can ligate pre-tRNA half molecules produced by endonuclease yielding a product that does not contain a 2'phosphate. The 2'-Phosphotransferase Early investigation into the activity responsible for 2' -phosphate removal at the splice junction demonstrated the requirement for two separate components present in yeast extracts (McCraith and Phizicky, 1990). The first component was purified and determined to be the cellular co-factor NAO+ (McCraith and Phizicky, 1991). As described earlier, NAO+ is the receptor in a reaction which involves the transfer of the 2' phosphate at the splice junction to the 2" position of the ribose of NAO+ yielding the high energy compound AOP-ribose 1"-2" cyclic phosphate (Appr>p) (Culver et al., 1993). The second component was purified and shown to be a 26.2kDa polypeptide encoded by the TPTI gene (Culver et al., 1997). Expressed in E. coli, the single polypeptide could catalyze the transfer of the 2'-phosphate to NAO+. The gene was shown to be essential in yeast and it was suggested that its essential role could be either removal of the 2' phosphate from all intron-containing tRNA molecules or generation of the novel molecule Appr>p. To further understand the role of each of these potential functions, a conditional lethal phosphotransferase mutant was generated (Spinelli et al., 1997). Yeast expressing this mutant accumulated tRNA molecules with a 2' -phosphate at the splice junction. Interestingly, these tRNA molecules were undermodified at positions near the splice junction residue while other modifications appeared to be normal. These results suggest that the removal of the 2' -phosphate is essential for correct modification of the residues near the splice junction and that tRNA containing the 2' -phosphate is not a substrate for 15 the modification enzyme and likely inactive in carrying out its function in translation. The results also suggest a temporal order for the maturation of pre-tRNA with splicing and 2' phosphate removal occurring before modification of certain positions in the tRNA molecule. The TPTl gene was also found to have homologs in other eukaryotes and surprisingly in E. coli and some Archaea (Culver et al., 1997; Trotta, 1997). The function of this protein in Bacteria and Archaea is unclear since a source of 2'-phosphates is not apparent (see the discussion on the archaeal ligase mechanism below). E. coli does not possess intron-containing tRNA or a eukaryotic-like RNA ligase. It will therefore be interesting to elucidate its role in bacteria hopefully providing a clue as to the origin of this unique processing event. The Archaeal tRNA Endonuclease The endonuclease of members of the Archaea proved to be just as difficult to purify as those of Eukaryotes. Daniels and co-workers chose the halophilic archaeon Haloferax volcanii to carry out the purification. The enzyme is present in low abundance and is extremely difficult to stabilize during the purification protocol. When finally purified the endonuclease of H. volcanii proved to be composed of a single 37kDa protein encoded by the EndA gene (Kleman-Leyer et al., 1997). Unlike the heterotetrameric yeast enzyme, the active enzyme is composed of a single subunit shown by gel filtration to behave as a homodimer in solution. The protein was shown to contain a 130 amino acid domain homologous to the two yeast subunits SEN2 and SEN34. This was the observation that unified the two lines of research and led to the progress described below. The Archaeal tRNA Ligase The identity of the archaeal tRNA ligase is at present unknown; however, with the complete genome sequence of four members of the Archaea, it is clear that there is no 16 homolog of the yeast tRNA splicing ligase (a remarkable statement to be able to make, now common in the genomic sequence era). Thus it appears that ligation of tRNA half molecules occurs via a different reaction mechanism to that of the eukaryotic ligation reaction. All sequenced archaeal genomes do, however, contain an open reading frame encoding a protein homologous to the 2' -5' ligase of E. coli characterized by E. Arn (Arn and Abelson, 1996). This protein functions in vitro as an RNA ligase joining tRNA half molecules to form a 2' -5' phosphodiester bond at the splice junction. The exact mechanism of this reaction is unknown but it does not require the addition of exogenous ATP. Thus it is tempting to speculate that the homolog found in Archaea may function to ligate the tRNA half-molecules of the splicing reaction. 2'-5' linkages have not been detected in tRNA but it is possible that the archaeal ligase has an altered specificity and catalyzes the formation of a 3' -5 ' linkage. Clearly, it will be interesting to further investigate the archaeal homologues of the 2'-5' ligase. tRNA Splicing in Higher Eukaryotes Like the yeast tRNA introns, the position of the introns in higher eukaryotes is conserved and the cleavage reaction operates in a manner identical to the yeast tRNA splicing endonuclease as exemplified by the Xenopus endonuclease. Ligation of the half molecules to yield mature tRNA has been shown to proceed through a mechanism similar to the yeast ligase (Konarska et al., 1981 ; Schwartz et al., 1983; Zillmann et al., 1991). Using HeLa cell extracts, Zilmann et al. (1991) have demonstrated the incorporation of exogenous y-phosphate from ATP upon ligation of yeast tRNAPhe, a hallmark of the yeast tRNA ligase reaction mechanism. Furthermore, they demonstrated the presence of a 2'phosphate at the splice junction. A 2' -phosphotransferase activity was detected in both 17 HeLa extracts (Zillmann et al., 1992) and is implicated in the dephosphorylation of ligated tRNA in rnicroinjected Xenopus oocytes (Culver et al. , 1993). However, early studies implicated a different ligase in tRNA splicing in HeLa extracts and Xenopus oocytes. It was demonstrated that the product of the tRNA splicing reactions catalyzed by these extracts did not contain a 2' -phosphate at the splice junction. This novel ligase uses the cyclic phosphate at the end of the 5' half of tRNA to generate the 3' ,5 ' -phosphodiester bond in the mature tRNA as there is no incorporation of exogenous phosphate at the splice junction (Filipowicz et al., 1983; Laski et al., 1983; Nishikura and De Robertis, 1981). Thus there appears to be redundant pathways for ligation of tRNA molecules in metazoans. To date, no homolog of the bacterial 2' -5' RNA ligase has been sequenced in higher eukaryotes. The further characterization of the alternate eukaryotic ligase activity could conceivably shed light on the nature of the archaeal ligase mechanism. RECOGNITION OF PRE-tRNA BY THE tRNA SPLICING ENDONUCLEASE Eukaryotic Endonuclease: the Ruler Mechanism Early speculation concerning the ability of the tRNA endonuclease to recognize and cleave tRNA substrates relied on comparison of the primary and tertiary sequences of the precursor molecules. As previously mentioned there is no sequence conservation around the splice sites and the precursors all fold into the same tertiary structure. The intrans are of variable sequence and can be altered by both addition and deletion of nucleotides as well as changes in the sequence with little effect on removal by endonuclease (Johnson et al., 1980; Reyes and Abelson, 1988; Strobel and Abelson, 1986). The exception to this generalization is the A-I base pair necessary for accurate cleavage at the 3' splice site. Mutations which abolish this base pair are not substrates for the splicing endonuclease (Baldi et al., 1992; Willis et al. , 1984). This led to the 18 hypothesis that recognition involved specific interactions with the conserved tertiary structure present in the mature domain of the pre-tRNA. Experimental support for this hypothesis involved the engineering of mutant pre-tRNA which altered distinct features of the primary, secondary or tertiary structures of the molecule (reviewed in Culbertson and Winey, 1989). This suggested a ruler model for recognition and cleavage by the endonuclease in which the enzyme recognizes the mature domain and measures the conserved distance to the splice sites (Greer et al., 1987; Reyes and Abelson, 1988) (Figure I-4) . To test this model, tRNA molecules were constructed in which the conserved distance was altered by insertion and deletion of base pairs in the anticodon stem. Addition of a single base pair increased the length of the intron by 2 nucleotides , one at either end (Reyes and Abelson, 1988) as predicted by the model. Archaeal Endonuclease: the Bulge-Helix-Bulge Motif In contrast to the yeast results, it was demonstrated that for the archaeal endonuclease, the pre-tRNA substrate could be drastically altered and cleavage would still occur. Synthetic tRNA molecules where almost the entire tRNA mature domain was removed were accurately cleaved by the endonuclease (Thompson and Daniels, 1988). As in the case of the eukaryotic endonuclease, the intron plays a mostly passive role as large deletions can be tolerated by the enzyme. There is, however, a strict requirement for sequence and structure present at the exon-intron boundaries (Thompson and Daniels, 1988). This helix-bulge-helix (BHB) motif is the only requirement for intron recognition and removal by the archaeal tRNA splicing endonuclease. 19 FIGURE 1-4. Ruler model for interaction of the yeast endonuclease with pre-tRNA (see text; taken from Reyes and Abelson, 1988). 20 3'Splice Site Yeast Endonuclease Ruler Model Figure I-4 21 THE GENES FOR THE ENDONUCLEASE: A BRIDGE BETWEEN KINGDOMS (Discussed in detail in Chapter II) Once the genes for the eukaryotic and archaeal endonuclease were in hand, a comparison between the yeast subunits and the single H. volcanii endonuclease revealed that these proteins were related. A domain of 130 amino acid is found in the C-terrninus of the SEN2 and SEN34 genes of the yeast endonuclease and in the H. volcanii endonuclease. Furthermore, homo logs of this domain were discovered in the Archaea M. jannaschii and P. aerophilum where the domain represented the entirety of the protein (Figure I-5). This homology led to the suggestion that the two yeast subunits each contain an active site for the cleavage reaction. It had been demonstrated that a mutant in the SEN2 gene, sen2-3, was defective in 5' splice site cleavage (Ho et al., 1990). This mutation, glycine292 to glutamate, lies within the conserved domain suggesting that SEN2 contains the active site for 5' splice site cleavage and by extension that SEN34 carries the active site for 3' splice site cleavage. A mutant in SEN34 changing a conserved histidine at position 242 to alanine resulted in a marked decrease in cleavage at the 3' splice site, while cleavage at the 5' splice site was normal, strongly supporting the two active site model (Trotta et al., 1997). In the dimeric H. volcanii endonuclease, identical subunits cleave the symmetrically disposed splice sites in the BHB substrate. Thus the arrangement of active sites in eukaryotic and archaeal endonucleases is similar (as depicted in Figure I-6) . tRNA SPLICING AT THE ATOMIC LEVEL (Discussed in detail in Chapter ID) Given the relatedness of the tRNA splicing endonuclease from the two kingdoms, it seemed expedient to choose the archaeal system for structural work. The endonuclease of the archaeon M. jannnaschii was chosen as this enzyme consists of only the 22 FIGURE 1-5. Graphic representation of archaeal and eukaryotic endonuclease proteins. The sizes of the various proteins (open rectangle) and location of the conserved sequence regions (black box) are indicated. Light grey blocks represent a degenerative domain repeat present in several archaeal endonucleases (see text). Proteins are placed on the Woesnian evolutionary tree. Endonuclease configuration is noted (see text for details): a4=tetramer; a2=dimer; a~y8=heterotetramer; ?=configuration unknown. Protein designations: H.volcanii, H. volcanii EndA (AF001578), A.fulgidus, Archaeoglobus fulgidus (AE001041); M. jann, M. jannaschii (MJ1424); M. thermo, Methanobacterium thermoautotrophicum (AF001577); P. horik, Pyrococcus horikoshii (AB009474); Sc Sen2, Saccharomyces cerevisiae Sen2 (P16658); Sc Sen34, S. cerevisiae Sen34 (P39707); Sp Sen34, Schizosaccharomyces pombe Sen34; Z. mays, Zea mays open reading frame contained in the intron of HMG (X72692); Mouse, Drosophila, human, represent homo lo gs of the endonuclease domain which have been detected in partially sequenced genomic database submissions (Trotta, unpub.). 23 Endonuclease Gene Family I ARCHAEAI EURYARCHAEOT A Mouse • Drosophila• Human • •= IEUCARYA I ? Sc. SEN34 ~ Sc.SEN2 ~ Sp. SEN34 ai3y8 . mays ■ Figure 1-5 • • 24 FIGURE I-6. Comparison of the models proposed for the eukaryotic (yeast) and archaeal (H. volcanii) endonucleases (see text for details) (Taken with permission from Li et al., 1998). 25 SEN15 Eukaryotic Endonuclease Archaeal Endonuclease Figure 1-6 26 homologous domain implicated by the endonuclease sequence alignment. It is small and proved easy to express and purify. Extensive biochemical characterization of this enzyme by Lykke-Andersen and Garrett (1997) showed it to be a homotetramer in solution, an observation confirmed by the crystal structure (Li et al. , 1998)(Figure 1-7B). Each monomer consists of two domains: the N-terminal domain (residues 9-84) which is composed of three alpha helices and a mixed antiparallel/parallel P-pleated sheet of four strands and the C-terminal domain (residues 85-179) which contains two alpha helices flanking a five-stranded mixed B-sheet (Figure I-7A). The manner in which four of these monomers are brought together provides insight into the architecture and evolution of other members of the endonuclease family of both Archaea and eukaryotes. Two sets of interactions are crucial to formation of the tetramer. The first interaction involves the formation of an isologous dimer between two monomers. Formation of this dimer is mediated by the interaction of P9 from one monomer and P9 ' from another monomer. This tail-to-tail interaction results in main chain hydrogen bonds between monomers and leads to a two-stranded P-sheet spanning the subunit boundary. Loop L8 also interacts to form more hydrogen bonds and together these interactions enclose a hydrophobic core at the subunit interface. This leads to an extremely stable dimeric unit that Li et al. (1998) have suggested is conserved in endonucleases from other Archaea and eukaryotes (see below). The second interaction involves the heterologous interaction between the two dimers which compose the tetramer. The main interaction is an insertion of acidic residues in Loop LIO into a polar groove formed between the N and C terminal domains of the 27 FIGURE 1-7. Crystal structure of the M. jannaschii tRNA splicing endonuclease. A. Ribbon representation of the endonuclease monomer. The proposed catalytic triad residues are within 7 A of one another and are shown in red ball-and-stick. Also shown is electron density detected near the putative catalytic triad, where the scissile phosphate is proposed to be located for cleavage. B. Endonuclease tetramer. Each subunit is represented by a distinct label and color. The main chain hydrogen bonds formed between ~9 and ~8' and between loops L8 and L8' for the formation of isologous dimers are shown as thin lines. Side chains of the hydrophobic residues enclosed at the dimer interface are shown as blue ball-and-stick models. The heterologous interaction which forms the tetramer is mediated by interaction between subunits Al and B2 (or Bl and A2) through the acidic loops LIO and L8 and are highlighted by dotted surfaces (Taken with permission from Li et al. , 1998). 29 endonuclease monomer (Figure I-7B). This causes the two dimers to be translated relative to each other by about 20 A and brings two subunits, Al and Bl , much closer together than the other two subunits A2 and B2. This interaction is essential to the configuration of the two symmetrically disposed active sites and is also proposed to be conserved in the endonucleases from other organisms (see below). Active Site of the tRNA Splicing Endonuclease As discussed earlier, the product of the endonuclease cleavage reaction is a 2',3' cyclic phosphate and a 5-hydroxyl, a product identical to that of other ribonucleases such as RNaseA. In a well characterized acid-base catalyzed reaction RNaseA utilizes a histidine, His 12, to abstract a proton from the 2' -hydroxyl of the ribose leading to an inline attack on the adjacent phosphodiester bond and the formation of a pentacovalent reaction intermediate stabilized by lysine41. The general acid, Hisl 19, then protonates the 5' -leaving group leading to the 2',3' cyclic phosphate and 5' -hydroxyl product. In a second step, a proton is abstracted from a water molecule, OH- attacks, and the 2' ,3 ' cyclic phosphate is hydrolyzed to the 3 ' -phosphate (Thompson and Raines, 1994; Walsh, 1979). Inspection of the sequence alignment between the conserved members of the endonuclease family showed a single histidine residue that is absolutely conserved. Recall that the histidine to alanine mutation in Sen34 causes a marked reduction in the catalytic efficiency of 3' splice site cleavage. Similar mutants have been created in this conserved histidine of M. jannaschii and H. volcanii with a similar reduction in the catalytic efficiency of the cleavage reaction (Daniels, pers. comm.; Lykke-Andersen and Garrett, 1997). Thus it would appear that the tRNA splicing endonucleases catalyze the cleavage of RNA by a general acid-base catalysis. 30 With the crystal structure for the M. jannaschii endonuclease, this histidine was localized at the molecular level. Figure 1-7 A shows the environment surrounding histidine125 which is found in a cluster with the absolutely conserved residues Tyrl 15 (from Loop L7) and Lys156 (from a5), forming a pocket into which the scissile phosphate is proposed to fit. The three residues can be spatially superimposed with the catalytic triad of RNaseA (Li et al., 1998) leading to the prediction that His125 of the M. jannaschii enzyme is equivalent to the general base, Tyrl 15 is equivalent to the general acid and Lys156 stabilizes the transition state. Preliminary experiments indicate a role for both His125 and Lys156, but the role of Tyrl 15 has yet to be established (Trotta, unpub). Since the M. jannaschii enzyme is a homotetramer, each monomer is expected to contain a separate active site for cleavage. However, it is expected that only two of these active sites function in cleavage of the symmetric substrate. Recently, Diener and Moore (1998) have solved the NMR solution structure of a true archaeal substrate molecule. This molecule and a similar modeled substrate based on the TAR RNA structure (Li et. al. 1998) dock in plausible fashion with the enzyme (Figure I-8). The scissile phosphates are shown to fit nicely into the Al and B 1 active sites of the tetramer. The A2 and B2 active sites are too far apart to allow for interaction with the substrate. ON THE EVOLUTION OF THE ENDONUCLEASES The Archaeal Endonucleases The archaeal endonuclease of H. volcanii appears to be configured in a different manner than the M. jannaschii endonuclease. The protein behaves as a homodimer, not a homotetramer, but must be arrayed in a similar manner due to the identical substrate specificity. Based on the observation that the H. volcanii endonuclease is actually an inframe duplication of the endonuclease domain, Li et al. (1998) proposed a model to 31 FIGURE 1-8. Endonuclease docked with a hypothetical pre-tRNA substrate. The hypothetical tRNA substrate was constructed from joining the known crystal structure of yeast tRNAPhe with the archaeal substrate solved by Diener and Moore (1998). This represents the pre-tRNAArch-Euk demonstrated to be a substrate for both the eukaryotic and archaeal endonucleases (Fabbri et al., 1998) (see text) . The scissile phosphates are shown to dock directly in the active sites present on subunits Al and B 1 of the endonuclease. 32 Figure 1-8 33 describe how this enzyme is configured (Figure I-9B). The H. volcanii enzyme can be thought of as a pseudo-dimer with the N-terminal endonuclease domain repeat comprising one pseudo-monomer and the C-terminal repeat comprising the other. The repeats are connected by a stretch of polypeptide and serve to form the pseudo-dimer by the j39-j39' hydrophobic interaction which occurs between monomers in the M. jannaschii enzyme. Loop LIO present in the N-terminal repeat allows dimerization of the pseudo-dimers to form the active enzyme. Careful examination of the sequence of the N-terminal repeat shows the absence of the amino acids for the catalytic triad, thus the dimeric enzyme contains only two active sites present in the C-terminal repeat. The occurrence of an in-frame gene duplication event explains the configuration of H. volcanii enzyme (Lykke-Andersen and Garrett, 1997). This event has been localized to a set of closely related organisms of the Euryarchaeota branch of the Archaea (Figure I-5) which includes the genus Haloferax, Archaeoglobus and Methanosarcina. The Yeast Endonuclease Like the M. jannaschii endonuclease, the yeast tRNA splicing endonuclease is a tetramer. The two active site containing subunits, Sen2 and Sen34, must be configured in a similar manner to the Al and B 1 subunits of M. jannaschii. This arrangement is proposed to be facilitated by interactions with Sen54 and Sen15. Sen54 and Sen15 both contain a stretch of amino acids at their C-terminus that is homologous to the M. jannaschii endonuclease C-terrninus. This homology includes the crucial regions involved in the two sets of interactions seen in formation of the tetramer: the j39-j39' hydrophobic tail interaction and the LooplO insertion sequence. Thus it is proposed (Figure I-9C) that the Sen54-Sen2 and Sen34-Senl5 interactions detected in the two hybrid assay are 34 FIGURE I-9. Model of the tRNA splicing endonucleases of M. jannaschii, H. volcanii and S. cerevisiae. A. The M. jannaschii endonuclease is graphically depicted. Several important features are shown in the primary sequence and by a model: loop LlO for tetramerization, the COOH-terrninal P9 strands (arrows) for dirnerization and the conserved catalytic residue His125 (pentagon). B. The H. volcanii endonuclease consists of a tandem repeat of the endonuclease domain. The N-terrninal repeat has degenerated to possess only the loop LlO and COOH-terrninal interaction domains present in the M. jannaschii. The carboxy-terminal repeat contains the groove for loop LIO insertion as well as the catalytic triad including the conserved histidine (pentagon) . The dashed line represents the polypeptide chain connecting the amino-terminal repeat and the carboxy-terminal repeat. C. Proposed structural model of the yeast endonuclease. Upper: the sequence of the carboxy terminus of the M. jann. M. jannaschii, H vol. Nt., H. volcanii amino-terminal repeat and Sc. Sen54 and Sc. Sen15 is shown to be homologous. Lower: These important interaction elements are modeled in the yeast endonuclease. The two interaction elements are modeled as Loop LlO conserved in Sen54 and Senl5 and the COOH-terminal interaction (circled arrows) (see text for details) (adapted from Li et al., 1998). 35 ~ A M. jannaschii y HKL --■■ 115 125156 L10 NH2 Y H KL ... •--·-■■ ... .. r • • • • • • • • ' ~ C -------- . ~ M.jann. 15s L I H.vol Nt. 13sVVLAIViD r Sc.Sen54 425 FIIAIMD I Sc.Sen15 102 ILLAIIN Figure I-9 115 125156 36 mediated by the f39-f39' interaction. The Loop LIO interactions are proposed as in the M. jannaschii structure to mediate dimer-dimer interactions. Interestingly it has recently been shown that the eukaryotic endonuclease can recognize and cleave an archaeal bulge-helix-bulge containing tRNA substrate (Fabbri et al., 1998). This tRNA substrate was constructed by creating a hybrid tRNA containing an archaeal intron fused to yeast pre-tRNAPhe_ The eukaryotic enzyme can recognize and correctly cleave the 5' and 3' splice sites in this substrate. In doing so, the enzyme dispenses with the ruler mechanism. There is no change in the size or location of the intron released from the pre-tRNA upon addition or deletion of base-pairs in the anticodon stem as was seen for endonucleolytic cleavage of eukaryotic substrates. This result proves that the disposition of the active sites is identical in the archaeal and eukaryotic endonucleases and that this architecture has been conserved since the divergence from a common ancestor. It must be through specialization of the yeast Sen54 and Senl5 subunits that the ruler mechanism has evolved. The SEN2 gene has acquired a transrnembrane sequence, which likely anchors the endonuclease in the inner nuclear membrane perhaps near nuclear pore structures. This has been shown to be a primary localization of the tRNA splicing ligase (Clark and Abelson, 1987) and the two enzymes likely act in concert at this site (Greer, 1986). ORIGIN OF tRNA INTRONS The relatedness of the eukaryotic and archaeal enzymes almost certainly proves that the endonuclease gene was present in their last common ancestor. The simplest hypothesis is that its function then was to splice tRNA precursors, but it could also be that it had a different function and has been independently recruited in both lines to a tRNA splicing function. 37 Certainly recruitment is a theme that is a feature of the enzymes in this system. It has recently been shown that tRNA ligase functions in the ligation of an unusually spliced yeast messenger RNA (Cox and Walter, 1996; Sidrauski et al., 1996). The rnRNA encodes a transcription factor, HACl, which upregulates transcription of genes involved in the maintenance of unfolded protein in the endoplasmic reticulum (ER). The presence of unfolded proteins appears to be sensed by a receptor tyrosine kinase-like protein, IREl, which spans the membrane of the ER. The Irel protein contains a nuclease domain capable of cleaving the Hacl rnRNA and releasing an intron (Sidrauski and Walter, 1997). Ligation of the two Hacl exons is dependent on the function of tRNA ligase both in vivo and in vitro (Sidrauski et al., 1996; Sidrauski and Walter, 1997). Two hypothesis have been postulated for the origin of tRNA introns. The first (Cavalier-Smith, 1991) posits that existing proteins were recruited to splice introns resulting from the partial deletion of pre-existing group I or group II self-splicing introns in tRNA or rRNA genes. Support for such a hypothesis is derived from the presence of group I and group II introns in the tRNA genes of a handful of bacterial species. In particular a group I intron found in the tRNALeu gene of cyanobacteria and bacteria interrupts the anticodon loop at the same position as a protein-dependent intron found in some Archaea (Wich et al., 1987). A second hypothesis proposes the expansion of loops of the tRNA molecule. In both instances the pre-existence of compatible splicing machinery in the cell would insure against lethality of the insertion (Belfort and Weiner, 1997). It is unlikely that agreement on a time for the origin of tRNA introns is possible; however, the new addition to the ongoing dialogue is that at least the splicing endonuclease is ancient. 38 CONCLUDING REMARKS Whatever the origin of tRNA intrans, it is clear that they are here to stay. Both Archaea and eukaryotes have failed to displace all intrans from tRNA genes. This would suggest that the intrans serve some selective advantage for the organism and that maintenance of the splicing system is necessary for survival. It is clear from studies in yeast that one function of intrans is to aid in the modification of tRNAs (for review see Grosjean et al., 1997). Thus, the modification enzymes have evolved to depend on the presence of the intron for correct substrate recognition and modification. By studying the enzymology of the removal process, we have begun to understand just how intrans have evolved. It will be interesting to fully delineate the splicing pathway of the archaeal system through discovery of the ligase that must function to ligate the products of the endonuclease reaction. Perhaps we will again be afforded a glimpse into the RNA world that existed before the invention of protein enzymes. REFERENCES Abelson, J. , Trotta, C.R., and Li, H. (1998). tRNA splicing. J. Biol. Chem. 273, 1268512688. Apostol, B. L. , Westaway, S. K. , Abelson, J. , and Greer, C. L. (1991). Deletion analysis of a multifunctional yeast tRNA ligase polypeptide: identification of essential and dispensable functional domains. J. Biol. Chem. 266, 7445-7455 . Arn, E. A , and Abelson, J. (1998). RNA ligases: function, mechanism, and sequence conservation. In RNA structure and function (Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press), pp. 695-726. Arn, E. A , and Abelson, J. N. (1996) . The 2'-5' ligase from Escherichia coli: purification, cloning and genomic disruption. J. Biol. Chem. 271, 31145-31153. Baldi, M. I. , Mattoccia, E., Bufardeci, E., Fabbri, S. , and Tocchini-Valentini, G. P. (1992). Participation of the intron in the reaction catalyzed by the Xenopus tRNA splicing endonuclease. Science 255, 1404-1408. 39 Belford, H. G., Westaway, S. K., Abelson, J., and Greer, C. L. (1993). Multiple nucleotide cofactor use by yeast ligase in tRNA splicing: evidence for independent ATPand GTP- binding sites. J. Biol. Chem. 268, 2444-2450. Belfort, M., and Weiner, A (1997). Another bridge between kingdoms:tRNA spling in Archaea and Eukaryotes. Cell 89, 1003-1006. Biniszkiewicz, D. , Cesnaviciene, E., and Shub, D. A (1994). Self-splicing group-I intron in cyanobacterial initiator methionine tRNA: evidence for lateral transfer of intrans in bacteria. EMBO Journal 13, 4629-4635. Cavalier-Smith, T. (1991). Intron phylogeny: a new hypothesis. Trends Genet. 7, 145-148. Clark, M. W., and Abelson, J. (1987). The subnuclear localization of tRNA ligase in yeast. Journal Of Cell Biology 105, 1515-1526. Cox, J. S., and Walter, P. (1996) . A novel mechanism for regulating activity of a transcription factor that controls the unfolded protein response. Cell 86, 391-404. Culbertson, M. R., and Winey, M. (1989). Split tRNA genes and their products: a paradigm for the study of cell-function and evolution. Yeast 5, 405-427. Culver, G. M., McCraith, S. M., Consaul, S. A, Stanford, D. R., and Phizicky, E. M. (1997). A 2'-phosphotransferase implicated in tRNA splicing is essential in Saccharomyces cerevisiae. J. Biol. Chem. 272 , 13203-13210. Culver, G. M., McCraith, S. M., Zillmann, M., Kierzek, R., Michaud, N., Lareau, R. D., Turner, D. H., and Phizicky, E. M. (1993). An NAD derivative produced during tRNA splicing: ADP-ribose 1"-2" cyclic phosphate. Science 261, 206-208. Daniels, C. J., Gupta, R., and Doolittle, W. F. (1985). Transcription and excision of a large intron in the tRNATrp gene of an Archaebacterium, Halobacterium volcanii. J. Biol. Chem. 260, 3132-3134. Diener, J., and Moore, P. (1998). Solution structure of a substrate for the archaeal pretRNA splicing endonucleases: the bulge-helix-bulge motif. Molecular Cell. Fabbri, S., Fruscoloni, P., Bufardeci, E., Di Nicola Negri, E., Baldi, M. I., Gandini Attardi, D., Mattoccia, E., and Tocchini-Valentini, G. P. (1998). Conservation of substrate recognition mechanisms by tRNA splicing endonucleases. Science 280, 284-286. Filipowicz, W., Konarska, M., Gross, H., and Shatkin, A (1983). RNA 3'-terrninal phosphate cyclase activity and RNA ligation in HeLa cell extract. Nucleic Acids Res . 11, 1405-1418. Gomes, I., and Gupta, R. (1997). RNA splicing ligase activity in the archaeon Haloferax volcanii. Biochemical and Biophysical Review 237, 588-594. 40 Goodman, H. M. , Olson, M. V., and Hall, B. D. (1977). Nucleotide sequences of a mutant eukaryotic gene: the yeast tyrosine inserting suppressor, SUP4-0. Proc. Natl. Acad. Sci. USA 74, 5453-5457. Green, C. J. , Sohel, I. , and Vold, B. S. (1990). The discovery of new intron-containing human tRNA genes using the polymerase chain reaction. J. Biol. Chem. 265, 1213912142. Greer, C. L. (1986). Assembly of a tRNA splicing complex: evidence for concerted excision and joining steps in splicing in vitro. Mol. Cell. Biol. 6, 635-644. Greer, C. L. , Soll, D., and Willis, I. (1987). Substrate recognition and identification of splice sites by the tRNA-splicing endonuclease and ligase from Saccharomyces cerevisiae. Molecular Cellular Biology 7, 76-84. Grosjean, H., Szweykowska-Kulinska, Z., Motorin, Y. , Fasiolo, F. , and Simos, G. (1997). Intron-dependent enzymatic formation of modified nucleosides in eukaryotic tRNAs: a review. Biochimie 79, 293-302. Ho , C. K., Rauhut, R. , Vijayraghavan, U., and Abelson, J. (1990). Accumulation of pretRNA splicing 2/3 intermediates in a Saccharomyces cerevisiae mutant. Embo Journal 9, 1245-1252. Hopper, A. K., Banks, F. , and Evangelidis, V. (1978). A yeast mutant which accumulates precursor tRNAs. Cell 14, 211-219. Johnson, J., Ogden, R. , Johnson, P. , Abelson, J., Dembeck, P., and Itakura, K. (1980). Transcription and processing of a yeast tRNA gene containing a modified intervening sequence. Proceedings of the National Academy of Sciences, USA 77, 2564-2568. Kjems, J. , and Garrett, R. A. (1985). An intron in the 23S rRNA gene of the Archaebacterium Desulfurococcus mobilis. Nature 318, 675-677. Kjems, J. , Jensen, J., Olesen, T., and Garrett, R. A. (1989). Comparison Of tRNA and rRNA intron splicing in the extreme thermophile and Archaebacterium Desulfurococcus mobilis. Canadian Journal Of Microbiology 35, 210-214. Kleman-Leyer, K. , Armbruster, D. A, and Daniels, C. J. (1997). Properties of H. volcanii tRNA intron endonuclease reveal a relationship between the archaeal and eucaryal tRNA intron processing systems. Cell 89, 839-848. Klenk, H., Clayton, R. , Tomb, J., White, 0. , Nelson, K., Ketchum, K., Dodson, R., Gwinn, M., Hickey, E., Peterson, J. , Richardson, D., Kerlavage, A. , Graham, D., Kyrpides, N. , Fleischmann, R. , Quackenbush, J. , Lee, N., Sutton, G., Gill, S., Kirkness, E., Dougherty, B., McKenney, K. , Adams, M., Loftus, B. , Peterson, S. , Reich, C., McNeil, L., Badger, J., Glodek, A. , Zhou, L. , Overbeek, R. , Gocayne, J., Weidman, J. , 41 McDonald, L., Utterback, T., Cotton, M., Spriggs, T., Artiach, P., Kaine, B., Sykes, S., Sadow, P., D'Andrea, K., Bowman, C., Fujii, C., Garland, S., Mason, T., Olsen, G., Fraser, C., Smith, H., Woese, C., and JC, V. (1998). The complete genome sequence of the hyperthermophilic, sulphate-reducing archaeon Archaeoglobus fulgidus. Nature 390, 364. Knapp, G., Beckmann, J. S., Johnson, P. F., Fuhrman, S. A, and Abelson, J. N. (1978). Transcription and processing of intervening sequences in yeast tRNA genes. Cell 14, 221236. Konarska, M., Filipowicz, W., Domdey, H., and Gross, H. (1981). Formation of a 2'-· phosphomonoester, 3',5'-phosphodiester linkage by a novel RNA ligase in wheat germ. Nature 293. Kuhsel, M. G., Strickland, R., and Palmer, J. D. (1990). An ancient group I intron shared by eubacteria and chloroplasts. Science 250, 1570-1573. Laski, F., Fire, A, RajBhandary, U., and Sharp, P. (1983). Characterization of tRNA precursor splicing in mammalian extracts. J. Biol. Chem. 258, 11974-11980. Lee, M. C., and Knapp, G. (1985). tRNA splicing in Saccharomyces cerevisiae: secondary and tertiary structures of the substrates. J. Biol. Chem. 260, 3108-3115. Li, H., Trotta, C. R., and Abelson, J. N. (1998). Crystal structure and evolution of a tRNA splicing enzyme. Science 280, 279-284. Lykke-Andersen, J., and Garrett, R. A (1997). RNA-protein interactions of an archaeal homtetrameric splicing endoribonuclease with an exceptional evolutionary history. EMBO Journal 16, 6290-6300. McCraith, S. M., and Phizicky, E. M. (1991). An enzyme from Saccharomyces cerevisiae uses NAO+ to transfer the splice junction 2'-phosphate from ligated tRNA to an acceptor molecule. J. Biol. Chem. 266, 11986-11992. McCraith, S. M., and Phizicky, E. M. (1990). A highly specific phosphatase from Saccharomyces cerevisiae implicated in tRNA splicing. Mol. Cell. Biol. 10, 1049-1055. Nishikura, K., and De Robertis, E. (1981). RNA processing in microinjected Xenopus oocytes. Sequential addition of base modification in the spliced transfer RNA J. Mol. Biol. 145, 405-420. Ogden, R. C., Lee, M. C., and Knapp, G. (1984). Transfer RNA splicing In Saccharomyces cerevisiae: defining the substrates. Nucleic Acids Res. 12, 9367-9382. Peebles, C. L., Gegenheimer, P., and Abelson, J. (1983). Precise excision of intervening sequences from precursor tRNAs by a membrane-associated yeast endonuclease. Cell 32, 525-536. 42 Phizicky, E. M. , Consaul, S. A., Nehrke, K. W., and Abelson, J. (1992). Yeast tRNA ligase mutants are nonviable and accumulate tRNA splicing intennediates. J. Biol. Chem. 267, 4577-4582. Phizicky, E. M., Schwartz, R. C., and Abelson, J. (1986). Saccharomyces cerevisiae tRNA ligase: purification of the protein and isolation of the structural gene. J. Biol. Chem. 261, 2978-2986. Rauhut, R., Green, P.R. , and Abelson, J. (1990). Yeast tRNA-splicing endonuclease is a heterotrimeric enzyme. J Biol. Chem. 265, 18180-18184. Reinhold-Hurek, B., and Shub, D. A. (1992). Self-splicing intrans in tRNA genes of widely divergent bacteria. Nature 357, 173-176. Reyes, V. M., and Abelson, J. (1988). Substrate recognition and splice site determination in yeast tRNA splicing. Cell 55, 719-730. Schneider, A., Perry-McNally, K., and Agabian, N. (1993). Splicing and 3'-processing of the tyrosine tRNA of Trypanosoma brucei. J. Biol. Chem. 268, 21868-21874. Schwartz, R., Greer, C., Gegenheimer, P., and Abelson, J. (1983). Enzymatic mechanism of an RNA ligase from wheat genn. J. Biol. Chem. 258, 8374-8383. Sidrauski, C. , Cox, J. S. , and Walter, P. (1996). tRNA ligase is required for regulated mRNA splicing in the unfolded protein response. Cell 86, 405-413. Sidrauski, C., and Walter, P. (1997). The transmembrane kinase Irelp is a site-specific endonuclease that initiates mRNA splicing in the unfolded protein response. Cell 90, 10311039. Spinelli, S., Consaul, S., and Phizicky, E. (1997). A conditional lethal yeast phophotransferase (tptl) mutant accumulates tRNAs with a 2'-phosphate and an undennodified base at the splice junction. RNA 3, 1388-1400. Stange, N., and Beier, H. (1986). A gene for the major cytoplasmic tRNATyr from Nicotiana rustica contains a 13 nucleotides long intron. Nucleic Acids Res. 14, 8691. Strobel, M., and Abelson, J. (1986). Effect of intron mutations on processing and function of Saccharomyces cerevisiae SUP53 tRNA in vitro and in vivo. Mol. Cell. Biol. 6, 26632673. Swerdlow, H. , and Guthrie, C. (1984). Structure of intron-containing tRNA precursors: analysis of solution confonnation using chemical and enzymatic probes. J. Biol. Chem. 259, 5197-5207. 43 Thompson, J. E., and Raines, R. T. (1994). Value of general acid-base catalysis to ribonucleaseA. Journal of the American Chemical Society 116, 5467-5468. Thompson, L. D. , Brandon, L. D., Nieuwlandt, D. T., and Daniels, C. J. (1989). Transfer RNA intron processing in the halophilic archaebacteria. Canadian Journal Of Microbiology 35, 36-42. Thompson, L. D., and Daniels, C. J. (1988). A tRNAtrp intron endonuclease from Halobacterium volcanii: unique substrate recognition properties. J. Biol. Chem. 263, 17951-17959. Trotta, C.R., Miao, F., Am, E. A, Stevens, S. W., Ho, C. K., Rauhut, R. , and Abelson, J. N. (1997). The yeast tRNA splicing endonuclease: A tetrameric enzyme with two active site subunits homologous to the archaeal tRNA endonucleases. Cell 89, 849-858. Valenzuela, P., Venegas, A , Weinberg, F., Bishop, R., and Rutter, W. J. (1978). Structure of yeast phenylalanine-tRNA genes: an intervening DNA segment within the coding region for the tRNA Proceedings of the National Academy of Sciences USA 75, 190-194. Walker, G., Uhlenbeck, 0. C., Bedows, E., and Gumport, R. I. (1975). T4-induced RNA ligase joins single-stranded oligoribonucleotides. Proceedings of the National Academy of Sciences, USA 72, 122-126. Walsh, C. (1979). Enzymatic reaction mechanisms (San Francisco: W.H. Freeman and Company). Wich, G. , Leinfelder, W., and Bock, A (1987). Genes for stable RNA in the extreme thermophile Thermoproteus-tenax intrans and transcription signals. EMBO J. 6, 523-528. Willis, I., Hottinger, H., Pearson, D., Chisholm, V., Leupold, U., and Soll, D. (1984). Mutations affecting excision of the intron from a eukaryotic dimeric tRNA precursor. EMBO J. 3, 1573-1580. Winey, M., and Culbertson, M. R. (1988). Mutations affecting the tRNA splicing endonuclease activity of Saccharomyces cerevisiae. Genetics 118, 49-63. Xu, Q., Teplow, D., Lee, T. D., and Abelson, J. (1990) . Domain structure in yeast tRNA ligase. Biochemistry 29, 6132-6138. Zillmann, M., Gorovsky, M. A, and Phizicky, E. M. (1991). Conserved mechanism of tRNA splicing in Eukaryotes. Mol. Cell. Biol. 11, 5410-5416. Zillmann, M., Gorovsky, M. A, and Phizicky, E. M. (1992). Hela cells contain a 2'phosphate-specific phosphotransferase similar to a yeast enzyme implicated in tRNA splicing. J. Biol. Chem. 267, 10289-10294. 44 Chapter II - The Yeast tRNA Splicing Endonuclease: A Two Active Site Tetrameric Enzyme Homologous to Archaeal tRNA Splicing Endonucleases (Christopher R. Trotta, Feng Miao , Eric A. Arn, Scott W. Stevens, Calvin K. Ho , Reinhard Rauhut, and John N. Abelson. Published in Cell 89:849-858, 1997; Appendix II) Abstract The splicing of tRNA precursors is essential for the production of mature tRNA in organisms from all major phyla. In yeast, the tRNA splicing endonuclease is responsible for identification and cleavage of the splice sites in pre-tRNA. We have cloned the genes encoding all four protein subunits of endonuclease. Each gene is essential. Two subunits, Sen2p and Sen34p, contain a homologous domain of approximately 130 amino acids. This domain is found in the gene encoding the archaeal tRNA splicing endonuclease of H. volcanii and in other Archaea. Our results demonstrate that the eucaryal tRNA splicing endonuclease contains two functionally independent active sites for cleavage of the 5' and 3' splice sites, encoded by SEN2 and SEN34 respectively. The presence of endonuclease in Eucarya and Archaea suggests an ancient origin for the tRNA splicing reaction. Introduction Although RNA splicing is a required step for gene expression in all three taxonomic kingdoms, at least four different mechanisms of splicing have evolved. The most primitive mechanism, autocatalytic self splicing of Group I and Group II introns, involves two phosphotransfer reactions (Cech, 1993). Pre-mRNA splicing requires ATP hydrolysis and more than 100 proteins, but it proceeds via the same chemical mechanism 45 used by Group II introns suggesting that it is also RNA-catalyzed and that it shares a common origin with Group II intron splicing (Madhani and Guthrie, 1994; Moore, 1993). In contrast to these RNA-catalyzed reactions, splicing of eucaryotic tRNA introns occurs by a mechanism involving three proteinaceous enzymes (Abelson, 1991; Culbertson and Winey, 1989; Phizicky and Greer, 1993). In the archaeal kingdom tRNA splicing proceeds by a similar protein-based enzymatic mechanism, but this splicing is distinct by a number of functional criteria (Thompson et al., 1989). Splicing of all known prokaryotic tRNA intrans occurs via the primitive autocatalytic mechanism (Biniszkiewicz et al. , 1994; Reinhold-Hurek and Shub, 1992). So while tRNA splicing occurs in all three kingdoms, the mechanism for this process is not conserved. Major questions regarding the origin of tRNA intrans and the relationship between the disparate enzymatic machineries required for their removal therefore remain. Do the different splicing mechanisms reflect independent origins or a single origin and later divergence? The results of this work together with those in the accompanying papers of Di Nicola Negri et al. (1997) and Kleman-Leyer et al. (1997) provide new insight into the origin of the tRNA spicing machinery. Much of our knowledge of the mechanism of eucaryotic tRNA splicing comes from studies in Saccharomyces cerevisiae. The S. cerevisiae genome contains 272 tRNA genes of which 59, encoding ten different tRNAs, contain introns (CRT, unpublished data) . These intrans are 14 to 60 bases in length and all interrupt the anticodon loop immediately 3' to the anticodon (Ogden et al., 1984). Splicing of intrans from precursor tRNA (pre-tRNA) is accomplished through the action of a site-specific endonuclease, an RNA ligase, and a phosphotransferase. The tRNA splicing endonuclease (referred to herein as "endonuclease") cleaves pre-tRNA at the 5' and 3' splice sites to release the 46 intron. The products of the endonuclease reaction are an intron and two tRNA halfmolecules bearing 2' ,3 ' cyclic phosphate and 5'-0H termini (Peebles et al., 1983). tRNA ligase covalently joins the half-molecules through a complex series of ATP- and GTPdependent reactions (Belford et al., 1993; Westaway et al. , 1993; Westaway et al., 1988). After ligation, a 2' phosphate remains at the splice junction and is removed by a specific phosphotransferase (Culver et al. , 1993; McCraith and Phizicky, 1991). Figure II-IA summarizes the conserved features of the ten yeast pre-tRNAs containing intrans (as depicted by Lee and Knapp, 1984). There are no conserved sequences at the splice sites, but the intron is invariably located at the same site in the gene, placing the splice sites an invariant distance from the constant structural features of the tRNA body. This conserved placement implies that recognition of the splice sites may be accomplished via interaction with the mature tRNA domain. The hypothesis that the endonuclease measures the distance from the mature domain to the splice sites to correctly place cleavages (Figure II-lB) was confirmed by experiments where insertions in the anticodon stem of pre-tRNA which changed the distance between the mature domain and the splice sites resulted in a predictable shift of the cleavage sites (Reyes and Abelson, 1988). Further evidence for a strict measuring mechanism was provided by the fact that an artificial tRNA intron composed almost entirely of uridine residues was spliced correctly in the yeast system (Reyes and Abelson, 1988). The intron is not, however, completely passive in the recognition of the splice sites. The Xenopus tRNA endonuclease has been shown to recognize a crucial structural element in tRNA intrans (Baldi et al. , 1992). Yeast tRNA intrans contain a conserved purine residue three nucleotides upstream of the 3' splice site. This base must be able to pair with a pyrimidine at position 32 in the anticodon loop in order for the intron to be 47 FIGURE II-1. Conserved Sequence and Structure Elements of Yeast Pre-tRNA and Archaeal Minimal Pre-tRNA A. A composite, modified from Lee and Knapp (1985), of yeast pre-tRNA sequences (Ogden et al., 1984). X = unconserved base in a region of variable length or secondary structure, 0 = an unconserved base in a region of conserved secondary structure. G, C and A indicate conserved bases conserved among pre-tRNAs. Y and R = conserved pyrimidines and purines, respectively. The conserved intron-anticodon tertiary interaction is depicted by a line between the conserved R and Y bases. B. Taken from Reyes and Abelson (1988). Presumed tertiary structure of pre-tRNAPhe (from Lee and Knapp, 1984). Endonuclease is proposed to interact with the mature domain and measure the distance to the two splice sites. C. A composite of archaeal pre-tRNA sequences, modified from Thompson and Daniels (1988). Position designations are as in panel A. This represents the minimal substrate for cleavage by the archaeal endonuclease. 48 A GG A-OH C C XXX XX C XOG I I I I xx XGXOY B C 0-0 0-0 0-0 0-0 0-0 0-0 UO-O OOOOGyOA R I I I I I R xOOOOCT Uc X xXX 00-0 0-0 0-0 0-0 0-00 / 0 v-FP U X 0-0 0-0 0-0 R X -x. X · xxx Yeast Splicing Substrate Consensus Yeast Endonuclease Ruler Model Figure II-1 Archaeal Splicing Substrate Consensus 49 recognized by either yeast or Xenopus endonuclease (Figure 11-lA) (Baldi et al., 1992; Di Nicola Negri et al., 1997). This data suggested there could be different requirements for the recognition of the 5' and 3' splice sites. More recent evidence in this paper and the accompanying paper by Di Nicola Negri et al. support the hypothesis that the endonuclease contains two independent active sites with different substrate recognition mechanisms. The mechanism of intron removal in tRNA splicing has also been investigated through purification of the endonuclease enzyme from yeast (Rauhut et al., 1990). Endonuclease behaves as a membrane-bound protein present in a low abundance of~ 100 molecules per (Rauhut et al., 1990). The purified enzyme was seen to contain three subunits of 54, 44 and 34kD and, as we report here, a fourth subunit of 15kD. The molecular weight of the holoenzyme as judged by gel exclusion chromatography implies that the enzyme is an apy8 tetramer. In this paper we report the primary sequences of all four of the endonuclease subunits. Analysis of the peptide sequence of the endonuclease subunits presented here sheds light on the architecture of the enzyme and its possible interactions with other components of the nucleus. The pre-tRNAs of Archaea are similar to those of Eucarya in that they contain introns which must be removed by protein enzymes in trans. Superficially, these introns are similar to yeast tRNA introns in that they are often (though not always) located in the anticodon loop. Removal of tRNA introns by either eucaryal or archaeal endonucleases yield 2',3' cyclic PO4 and 5'-OH terrninii. Studies of removal of tRNA introns in Archaea have, however, shown notable differences between the yeast and archaeal tRNA endonucleases. Yeast pre-tRNAs are not substrates for the Haloferax volcanii endonuclease (Palmer et al., 1994), and the mature domain of a Halobacterium volcanii 50 pre-tRNA is not required for correct excision of the intron (Thompson and Daniels, 1988). Rather, archaeal tRNA splice sites must be located in bulged loops separated by a helix of four base-pairs (Figure II-lC). This evidence would seem to suggest a different origin of the splicing process in these two kingdoms. The results presented here, however, together with those in the accompanying paper by Kleman-Leyer et al. (1998) establish that two of the yeast endonuclease subunits define a gene family that includes the single homodimeric 40kDa Haloferax volcanii endonuclease as well as related genes from other Archaea. We discuss the evolutionary and mechanistic implications of these findings . Results The gene for the 44kD subunit, SEN2. Our primary goal in the characterization of the yeast tRNA splicing endonuclease has been to clone the genes for each of the subunits of this enzyme. We previously reported the isolation and characterization of the cold sensitive sen2-3 mutation, an allele of the SEN2 gene found by Winey and Culbertson (1988) in their search for yeast tRNA mutants defective in tRNA splicing in vitro. At the non-permissive temperature sen2-3 accumulates "2/3" pre-tRNA molecules cleaved at the 3' splice site only (Ho et al., 1990) strongly suggesting that Sen2p is a component of endonuclease. In this paper we report the cloning of the SEN2 gene by complementation of the sen2-3 mutant and the determination of its nucleotide sequence (Genbank: accession #M32336). Analysis of the amino acid sequence of Sen2p reveals the presence of a probable transmembrane sequence between residues 225 and 243 (Figure II-5B). During purification endonuclease behaves as a membrane-bound enzyme (Rauhut et al. , 1990). None of the other subunits (see below) contain a convincing transmembrane sequence, so it is possible that Sen2p is the 51 subunit which anchors the endonuclease complex to the nuclear membrane. Sen2p also contains a domain predicted to form a canonical anti-parallel coiled-coil structure between residues 122 and 184 (Figure II-5E). Sequencing of the mutant sen2-3 allele reveals a single point mutation causing the change of glycine 292 to glutamate (Figure II-3B). A new purification strategy for the tRNA splicing endonuclease We had previously developed a chromatographic procedure for the purification of endonuclease which resulted in homogenous enzyme (Rauhut et al., 1990), and provided sufficient material for the characterization of two of the endonuclease subunits. However, this procedure was extremely time consuming and gave a low yield of enzyme. We have developed a highly efficient affinity purification procedure for large-scale preparation of the enzyme based on the cloned SEN2 gene. The SEN2 gene was modified by addition of an amino-terminal peptide tag consisting of eight histidine residues followed by the 4 residue FLAG epitope. This gene, cloned in a yeast 2µ vector, complemented the haplo-lethal phenotype of a SEN2 deletion, proving that the affinity tag did not affect the activity of Sen2p in vivo. The complemented sen2 deletion strain, CRT44, was used as a source of tagged enzyme for purification by a procedure utilizing standard chromatographic methods and two specific affinity steps, as diagrammed in Figure II-2A. This rapid procedure resulted in an efficient (10-20%) yield of 5 µg of pure enzyme per 150 grams of cells. A fourth subunit in the tRNA splicing endonuclease Pure endonuclease preparations had previously been characterized by SDS-PAGE and silver staining, which detected three subunits with apparent molecular masses of 52 FIGURE II-2. Affinity Purification of Endonuclease from Yeast A. Vector YpH8/FLAG SEN2 constructed as per Experimental Procedures. HIS8 = eight histidine tag, R.,AG = FLAG peptide epitope tag. This vector was used to complement a SEN2 chromosomal deletion. B. Purification procedure for isolation of affinity tagged endonuclease from yeast strain CRT44 (see Experimental Procedures for details). 53 A B ~. Crude Extract + 1%TritonX-100 1100,000xg 1hr + Pellet + Supernatant ,~-......... IDEAE-Sephacryl Flowthrough High Salt Wash I Heparin Hyper-D + Flow- + 0.1 M_.0.8M NaCl Gradient th rough I Ni-NTA Agarose + + Flowthrough YpHS/FLAG SEN2 150mM lmidazole Elution Ia -FLAG Mab Agarose + Flowthrough + FLAG Peptide Elution i ~1 Figure II-2 54 51kD, 42kD and 3lkD (Rauhut et al. , 1990). To compare endonuclease purified by the affinity method to that obtained in our original purification, we employed a high-resolution gradient gel procedure. This procedure revealed that both preparations contain proteins of 54, 44 and 34kD (cf. lanes A and R Figure II-6A), although the 44kD band in lane A has a lower mobility due to its affinity tag (the molecular weights of the proteins are designated here as determined by comparison with markers in gradient gel electrophoresis). Both preparations contained, in addition, a fourth band with an apparent molecular weight of 15kD. This band would have run with the dye front in our original procedure. The fact that all four proteins are present in stoichiometric amounts in endonuclease purified by different purification procedures and that all four proteins cosediment in glycerol density gradients (data not shown) strongly suggests that all four polypeptides are subunits of the enzyme. The molecular mass of the enzyme as measured by gel filtration chromatography in our earlier study was 140kD (Rauhut et al., 1990). With the fourth subunit, the predicted mass of a 1: 1: 1: 1 endonuclease complex is 146kD, very close to the observed molecular mass. It is therefore likely that the subunit constitution of this enzyme is a~yo. The genes for the 34kD, 54kD, and lSkD subunits A 50 microgram aliquot of purified endonuclease, produced by the original Rauhut procedure, was fractionated by SOS-PAGE, and individual bands were excised from the gel. These polypeptides were digested with lysyl-protease and a peptide resulting from digestion of the 34kD subunit yielded the sequence FIAYPGDPLRFX(R ?)XLTIQ. A search of the yeast genome database revealed that this peptide is uniquely found in the open reading frame (ORF) YAR008w on S. cerevisiae Chromosome I (Figure II-3C). 55 This ORF potentially encodes a protein with a predicted molecular weight of 34kD. We have renamed this gene SEN34. Further evidence described below confirms that this protein is the 34kD subunit of the endonuclease. Five micrograms of endonuclease purified by the affinity method described above were also fractionated by SDS-PAGE, the 54kD band was digested with lysyl-protease and one unambiguous peptide sequence was obtained, NDDLQHFPTYK, which matches the sequence encoded by ORF YPL083c of S. cerevisiae Chromosome XVI (Figure II3A). This 467 amino acid protein has a predicted molecular weight of 54kD. We have renamed this gene SEN54. Sen54p is overall a very basic protein (pI=9.5) but it contains several highly acidic regions, e.g., the sequence EEEEEEEEELQQD (residues 19-31) and EFLSEIEDDDDWE (residues 346-358). The significance of these extreme disparities in charge remains to be investigated. Further evidence described below suggests that this protein is a subunit of endonuclease. Using a variation of the affinity purification method (see Experimental Procedures), several micrograms of pure 15kD subunit were prepared from SDS-PAGE:fractionated endonuclease. This material was analyzed by trypsin digestion followed by LC-MS/MS analysis and three peptides were sequenced. Two of these were identical except for a possible acetylation modification of the methionine in the peptide sequence MATTDIISLVK, suggesting that this peptide is the amino terminus of the protein (K. Swiderek, pers. comm). The two unique peptides are found in a protein encoded by ORF YMR059w on Chromosome XIII (Figure 11-3D). This ORF potentially encodes a protein of 17kD. However, if the methionine codon at position 20 in this ORF is the translation start site for this protein, as suggested by the peptide sequencing, the resulting protein would be predicted to have a length of 124 amino acids with a molecular weight of 56 FIGURE II-3. Primary Amino Acid Sequence of the Four Subunits of the Yeast tRNA Splicing Endonuclease A. Sen54p B. Sen2p C. Sen34p D. Sen15p. Residues which are underlined represent peptides which were directly sequenced. Residues shown in gray boxes represent either bi-partite or basic nuclear localization sequences as described by Robbins et al., (1991) or Kalderon et al., (1984), respectively. Residues shown in a black box represent a predicted transmembrane domain (see figure II5 for details) . Residues shown in open boxes represent potential nuclear export signals as described by Taylor and co-workers (1995) and Ltihrmann and co-workers (1995). In panel B residue 292, the underlined and bold G represents the mutation in the sen2-3 mutant (Ho et al., 1990). In panel D the* represent the presumed initiator methionine, and the italicized residues represent the presumed mitochondrial pre-sequence (see text for details). 57 A Sen54p MQFAGKKTDQVTTSNPGFEEEEEEEEELQQDWSQLASLVSKNAALSLPKR GEKDYEPDGTNLQDLLLYNASKAMFDTISDSIRGTTVKSEVRGYYVPHKH QAVLLKPKGSFMQTMGRADSTGELWLDFHEFVYLAERGTILPYYRLEAGS NKSSKHETEILLSMEDLYSLFSSQQEMDQYFVFAHLKRLGFILKPSNQEA AVKTSFFPLKKQRSNLQAITWR~LSLFKIQELSLfSGFFYSKWNFFFRKY TTSPQLYQGLNRLVRSVAVPKNKKELLDAQSDREFQKVKDIPLTFKVWKP HSNFKKRDPGLPDFQVFVYNKNQDLQHFPTYKELRSMFSSLDYKFEFLSE IEDDDDWETNSYVEDIPRKEYIHKRSAKSQTEKSESSMKASFQKKTAQSS T KKK R KAY PP H I NIR R L K '!'.JG YRS F I I A I MD NG L I S F V KM S EA DF GS E S V WYTPNTQKKVDQRWKKH 50 100 150 200 250 300 350 400 450 467 B Sen2p MSKGRVNQKRYKYPLPIHPVDDLPELILHNPLSWLYWAYRYYKSTNALND 50 KVHVDFIGDTTLHITVQDDKQMLYLWNNGFFGTGQFSRSEPTWKARTEAR 100 LGLNDTPLHNRGGTKSNTETEMTLEKVTQQRRLQRLEFKKERAKLERELL__I 150 IELRKKPGHIDEENILLEKQRESLRKFKLKQTEDVGIVAQQQDISESNLRD 200 EDNNLLDENGD~LPLESLELMPVE KLFQFD 250 AKYKDIHSFVRSYVIYHHYRSHGWCVRSGIKFGCDYLLYKRGPPFQHAEF 300 CVMGLDHDVSKDYTWYSSIARVVGGAKKTFVLCYVERLISEQEAIALWKS 350 NNFTKLFNSFQVGEVLYKRWVPGRNRD 377 C Sen34p MAEEGGTRIAINIYAKRTAKGEEVFMPPLVFDIDHIKLLRKWGICGVLSG TLPTAAQQNVFLSVPLRLMLEDVLWLHLNNLADVKLIRQEGDEIMEGITL ERGAKLSKIVNDRLNKSFEYQ~K F KKDEHIAKLKKIGR~NDKTTAEELQR LDKSSNNDQLIESSLFIDIANTSMILRDIRSDSDSLSRDDISDLLFKQYR QAGKMQTYFLYKALRDQGYVLSPGGRFGGKFIAYPGDPLRFHSHLTIQDA IDYHNEPIDLISMISGARLGTTVKKLWVIGGVAEETKETHFFSIEWAGFG 50 100 150 200 250 300 D Sen15p MKDIINCRREITSRRLRKIRi.8._T_T_l)_I_I__S_L_V_K_NNLLYFQMWTEVEILQDDL 50 SWKGNSLRLLRGRPPHKLSNDVDTEHENSLSSPRPLEFILPINMSQYKEN 100 FLTLECLSQTFTHLCSPSTERILLAIINDDGTIVYYFVYKGV KPK N 144 Figure II-3 58 14.8kD, exactly the size determined by SDS-PAGE. We have renamed this gene SEN15. Further evidence described below suggests that this protein is a subunit of endonuclease. Chromosomal disruptions of each of the subunit genes are lethal in haploid yeast cells, indicating that all of their gene products are essential for vegetative growth, and employing an essential function for each in the tRNA splicing process (Figure II-4A-D). Analysis of the tRNA splicing endonuclease subunit sequences During all purification procedures endonuclease behaves as a membrane-bound protein. We therefore searched for predicted membrane-spanning sequences in each of subunits. Figure II-5 A-D shows hydropathy plots for each of the four subunits. The most probable transmembrane sequence is found in Sen2p, the 44kD subunit, between residues 225 and 243 (see Figure II-3B). Less convincing but still potential transmembrane sequences are found in Sen54p (residues 226 to 247) and Sen34p (residues 42 to 60). Sen2p contains a domain predicted to form a canonical anti-parallel coiled-coil structure (see Figure II-5E). This structure can form distinct protrusions from the surface of a protein, as seen in the E. coli seryl-tRNA synthetase (Cusack et al., 1996), and may mediate protein-protein or RNA-protein interactions. All four subunits contain probable basic and bipartite nuclear localization sequences of the type described by Kalderon et al., (1984) and Robbins et al., (1991) (Figure II-3 A-D shaded boxes) . Sen2p and Sen54p also contain potential nuclear export signals (NES) similar to those originally described by Taylor and co-workers for export of the PKI protein and Ltihrmann and co-workers for export of the HIV Rev-1 protein (Fischer et al., 1995; Wen et al., 1995) (Figure II-3A,B, open boxes). When fused to a 59 FIGURE II-4. All Four Subunits of Endonuclease are Essential (A-D) Diploid strains containing single locus deletion of the ORFs of the various endonuclease subunits were sporulated and tetrads were dissected on YPD. Image was taken after 4 days growth at 30°C. All spores were examined for germination and only those in which all four spores germinated are depicted. 60 A SEN54 B SEN2 • • • -• • • C SEN34 D Figure II-4 • SEN15 • I 61 FIGURE II-5. Sequence Analysis of the Subunits of Endouclease (A-D) Hydropathy plots of each subunit of endonuclease using the Kyte and Doolittle algorithm and a window of 19 amino acids. (E) Coiled coil prediction for Sen2p utilizing the Coils program (Lupas et al., 1991). 62 HAydrophilic~:~~1 Hydrophobic 3~ MwA I '·~11Jtu_,....,,Jilj,,..... - , . _ _ _ _ _ _ _ _ _ _ __ _ _ _ S_e_n_5_4~p -3..----------------------, B -2 -1 ~ l - d tMI■ l.Wc 111 t I' !.ti&n,r ~i..Ju .11.L .. J 2 3_,____ _ _ _ _ __ _ _ _ _ _ _ _ _Sen2p __ ~ _3..----- - - - - - - - - - -- - - - - - - , -2 C -1 o 1 2 Sen34p 3.,____ _ _ _ _ _ _ _ _ _ __ _ _ _ __ ~ _ 3 . . - - - - - - - - - - - - - - - - - -- - - - , -2 D -6 1 2 Sen15p 3 -'-r-r-T""T""T"rn--,-,--,-,--,-.-r,--m,---,-,~rr-r-T""T""T"rn--,-,--,-,--,-.-r,--m""T""T"T--,-,--,'--T' 0 50 100 150 200 250 300 350 400 450 500 Residue Number E 1.0 0.9 0.8 0.7 Probability 0.6 of Coiled-Coil 0.5 Formation 0.4 0.3 0.2 0.1 O 50 Sen2p 100 150 200 250 300 350 400 Residue Number Figure II-5 63 heterologous protein, NES's direct export of a protein from the nucleus to the cytoplasm (Fischer et al., 1995; Wen et al., 1995). SEN15 encodes a possible mitochondrial import presequence in the first 20 amino acids (Figure 11-3D italicized residues) (Attardi and Schatz, 1988; Kiebler et al., 1993). This feature is surprising and suggests the possibility that Sen15p may also localize to the mitochondria. All four genes are characterized by a low codon usage bias, in accordance with the very low abundance of this enzyme (Rauhut et al., 1990). Antibodies to the subunits of endonuclease To unambiguously determine that each of these genes encodes a subunit of endonuclease, we prepared polyclonal antisera specific for each of the four gene products expressed in E. coli. A Western blot experiment was performed in which purified endonuclease was fractionated by SDS-PAGE and individually probed with antisera specific for each of the four gene products. As shown in Figure 11-6, probing of endonuclease purified by the affinity procedure (lane A), and endonuclease purified by the Rauhut procedure (lane R), reveals that each of the antisera react specifically with the appropriate polypeptide. The antibody to Sen15p was produced to a histidine/FLAG tagged antigen and as can be seen, the polyclonal antisera also contains antibodies to the affinity tag present on Sen2p and an apparent degradation product of Sen2p (compare 2C a-Sen 15, lane A to 2C a-Sen2, lane A). Since the two preparations of pure endonuclease were prepared by different purification protocols, these results strongly support the conclusion that each of these four genes encodes a subunit of endonuclease. 64 FIGURE II-6. Western Blot of Pure Endonuclease Utilizing Antibodies to Endonuclease Subunits A. Silver stained gel of: Lane 1, 5µg of crude yeast extract; lane 2, ~0.5µg of affinity tagged endonuclease purified as in Figure II-2; lane 3, ~0.5µg of endonuclease purified by Rauhut et al.(1990); MW represents molecular weight standards (Novex) and are labeled to the left in kiloDaltons. Panels B, C, D, and E represent duplicate blots of gel from panel A probed with 1:500 dilution of antibodies to Sen54p, Sen2p, Sen34p, and Senl5p, respectively. 65 MW 200kD A 116kD 97kD 66kD 55kD 36kD -- R a-Sen54 A R a-Sen2 A R a-Sen34 A R -- - ---- 31kD 21kD • 14kD 6kD Figure II-6 a-Sen15 A R 66 Two-hybrid analysis of the four subunits of endonuclease To begin to understand the functional interactions of the endonuclease subunits, we have tested the four subunit genes in a two-hybrid matrix experiment. Each of the four genes was translationally linked to either a DNA-binding or transcriptional activation domain for analysis in the Fields two-hybrid system (Bartel and Fields, 1995; Fields and Song, 1989). Figure 11-7 shows the results of combinatorial expression of subunit gene pairs using rescue of histidine auxotrophy as an assay of interaction. Two interactions were detected in vivo: interaction between Sen2p and Sen54p, and between Sen34p and Senl5p. The reciprocal interaction between Sen2p and Sen54p did not activate expression, indicating that this interaction is sensitive to the orientation of the subunits relative to the binding domain/activation domains. Equivalent results were obtained with a j3-galactosidase reporter gene (data not shown). Thus we have observed interactions between Sen34p and Sen15p and between Sen2p and Sen54p, but the other possible pairwise interactions were not observed. Sen2p and Sen34p are members of a gene family of endonucleases A search of the Genbank database (ver. 97) for sequences similar to the four endonuclease subunit genes did not reveal any potential homologs. However, similarity was detected between the Sen2 and Sen34 proteins themselves. The similarity occurs in a region of approximately 130 amino acids (see below). A sequence profile of the region of similarity between these proteins was used to search the translated Genbank database and was found to match one sequence, a 205 amino acid ORF contained in a presumed intron of an HMG-like gene of Zea mays (Yanagisawa and Izui, 1993). C. Daniels and colleagues have determined the nucleotide sequence of the gene for the tRNA splicing 67 FIGURE II-7. Two-Hybrid Matrix Analysis of Endonuclease Subunits Activation of histidine reporter gene by various hybrid combinations of Gal4 binding domain and Gal4 activation domain to the subunits of endonuclease. Reporter strain was transformed and then duplicate colonies from each were picked to a new selective plate. This plate was then replica plated to plates lacking histidine and containing 5mM 3aminotriazole (to suppress background growth) to determine activation of the reporter gene. 68 GAL4 Activation SEN15 SEN34 SEN2 No Insert GAL4 DNA Binding Domain Fusion SEN15 SEN34 . SEN2 SEN54 Figure II- 7 Domain Fusion SEN54 No Insert 69 FIGURE 11-8. Sequence Alignment Between the Archaeal Endonucleases and Two Subunits of the Yeast Endo nuclease Identifies a Family of tRNA Splicing Endonucleases Sequence alignment between endonuclease identified by Daniels and co-workers from Haloferax volcanii and Methanobacterium thermoautotrophicum, endonuclease from Methanococcus jannaschii, putative endonuclease from Pyrobaculum aerophilum, putative endonuclease from Zea mays and two subunits, Sen34p and Sen2p, of the yeast tRNA splicing endonuclease. Sequence alignment was performed using the GCG program Pileup. 70 M. jannaschii 28RHYlNVEGNFISISII ' YINLGW ••• • 1VKYKDNKP LS FEE LYE 71 M. thermo. 14 S H Y . K MY E DR Q S I • Y ERGK . . • K L M • K D DD E V S P E E F I S 56 H. volcanii 261 R LY NA D S GP Q S L • Y RAD A . . I . . DE ADV LS R GR D VE 301 P. aerophilum 1 . . . . . . . . . . . . . . . . . . . . . EKGR • • . . KVMGEDGREVAPEELAA ~ mays 73 A VE 9A G A E S Q W F Q [!IG P E F F C HAL K C • . • RV E S N N KT R I GA P E L W G 118 S.c. Sen2p 207 D E N 9D L L P LE S [!E MP V • F T FA L P V . . . D I S P • • AC L AG K L F Q F D 250 145 A E E L Q R L DK S S N ND Q [!II L F I D I A NT S M I RD I R S D S D S L S R DD I S D 193 S.c. Sen34p z. i 1 117 M. jannaschii M. thermo. 12 57LL YARNVE . . GERGLYSK ERLCLK . •• . t ilr,V ir,vlD Do· I R R9Yij!JicT 9 Y ri!JicT1•11D1L,ERGANID I E• L .ERGGAPG100 H. volcanii 302 . . . . . G DR F DR R • • LA V A • AK T VP KS I S DI V TE F E S VD D 345 P. aerophilum 25 L GR E RM R NF DE I • • I y F ,. L 9Y V E s LI s V • E K Gp G I • 69 Z. mays 119LLASGSEQFPEM . . A Y ·LKNWV S QY D A H.HPALV164 S.c. Sen2p 251 A KY K D I H S F V R S . . I HY 1· S H 9w C S I [3F DY LL R • GP P F • 295 194 L LFKQY RQAGK MQT L A • Q9Y VL SP RF K!)I A . . . PG DP L 239 S.c. Sen34p M. jannaschii 118 K E I Y LIVF PEDS SF • • • • • • • • L LSE LT GFVIAH S I I K L L I AIV 158 M. thermo. 101RT YL RVISENDTI . . . . . . • . HALDFSSYV· AHG KLLMAFL140 H. volcanii 346LS. FL RVVAPDHTF . . . . . . . . VPRDLSLDV·LAGG RMVFAL.385 P. aerophilum 70 . D , P M V F L E P DK G I S . . • , . . • . AT D I T R G G • L S H S T WT L ATV 110 Z. mays 165 . . . F T VA P E GK A F G T R C ARM EV W S EV L CAL • S GS I T L L V LT I 211 S.c. Sen2p 296 . Q . E F C G L DH D V S K DY T W . . . . . . . . Y S S I A • VG GA K T F V L CY V 335 240 R F L T I . . . . . Q DA I DY H NE P I D L I S M I S GA • L GT T i!Jic L WV I G G V 281 S.c. Sen34p Figure II-8 71 endonuclease of the Archaea Haloferax volcanii and Methanobacterium thermoautotrophicum (Kleman-Leyer et al., 1997). Homologs of this gene occur in other Archaeon, Pyrobaculum aerophilum and Methanococcus jannaschii (Bult et al., 1996). In order to determine if the M. jannaschii homo log is a functional homolog of the H. volcanii endonuclease, the gene product was expressed in E. coli, purified and shown to cleave intron containing archaeal pre-tRNAs (CRT, unpublished). Interestingly, the endonucleases of the Archaea align with the yeast endonuclease subunits, Sen2p and Sen34p, and the ORF of Z. mays (Figure 11-8). In all, with sequences for seven members of this gene family available, a convincing homologous core domain of ~50 amino acids is now apparent with lesser regions of homology extending over~ 130 amino acids. Two active sites in endonuclease The sequence similarity between Sen2p and Sen34p suggests that these proteins have a similar function. It seems that this function is very likely to be the catalysis of tRNA cleavage, since these proteins belong to a family which includes the homodirneric tRNA splicing endonuclease of H. volcanii. Interestingly, members of the endonuclease family found in the genomes of M. thermoautotrophicum and M. jannaschii are both small proteins made up entirely of the sequence domain found in the yeast and H. volcanii endonucleases. This further supports the notion that the sequence alignment reveals the catalytic core of the endonuclease enzyme. Previous results have indicated that Sen2p carries the active site for 5' splice site cleavage. Ho et al, 1990, showed that the sen2-3 mutation results in endoncuclease that is defective in cleavage only at the 5' splice site of tRNA precursors in vivo (Ho et al., 1990). As reported here, the glycine to glutamate change in this allele is located at position 292, within the domain conserved between the 72 eucaryal and archaeal endonucleases. This residue itself is not conserved in the endonuclease family, as perhaps expected for a conditional lethal mutation. Given that Sen34p is homologous to Sen2p and to the archaeal endonucleases we predicted that the Sen34p contains the active site for 3' splice site cleavage. This prediction led to the model shown in Figure 11-9 in which the 5' splice site is cleaved by Sen2p and the 3' splice site by Sen34p. This model clearly predicts that endonuclease containing a mutant Sen34p subunit should fail to cleave the 3' splice site of pre-tRNA while retaining full activity in cleavage of the 5' splice site. To test this prediction we changed the conserved histidine at position 242 of SEN34 to an alanine residue (H242A) . Extrachromosomal expression of a HIS/FLAG tagged mutant Sen34p H242A in wild type haploid cells had a dominant deleterious effect, increasing the doubling time from 2.5 hours to 6 hours in minimal media. Extrachromosomal expression of wild type Sen34p in the same strain had little or no effect on growth. Endo nuclease was purified from each of these strains by the affinity purification method described in the Experimental procedures to select only endonuclease complexes containing the tagged Sen34p subunit. tRNA splicing assays were performed with endonuclease containing wild type Sen34p and Sen34p H242A. Endonuclease containing affinity tagged wild type Sen34p cleaves pre-tRNAPhe normally indicating that the affinity tagged enzyme functions as wild type endonuclease (Figure 11-10). At the low enzyme concentrations used in this experiment, a slight accumulation of a 2/3 molecule consisting of the intron joined to the 3' exon, derived from cleavage at the 5' splice site only, is normally seen (Miao and Abelson, 1993). In contrast, endonuclease containing Sen34p H242A shows a marked accumulation of 5' exon and the intron-3 ' exon 2/3 molecule. There is a nominal amount of 3' exon and intron product, but we cannot 73 FIGURE II-9. Two Active Site Model for Cleavage of Pre-tRNA by Yeast tRNA Splicing Endonuclease Yeast pre-tRN APhe tertiary structure is depicted with a schematic diagram of the yeast endonuclease. In this model Sen2p and Sen34p each contain an active site for cleavage of the 5' and 3' splice sites, respectively. The Sen2p-Sen54p and Sen34p-Sen15p interaction, detected in the two-hybrid analysis, is depicted as well as the potential tRNA recognition mediated by Sen54p (see text for details). 74 en54 Figure II-9 75 FIGURE II-10. tRNA Splicing Activity of Sen34p H242A Containing Endonuclease Yeast pre-tRNAPhe was incubated without endonuclease (pre) or with endonuclease purified from HIS/FLAG tagged Sen2p (WT), HIS/FLAG tagged wild type Sen34p (Sen34p) or HIS/FLAG tagged Sen34p H242A (Sen34p H242A) expressing yeast. PretRNA was incubated for 5, 10 and 15 minutes and the resulting cleavage products (shown to the right) were fractionated on a 10% polyacrylamide gel. The amount of endonuclease used in each time course was similar based on silver staining (WT enzyme was used at lOOX concentration to allow for complete digestion of pre-tRNA). 76 pre WT Sen34p Sen34p H242A I 5 10 15 1 I 5 10 15 I -pre-tRNA -lntron-3'exon -3'exon -S'exon -lntron Figure II-10 77 determine if this is derived from retention of some 3' splice site cleavage activity in Sen34p H242A or from the presence of some wild type endonuclease in the purified mutant extract. However, the assay clearly demonstrates that endonuclease containing Sen34p H242A is, as predicted by the model, impaired in 3' splice site cleavage. Discussion Yeast tRNA endonuclease contains two active sites The sequence similarity between Sen2p and Sen34p, and their homology to the archaeal tRNA splicing endonucleases, implied that each of these two subunits contains an active site for tRNA cleavage. As described above, endonuclease containing Sen34p H242A appears to be significantly impaired for 3' splice site cleavage, but 5' splice site cleavage occurs normally. A complete blockage of 5' splice site cleavage was seen in endonuclease containing sen2-3 (Ho et al., 1990). These results taken together provide strong support for the model in Figure II-9 indicating that the yeast tRNA splicing endonuclease contains two distinct, functionally independent active sites. Sen2p contains the active site for cleavage at the 5' splice site and Sen34p contains the active site for cleavage at the 3' splice site. We are currently undertaking a more extensive mutagenesis of SEN34 and SEN2 in order to generalize these results. Additional support for the model shown in Figure II-9 is derived from previous work on the Xenopus endonuclease where it has been shown that the two cleavage sites in yeast tRNA precursors have different properties. The Xenopus tRNA splicing endonuclease recognition of the 3' splice site requires a base-pair between a purine located 3 bases upstream of the 3' splice site and the conserved pyrimidine at position 31 in the anticodon loop (the A-I basepair) (Baldi et al., 1992). Cleavage at the 5' splice site 78 requires only a purine as the adjacent 5' nucleotide. It is conceivable that different specificities in cleavage could be obtained by a conformational change in a single active site occurring between cleavage reactions. For this to occur one might expect an obligatory reaction pathway, i.e., cleavage at a primary site followed by cleavage at the second site. However, we have shown that the yeast endonuclease can and does cleave the 5' and 3' splice sites in a random order (Miao and Abelson, 1993). More recently, the accompanying paper by Tocchini-Valentini's laboratory has shown that accurate 3' splice site cleavage does not require the mature tRNA domain of the pre-tRNA (Di Nicola Negri et al., 1997). A small substrate containing only an intron and anticodon stem is cleaved by the Xenopus endonuclease at the 3' splice site but not at the 5' splice site. While both splice sites in an intact pre-tRNA are subject to the measuring mechanism, it appears that only the 5' splice site absolutely requires this mechanism. The H. volcanii tRNA splicing endonuclease cleaves with the same general mechanism as the eucaryal endonuclease, producing 2',3' cyclic PO4 and 5'-OH termini, but has a different splice site recognition mechanism which fails to cleave yeast pre-tRNAs (Thompson and Daniels, 1990; Thompson and Daniels, 1988). The mature tRNA domain is not a requirement for the archaeal endonuclease. Archaeal tRNA splice sites are located in bulged regions separated by four helical base-pairs (Fig. II-lC). In the accompanying paper by Kleman-Leyer et al., the H. volcanii endonuclease is shown to exist as a homodirner, and thus contains two active sites. The view that each of these active sites is functional is strengthened by the evidence presented here for two distinct active sites in the yeast endonuclease. However, because of the near symmetry of the archaeal splice site structure, we cannot rule out that a single active site could act interchangeably. 79 Interestingly, most archaeal pre-tRNAs appear to contain the conserved A-I basepair of eucaryotic pre-tRNAs in the context of the four basepair stem separating the bulged loops of the splice sites. In this regard, the requirements for 3' splice cleavage are more similar between Archaea and Eucarya than 5' splice site cleavage. In both cases the splice site must be in a single strand bulge near a base-paired element (in the eucaryotic pre-tRNAs it is the A-I basepair). Subunit interactions in endonuclease In a complete matrix of two-hybrid interactions between the four subunits of endonuclease, strong interactions were seen only between the Sen2p and Sen54p pair and the Sen34p and Senl5p pair. Since all four subunits remain tightly associated through two different purification regimens, it seems likely that there are also interactions between the two presumed heterodimers. These interactions, however, would not be detected in a two-hybrid experiment if they exist only within the context of the intact enzyme, or if three of the subunits are required for the interaction. The involvement of each of the active site subunits in a specific heterodimeric interaction suggests that the different specificities of the two active sites may be determined by these interactions. For example the measuring mechanism, seen primarily in the recognition of the 5' splice site, may be determined through the interaction between Sen2p and Sen54p as indicated in Figure II-9. Sen54p is a very basic protein and as such might be expected to interact with the phosphate backbone of the pre-tRNA substrate, presumably in the mature tRNA domain. Sen15p could be primarily responsible for the recognition of the A-I basepair. 80 Evolutionary implications of the tRNA splicing endonuclease gene family The existence of homologous tRNA splicing endonuclease genes in Eucarya and Archaea suggest that this enzyme was present in their common ancestor. Since endonuclease participates in tRNA splicing in both Eucarya and Archaea, it seems likely that the original function of this enzyme was the same. The discussion of the time of appearance of introns in evolution has been a lively one and it is safe to say that a consensus has not emerged. The simplest hypothesis consistent with the data presented here and in the accompanying papers is that the function of endonuclease in the common eucaryal-archaeal ancestor was the removal of introns from tRNA. A less parsimonius hypothesis is that the tRNA splicing function of this enzyme evolved independently in both kingdoms from an earlier, more generalized RNA processing function. In some present day Archaea the tRNA splicing endonuclease can apparently also function to remove introns from ribosomal RNA (Burggraf et al., 1993; Kjems and Garrett, 1985; Kjems and Garrett, 1991). Recently, Walter and co-workers have detected an intron in a yeast transcription factor gene that is not processed by the normal mRNA spliceosomal machinery (Cox and Walter, 1996; Sidrauski et al., 1996). They have demonstrated that tRNA ligase is involved in the splicing of this intron, but the tRNA splicing endonuclease has not as of yet been implicated. Clearly, members of this family of proteins have a number of functions, but it is not yet possible to determine which are original and which have been co-opted later in evolution. Although the endonuclease active site is apparently a very primitive protein domain, it is clear that its substrate specificity has been altered in evolution because the recognition of the splice sites is different in Eucarya and Archaea. We speculate that the peculiarities of the eucaryal system are in part explained by the heterodimeric interaction 81 of the divergent subunits-Sen2p with Sen54p and Sen34p with Senl5p. Many fascinating questions remain concerning this evolutionary specialization. Since Sen54p and Sen15p have no known homologs, where did these proteins come from? Do Sen54p and Senl5p play a role in splice site selection? Can the active site subunits function in conjunction with other protein subunits, thereby modifying endonuclease specificity for other RNA processing reactions? We are now able to begin to address these questions. A further complexity of the evolution of the tRNA splicing process is indicated by the lack of a yeast tRNA ligase (or T4 RNA ligase) homolog in the entire M. jannaschii genome (Bult et al., 1996; CRT and EAA, unpublished data). Interestingly, several archaeal genomes contain a clear homolog of a bacterial RNA ligase which can form a 2'5' linkage when joining tRNA half-molecules derived from endonucleolytic cleavage of yeast pre-(Arn and Abelson, 1996; Greer et al., 1983). The complete ligation mechanism for archaeal tRNA splicing has not yet been determined, but it has been demonstrated that an RNA splicing ligase activity present in H. volcanii extracts does not require ATP or exogenous phosphate (Gomes and Gupta, 1997). The E. coli 2'-5' RNA ligase is the only known RNA ligase activity which does not require a nucleotide triphosphate cofactor (Arn and Abelson, 1998). We are presently attempting to ascertain whether the archaeal 2'-5' RNA ligase homolog functions in tRNA splicing in vivo. Experimental Procedures Yeast Methods Yeast were transformed by the lithium acetate procedure (Gietz and Schiestl, 1995). Yeast minimal media (SD) contained Bacto yeast nitrogen base without amino acids (YNB - aa, 6.7g/1), dextrose (2%w/v) and a supplemental amino acid mix (BiolOl) 82 lacking various amino acids for auxotroph selection. All other yeast media was prepared according to Sherman et al., 1986. For fermentation growth of CRT44, SD:-histidinetryptophan, prepared as described above, was used and supplemented with 25ug/ml Kanamycin. The fermentation was fed using SD+ lOX-histidine-tryptophan amino acid mix with constant flow for 24 hours. Two-hybrid strains were obtained by a generous gift from P. Bartel of S. Fields laboratory and are described in Table I. All two hybrid techniques used are described by Hannon et al. (1995). Plates used for analysis contained 5rnM 3-aminotriazole which suppressed background growth associated with histidine reporter assay (Hannon and Bartel, 1995). Strains E.coli strains utilized - HB 101. XLl-Blue (Novagen). DH5 (Gibco-BRL). Saccharomyces cerevisiae strains are described in Table I (Appendix II). Knockout strains YPH274~SEN2 and CRT54 were created utilizing methods described in Guthrie and Fink, 1992. Knockout strains of SEN15 (CRT15) and SEN34 (CRT34) were created utilizing methods and plasmids generously obtained from A Wach and described by Wach et al. (1996). All knockouts were confirmed by PCR analysis and each knockout was complemented with the corresponding ORF only. Strain CRT44 was prepared by transforming YPH274~SEN2pC10 with YpH8/FLAG SEN2, plating on selective media (SD-his-trp-ura), and streaking out isolates on selective media (SD-his-trp) + 5 FOA to select for loss of pClO (URA3 CEN/ARS). 83 Plasmid Preparation All restriction enzymes used were obtained from New England Bio-Labs and Boehringer Mannheim and used in accordance with the manufacturer' s instructions. Restriction digests, DNA ligations and electrophoresis of DNA were performed by standard methods of Sambrook et al. (1990). pClO was prepared by subcloning a EcoRV/EcoRI fragment of DNA derived from the YcP50 sen2-3 complementing clone, into the vector pRS416 (URA3)(Sikorski and Hieter, 1989). This fragment contains an open reading frame coding for the 377 amino acid SEN2 gene plus 250 base pairs in the promoter region and 345 base pairs of downstream region. pCl0 was capable of complementing a knockout of the SEN2 ORF. YpH8/FLAG was prepared by PCR of the CUPJ promoter, from JD51, to allow cloning of the CUP1 promoter into the HindIII/EcoRI site of pBluescript (Stratgene) . This plasmid was then digested with Xbal, filled-in and cut with Pstl and the Pstl-Nael fragment from pGBT9 (from S. Fields), containing the ADHJ terminator sequence, was cloned into this site, yielding pBSCUPffERM. This plasmid was then cut with Hindlll/Sacl and the resulting fragment was subcloned into pRS424 (Sikorski and Hieter, 1989), to yield YpCUPffERM. This plasmid was cut with Ndel and EcoRI and the HIS/FLAG linker was cloned into this site, generating YpH8/FLAG. YpH8/FLAG SEN2 was prepared by digestion of pGBT9-SEN2 with EcoRI and Pstl, gel isolation of the fragment containing SEN2 and subcloning into EcoRI/PstI digested YpH8/FLAG. YpH8FLAG SEN34 was prepared by digestion of pGBT9-SEN34 with EcoRI and Pstl, gel isolation of the fragment containing SEN34 and subcloning into EcoRI/Pstl digested YpH8/FLAG. Oligonucleotide directed mutagenesis was performed on single stranded 84 YpH8/FLAG-SEN34 to effect an H242A mutation according to the protocol provided by the manufacturer (Amersham). Two-hybrid vectors were prepared by cloning PCR products of SENl5 (oligos SEN15-N 5' -GGAAITCGCAACGACAGATATCATATC-3' and -C 5'TTCTGCAGATCTTCAATITCTTITCGGTTITCG- 3'), SEN54 (oligos SEN54-N 5'GGAATTCCAATTCGCTGGGAAG-3' and -C 5'TTCTGCAGTTAATGCTTCTTCCATCTTTG-3'), SEN2 (oligos SEN2-N 5'GGAATTCATGTCTAAAGGGAGGG and -C 5' CGGGATCCCAATGAAAAGTITAGAGC) and SEN34 (oligos SEN34-N 5'CGGAATTCGGGATCCATATGCCACCGCTAGTATITGAC-3 ' and-C 5' CGGGATCCCGGGAAGCTTAACCAAATCCAGCCC-3') digested with EcoRI and BamHI and cloned into EcoRI/BamHI digested pGBT9 and pGAD424 (Hannon and Bartel, 1995) (these constructs were sequenced). Antibodies and Western Blots Polyclonal antibodies to Sen34p were produced by injection of rabbit (Cocalico Biosciences) with peptides NH2-DKTT AEELQRLDKSSC-COOH, and NH2DIRSDSDSLSRDDIC-COOH. Polyclonal antibodies to Sen2p were produced by injection of rabbits with peptide derived from the C-terminus of Sen2p (NH2MSKGRVNQKRYPLC-COOH) . All peptides were produced by the Caltech Peptide Facility. The antibodies were used directly in rabbit serum at the indicated dilution. Polyclonal antibodies to both Sen54p and Sen15p were prepared in the following manner: Antigen was derived from purification of HISS/FLAG tagged protein from E.coli. Each gene was cloned into the EcoRI/PstI site of YpH8/FLAG by subcloning from pGBT9 85 SEN54 and pGBT9 SEN15. Each construct was then cut with Ndel/Pstl and subcloned into Petl la. Purification was carried out under denaturing conditions by Ni-NTA chromotography as suggested by Qiagen. Protein eluted from the Ni-NTA resin was gel purified by Prep-Cell Chromatography (Biorad) as per the manufacturers suggested protocol with elution into buffer containing 0.05% SDS. The eluate was directly injected into Swiss Webster mice as per the protocol of Susan Ou of the Caltech Monoclonal Antibody Facility. SDS-PAGE was performed on 10%-20% gradient gels (Biorad). Proteins were transferred to PVDF (lmmobilon) by semi-dry transfer (Biorad). Western blots were performed as described by Harlow and Lane (1988) and were visualized by enhanced chemiluminescence (ECL, Amersham). Endonuclease Assay Endonuclease was assayed by incubation of 2µ1 of extract for specified amount of time at 30°C in 10µ1 reaction containing 20mM Na-HEPES pH 7.5, 5mM MgC12, 2.4mM spermidine-HCl pH 7.5, 0.lmM DTT, 0.4% Triton X-100 and 10 frnole pre-tRNAPhe body-labeled with [a-32p]UTP (specific activity ~500 cpm per femtomole tRNA) by T7 polymerase transcription according to Sampson and Saks (1993). RNA products were phenol/chloroform extracted, precipitated and fractionated on a 10% polyacrylamide gel containing 8M urea. Gels were exposed to phosphorimager for 8 hours. 86 Purification of tRNA Endonuclease Purification of endonuclease was carried out utilizing three separate purification strategies. The first is described by Rauhut et al. (1990). This protein was used to obtain the gene for the 34kD subunit. Affinity purification was carried out utilizing yeast strain CRT44 grown by fermentation at 30°C. The cells were harvested and stored at -70°C. All subsequent steps were performed at 4 °C. 150g of yeast cells resuspended in 1.5X breaking buffer (37 .5mM Tris pH8.0, 15%(v/v) Glycerol, 5mM B-ME, lmM EDTA, 0.1% (v/v) Triton X-100, 405 mM NH4SO4) to 300mls was ground in a Bead-Beater (Bio-Spec Corp) for 20xl minute intervals and Triton X-100 (Boehringer-Mannheim) was added to 1%. The extract was incubated for 1 hour and then centrifuged (40,000 rpm X 1.5 h). The supernatant was loaded onto a 300ml DEAE-sepharose CL-6B equilibrated in IX breaking buffer+ 1% Triton X-100. DEAE effluent was diluted approximately three-fold and loaded onto a 100ml Heparin-HyperD (BioSepra) column equilibrated in buffer A (25mM MOPS pH7.1, 10% Glycerol, lmM EDTA, 5mM BME, 0.9% Triton X-100, lO0mM NaCl), at lml/rninute. The column was washed with 500ml buffer B (25mM Tris pH8.0, 10% Glycerol, SmM BME, 0.3% Triton X-100, 150mM NaCl). A 2X200ml gradient of 0.10.SmM NaCl in buffer B was applied. Active fractions were pooled and bound in batch to 5ml Ni-NTA agarose (Quiagen, equilibrated in buffer B) for I.Sh. The suspension was poured into a 2 X 10 mm column and washed extensively with Smls of each of the following buffers: buffer N (25mM Tris-HCl ph 7.5, 150mM NaCl, 10% Glycerol, SmM B-mercaptoethanol, 0.3% Triton X-100)+ 650mM NaCl+ 15mM imidazole; buffer N + 650mM NaCl+ 20mM imidazole; buffer N + 20mM imidazole; buffer N + 40 mM imidazole. Endonuclease was eluted in Buffer N + 150mM imidazole. The eluate was 87 loaded onto a 1ml M2 monoclonal anti-FLAG antibody column (Kodak) equilibrated in Buffer N. The column was washed in buffer N and eluted with Buffer N + 150ug/rnl FLAG peptide. This endonuclease was used to prepare the 54 kD subunit. A modified purification protocol was developed to prepare the 15 kD subunit and purification of mutant containing endonuclease. Yeast cells (strain CRT44 or as indicated in the text) were resuspended and broken as above. The extract was then spun at 40,000 rpm X 2h to pellet the nuclear membranes. The pellet was resuspended in 50ml of Buffer R (50rnM Tris, pH7.5, 150rnM NaCl, 5% Glycerol, lrnM EDTA, 5mM BME, 0.9% Triton X-100) and incubated for 1 h. The extract was subjected to centrifugation as before and the supernatant was loaded directly onto a 2.5ml M2 monoclonal anti-FLAG antibody column. The column was washed with buffer R and endonuclease was eluted with Buffer S (30rnM Tris, pH8.0, 500rnM NaCl, 10% Glycerol, 5mM BME, 0.3% Triton X-100, 5rnM Imidazole) containing 150ug/ml FLAG peptide. The elution was incubated with 1ml of Ni-NT A agarose, equilibrated in Buffer S, for 1.5h and poured into a 3ml column. The column was washed with Buffer S + 40rnM imidazole and endonuclease was eluted in Buffer S + 150rnM imidazole. Protein Sequencing Three separate procedures were utilized for sequencing the subunits of endonuclease. In all three cases, protein was submitted as a gel slice with various amounts of protein in each. For the 34kD subunit, ~25ug of protein was submitted to the protein sequencing facility at Cold Spring Harbor and was digested in gel with lysly-endoprotease followed by extraction and HPLC purification of the resulting peptides following the 88 protocols of the facility (Dan Marshak, pers.comm.). Fractions containing single peaks were sequenced by Edman degradation. For the 54kD subunit, -5ug of protein was digested in gel with trypsin as described by Hellman et al. (1995) (Hellman et al., 1995), except that 0.3ug of Promega modified pig trypsin and speedvac drying of the gel slice was used. Peptides were HPLC purified and fractions containing single mass peaks, as determined by mass spectrometry on a Voyager Mass Spec, were submitted for Edman Degradation sequencing on an Applied Biosystems sequencer (Gary Hathaway, pers.comm.). For the 15kD subunit, -5ug of protein was submitted to the protein sequencing facility at the City of Hope (Duarte, CA) for in gel trypsin digestion followed by peptide sequencing utilizing a Beckman LC/MS-MS protein sequencer (K. Swiderek, pers. comm). Data obtained was used to search the yeast genome database (YGD). Acknowledgements We wish to acknowledge first and foremost the protein sequencing work done by Gary Hathaway and Dirk Krapf at Caltech Protein Analysis Laboratory, Dr. Terry Lee and Dr. Kristine Swiderek at the City of Hope Protein Sequencing Facility and Dr. Dan Marshak at the Cold Spring Harbor Protein Sequencing Facility. Special thanks to Paul Bartel and Stan Fields for two-hybrid strains and plasmids. We also want to thank Chuck Daniels for communicating the endonuclease sequences of Haloferax volcanii and Methanobacterium therrrwautotrophicum prior to publication and Ors. Mel Simon and Ung-Jin Kirn for sequence of the Pyrobaculum aerophilum homo log. We wish to thank Drs. Christine Guthrie, Eric Phizicky, Chris Greer and Anita Hopper for their insight and critical review of the manuscript and members of the Abelson laboratory for their advice 89 and support. This work was supported by generous grant from the American Cancer Society. References Abelson, J. (1991). RNA splicing in yeast. Harvey Lecture 85, 1-42. Arn, E. A, and Abelson, J. (1998). RNA ligases: function, mechanism, and sequence conservation. In RNA structure and function (Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press), pp. 695-726. Arn, E. A, and Abelson, J. N. (1996). The 2'-5' ligase from Escherichia coli: purification, cloning and genomic disruption. J. Biol. Chem. 271, 31145-31153. Attardi, G., and Schatz, G. (1988). Biogenesis of mitochondria. Ann. Rev. Cell Biol. 4, 289-333. Baldi, M. I., Mattoccia, E., Bufardeci, E., Fabbri, S., and Tocchini-Valentini, G. P. (1992). Participation of the intron in the reaction catalyzed by the Xenopus tRNA splicing endonuclease. Science 255, 1404-1408. Bartel, P. L., and Fields, S. (1995). Analyzing protein-protein interactions using 2-hybrid system. Meth. Enz. 254, 241-263. Belford, H. G., Westaway, S. K., Abelson, J., and Greer, C. L. (1993). Multiple nucleotide cofactor use by yeast ligase in tRNA splicing: evidence for independent ATPand GTP- binding sites. J. Biol. Chem. 268, 2444-2450. Biniszkiewicz, D., Cesnaviciene, E., and Shub, D. A (1994). Self-splicing group-I intron in cyanobacterial initiator methionine tRNA: evidence for lateral transfer of introns in bacteria. EMBO Journal 13, 4629-4635. Bult, C. J. , White, 0., Olsen, G. J., Zhou, L. X., Fleischmann, R. D. , Sutton, G. G., Blake, J. A, Fitzgerald, L. M., Clayton, R. A, Gocayne, J. D., Kerlavage, A R., Dougherty, B. A, Tomb, J. F., Adams, M. D., Reich, C. I., Overbeek, R., Kirkness, E. F. , Weinstock, K. G., Merrick, J.M., Glodek, A, Scott, J. L., Geoghagen, N. S. M., Weidman, J. F., Fuhrmann, J. L., Nguyen, D., Utterback, T. R., Kelley, J.M., Peterson, J. D., Sadow, P. W., Hanna, M. C., Cotton, M. D., Roberts, K. M., Hurst, M. A, Kaine, B. P., Borodovsky, M., Klenk, H.P. , Fraser, C. M., Smith, H. 0., Woese, C. R., and Venter, J. C. (1996). Complete genome sequence of the methanogenic archaeon, Methanococcus jannaschii. Science 273, 1058-1073. 90 Burggraf, S., Larsen, N., Woese, C.R., and Stetter, K. 0. (1993). An intron within the 16S rRNA gene of the archaeon Pyrobaculum aerophilum. Proceedings of the National Academy of Sciences 90, 2547-2550. Cech, T. (1993). Structure and mechanism of the large catalytic RNAs: group I and group II introns and ribonuclease P. In RNA World, R. F. Gesteland and J. F. Atkins, eds. (Cold Spring Harbor: Cold Spring Harbor Laboratory Press), pp. 206-208. Cox, J. S. , and Walter, P. (1996). A novel mechanism for regulating activity of a transcription factor that controls the unfolded protein response. Cell 86, 391-404. Culbertson, M. R., and Winey, M. (1989). Split tRNA genes and their products: a paradigm for the study of cell-function and evolution. Yeast 5, 405-427. Culver, G. M., McCraith, S. M., Zillmann, M., Kierzek, R., Michaud, N., Lareau, R. D. , Turner, D. H. , and Phizicky, E. M. (1993). An NAD derivative produced during tRNA splicing: ADP-ribose l "-2" cyclic phosphate. Science 261, 206-208. Cusack, S., Yaremchuk, A., and Tukalo, M. (1996). The crystal structure of the ternary complex of Thermus thermophilus seryl-tRNA synthetase with tRNAser and a seryladenylate analog reveals a conformational switch in the active site. Biological Chemistry Hoppe Seyler 377, 343-356. Di Nicola Negri, E., Fabbri, S., Bufardeci, E. , Baldi, M. I. , Gandini Attardi, D., Mattoccia, E., and Tocchini-Valentini, G. P. (1997). The eucaryal tRNA splicing endonuclease recognizes a tripartite set of RNA elements. Cell 89, 859-866. Fields, S., and Song, 0. K. (1989). A novel genetic system to detect protein- protein interactions. Nature 340, 245-246. Fischer, U. , Huber, J., Boelens, W. C., Mattaj, I. W. , and Luhrmann, R. (1995). The HIV1 rev activation domain is a nuclear export signal that accesses an export pathway used by specific cellular RNAs. Cell 82 , 475-483. Gietz, R. D., and Schiestl, R.H. (1995). Transforming yeast with DNA Methods In Molecular and Cellular Biology 5, 255-269. Gomes, I. , and Gupta, R. (1997). RNA splicing ligase activity in the archaeon Haloferax volcanii. Biochemical and Biophysical Review 237, 588-594. Greer, C. L. , Peebles, C. L., Gegenheimer, P., and Abelson, J. (1983). Mechanism of action of a yeast RNA ligase in tRNA splicing. Cell 32, 537-546. Guthrie, C., and Fink, G. R. (1991). Guide to yeast genetics and molecular biology (San Diego: Academic Press). 91 Hannon, G. , and Bartel, P. (1995). Identification of interacting proteins using the 2-hybrid system. Methods In Molecular and Cellular Biology 5, 289-297. Harlow, E. (1988). Antibodies: a laboratory manual (Cold Spring Harbor, NY: Cold Spring Harbor Laboratory). Hellman, U., Wernstedt, C., Gonez, J. , and Heldin, C.H. (1995). Improvement of an ingel digestion procedure for the rnicropreparation of internal protein fragments for amino acid sequencing. Analytical Biochemistry 224, 451-455. Ho, C. K. , Rauhut, R., Vijayraghavan, U., and Abelson, J. (1990). Accumulation of pretRNA splicing 2/3 intermediates in a Saccharomyces cerevisiae mutant. Embo Journal 9, 1245-1252. Kalderon, D., Roberts, B. L. , Richardson, W. D., and Smith, A E. (1984). A short aminoacid sequence able to specify nuclear location. Cell 39, 499-509. Kiebler, M., Becker, K., Pfanner, N., and Neupert, W. (1993). Mitochondrial protein import: specific recognition and membrane translocation of preproteins. Journal Of Membrane Biology 135, 191-207. Kjems, J. , and Garrett, R. A (1985). An intron in the 23S rRNA gene of the Archaebacterium Desulfurococcus mobilis. Nature 318, 675-677. Kjems, J. , and Garrett, R. A (1991). rRNA introns in Archaea and evidence for RNA conformational changes associated with splicing. Proceedings of the National Academy of Sciences 88, 439-443. Kleman-Leyer, K., Armbruster, D. A , and Daniels, C. J. (1997). Properties of H. volcanii tRNA intron endonuclease reveal a relationship between the archaeal and eucaryal tRNA intron processing systems. Cell 89, 839-848. Lee, M. C. , and Knapp, G. (1985) . tRNA splicing in Saccharomyces cerevisiae: secondary and tertiary structures of the substrates. J. Biol. Chem. 260, 3108-3115. Lupas, A , Vandyke, M., and Stock, J. (1991). Predicting coiled coils from protein sequences. Science 252, 1162-1164. Madhani, H. D., and Guthrie, C. (1994). Dynamic RNA-RNA interactions in the spliceosome. Ann. Rev. Genet. 28, 1-26. McCraith, S. M ., and Phizicky, E. M. (1991). An enzyme from Saccharomyces cerevisiae uses NAD+ to transfer the splice junction 2'-phosphate from ligated tRNA to an acceptor molecule. J. Biol. Chem. 266, 11986-11992. Miao, F. , and Abelson, J. (1993) . Yeast tRNA-splicing endonuclease cleaves precursor tRNA in a random pathway. J. Biol. Chem. 268, 672-677. 92 Moore, M., and Sharp, P. (1993). Splicing of precursor mRNAs by the spliceosome. In RNA World, R. F. Gesteland and J. F. Atkins, eds. (Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press), pp. 303-349. Ogden, R. C., Lee, M. C., and Knapp, G. (1984). Transfer RNA splicing In Saccharomyces cerevisiae: defming the substrates. Nucleic Acids Res. 12, 9367-9382. Palmer, J. R., Nieuwlandt, D. T., and Daniels, C. J. (1994). Expression of a yeast introncontaining tRNA in the Archaeon Haloferax volcanii. J. Bact. 176, 3820-3823. Peebles, C. L., Gegenheimer, P., and Abelson, J. (1983). Precise excision of intervening sequences from precursor tRNAs by a membrane-associated yeast endonuclease. Cell 32, 525-536. Phizicky, E. M., and Greer, C. L. (1993). Pre-tRNA splicing: variation on a theme or exception to the rule? Trends Biochem. Sci. 18, 31-34. Rauhut, R., Green, P.R., and Abelson, J. (1990). Yeast tRNA-splicing endonuclease is a heterotrimeric enzyme. J Biol. Chem. 265, 18180-18184. Reinhold-Hurek, B., and Shub, D. A (1992). Self-splicing intrans in tRNA genes of widely divergent bacteria. Nature 357, 173-176. Reyes, V. M., and Abelson, J. (1988). Substrate recognition and splice site determination in yeast tRNA splicing. Cell 55, 719-730. Robbins, J., Dilworth, S. M., Laskey, R. A, and Dingwall, C. (1991) . Two interdependent basic domains in nucleoplasmin nuclear targeting sequence: identification of a class of bipartite nuclear targeting sequence. Cell 64, 615-623. Sherman, F., Fink, G. R., and Hicks, J.B. (1986). Methods in yeast genetics (Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press). Sidrauski, C., Cox, J. S., and Walter, P. (1996). tRNA ligase is required for regulated mRNA splicing in the unfolded protein response. Cell 86, 405-413. Sikorski, R. S., and Hieter, P. (1989). A system of shuttle vectors and yeast host strains designed for efficient manipulation of DNA in Saccharomyces cerevisiae. Genetics 122, 19-27. Thompson, L. D. , Brandon, L. D., Nieuwlandt, D. T., and Daniels, C. J. (1989). Transfer RNA intron processing in the halophilic archaebacteria. Canadian Journal Of Microbiology 35, 36-42. 93 Thompson, L. D., and Daniels, C. J. (1988). A tRNAtrp intron endonuclease from Halobacterium volcanii: unique substrate recognition properties. J. Biol. Chem. 263, 17951-17959. Thompson, L. D., and Daniels, C. J. (1990) . Recognition of exon-intron boundaries by the Halobacterium volcanii tRNA intron endonuclease. J. Biol. Chem. 265, 18104-18111. Wach, A (1996). PCR-synthesis of marker cassettes with long flanking homology regions for gene disruptions in Saccharomyces cerevisiae. Yeast 12, 259-265. Wen, W. , Meinkoth, J. L., Tsien, R. Y. , and Taylor, S. S. (1995). Identification of a signal for rapid export of proteins from the nucleus. Cell 82, 463-473. Westaway, S. K., Belford, H. G., Apostol, B. L., Abelson, J. , and Greer, C. L. (1993). Novel activity of a yeast ligase deletion polypeptide: evidence for OTP-dependent tRNA splicing. J. Biol. Chem. 268, 2435-2443. Westaway, S. K., Phizicky, E. M., and Abelson, J. (1988). Structure and function of the yeast tRNA ligase gene. J. Biol. Chem. 263, 3171-3176. Winey, M. , and Culbertson, M. R. (1988). Mutations affecting the tRNA splicing endonuclease activity of Saccharomyces cerevisiae. Genetics 118, 49-63. Yanagisawa, S. , and Izui, K. (1993). Molecular cloning of 2 DNA-binding proteins of maize that are structurally different but interact with the same sequence motif. J. Biol. Chem. 268, 16028-16036. 94 Chapter ID - Three-Dimensional Structure and Evolution of a tRNA Splicing Enzyme (Hong Li, Christopher R. Trotta, John Abelson published in Science 280:279284, 1998; Appendix III; Chris Trotta's contributions include, identifying, cloning and purification of the M. jannaschii endonuclease, data collection from the crystals, extensive writing of the manuscript and intellectual contributions related to the understanding and generation of the models to describe the active site and enzyme architecture) Abstract The splicing of transfer RN A precursors is similar in Eucarya and Archaea. In both kingdoms an endonuclease recognizes the splice sites and releases the intron, but the mechanism of splice site recognition is different in each kingdom. The crystal structure of the endonuclease from the archaeon Methanococcus jannaschii was determined to a resolution of 2.3 angstroms. The structure indicates that the cleavage reaction is similar to that of ribonucleaseA and the arrangement of the active sites is conserved between the archaeal and eucaryal enzymes. These results suggest and evolutionary pathway for splice site recognition. Introduction Intrans are found in the tRNA genes of organisms in all three of the great lines of descent: the Eucarya, the Archaea and the Bacteria. In Bacteria tRNA intrans are selfsplicing group I intrans and the splicing mechanism is autocatalytic (1). In Eucarya tRNA intrans are small and they invariably interrupt the anticodon loop one base 3' to the anticodon. They are removed by the step-wise action of three enzymes: i) an endonuclease that cleaves the pre-tRNA, ii) a ligase that joins the exons together in a 95 reaction that requires the hydrolysis of both ATP and GTP (2), and iii) a phosphotransferase that transfers a 2'-phosphate left at the splice site to nicotinamide adenosine dinucleotide (NAD) acceptor (3). In Archaea the introns are also small and often reside in the same location as eucaryal tRNA introns. In some cases, however, they interrupt the tRNA gene at sites other than the anticodon (4). As in Eucarya, splicing in Archaea is also catalyzed by proteins. An endonuclease recognizes and cleaves the pretRNA It is virtually certain, however, that the mechanism of ligation is different from its counterpart in Eucarya as there is no obvious homolog of the eucaryal tRNA splicing ligase in the complete genome sequence of several members of the Archaea (5). Furthermore, the ligation mechanism in Haloferax volcanii extracts has been shown to proceed in an ATP-independent manner (6). The tRNA splicing endonucleases of both the Eucarya and the Archaea cleave the pre-tRNA substrate leaving 5'-hydroxyl and 2',3' cyclic phosphate termini, but these enzymes recognize their substrates differently. The eucaryal enzyme uses a measuring mechanism to determine the position of the universally positioned splice sites relative to the conserved domain of pre-tRNA (Figure III-1A)(7). In Archaea, where introns are not equivalently positioned, the enzyme recognizes a pseudo-symmetric substrate in which two bulged loops of three bases are separated by a stem of four base pairs (Figure III-IA) (8) . These observations suggested that the tRNA splicing mechanisms of the three major kingdoms evolved independently. Recently, however, characterization of eucaryal and archaeal endonucleases has provided new insights that reveal a common origin for this step in splicing. The yeast endonuclease is an a~yo heterotetramer with subunit molecular weights of 15kD, 34kD, 44kD and 54kD (9). The first archaeal endonuclease to be characterized, from Haloferax 96 F1GURE III-1. A. Consensus sequence and secondary structure of precursor tRNA substrate for yeast and archaeal endonucleases (modified from reference 9). Splice sites are indicated by short arrows. (0), (X) = unconserved bases in regions of conserved or variable secondary structure respectively, (Y) = pyrimidines, (R) = purines. The endonuclease in yeast is proposed to interact with the mature domain and measure the distance to the splice sites. In contrast, only a minimum region containing two three-base nucleotides separated by a four-base-paired helix (gray shaded region) is both necessary and sufficient for the splice site recognition by an archaeal endonuclease (16). B. Comparison of endonuclease models in Eucarya (heterotetramer) and Archaea (homodimer). 97 A A-OH C C 0-0 0-0 0-0 0-0 0-0 0-0 G G xx xXx UO-O OOOOG y OA C XOG I xx I I R O - OH 0 0-0 0-0 G-C R- Y A/ R C - GA R 11111 OOOOCT u C I pxov xxx 00-0 0-0 0-0 0-0 o-006 3' U-R C-G Y- R yp-o U X 0-0 8:8X R -+X xxx X R / RR-Y R-Y X c.' -' X X X X X Eucaryal substrate consensus Archaeal substrate consensus Eucaryal endonuclease Archaeal endonuclease B Fi g ure 111-1 98 volcanii, is a dimer of identical 37kD subunits (10). The H. volcanii enzyme was found to be homologous to both the 34kD and 44kD subunits of the yeast enzyme, suggesting that these distinct subunits contain active sites for tRNA splicing. This hypothesis was strongly supported by the finding that a mutant in the 44kD subunit, sen2-3, is selectively defective in 5' splice site cleavage (11) while a mutant in the 34kD subunit, H242A, is selectively defective in 3'-splice site cleavage (9). These results suggested related models for tRNA splice site recognition and cleavage in the two kingdoms (Figure III-1B). In yeast, the 34kD subunit recognizes the 3'-splice site and interacts strongly with the 15kD subunit (as seen in a yeast two-hybrid experiment) (9). The 44kD subunit recognizes the 5'-splice site and interacts with the 54kD subunit which, we have suggested, may embody the molecular ruler (9). In H. volcanii, the homodimeric enzyme is proposed to interact with the pseudo-symmetric substrate so that each splice site is cleaved by a symmetrically disposed active site monomer. Thus it is likely that the spatial disposition of the active site subunits has been conserved in evolution and in the eucaryal lineage a distinct mechanism of splice site recognition has evolved. There are apparent advantages in choosing the simpler archaeal system to begin to understand the enzymology of this conserved enzyme family. The euryarcheota group of Archaea which includes Methanococcus jannaschii contains genes that are homologous to the H. volcannii enzyme but whose products are about half the size as that of H. volcanii (179 amino acids in the case of M. jannaschii). M. jannaschii is a thermophilic organism collected at a thermal vent on the East Pacific Rise at a depth of 2600 meters. It can withstand 200 atmosphere of pressure and grows at an optimum temperature of about 85°C. Its complete genome sequence has been determined by the TIGR group (12). We believed that a high resolution structure of the simpler archaeal enzyme would shed light 99 on the mechanism of the more complicated but related eucaryal endonuclease and thus embarked on a structural study of the M. jannaschii endonuclease. We report here the structure of the M. jannaschii endonuclease determined by X-ray diffraction and refined to a resolution of 2.3A. Structure Detennination The M. jannaschii endonuclease gene with an 8xHis and Flag tag sequences (9) at its 5' end was cloned by PCR from genomic DNA into the pETl la vector. Soluble protein was obtained by overexpression in E coli (13). The structure was determined by the multiwavelength anomalous diffraction (MAD) method (14) on crystals derivatized by Au(CN)2. Details of the structure determination are described in the legend to Table 1. MAD data were treated as a special case of multiple isomorphous replacement with inclusion of anomalous scattering (Table 1). Phases were initially computed using a maximum likelihood procedure (Table 1) and were markedly improved by noncrystallographic averaging and solvent flattening (Table 1). A typical section of averaged electron density map is shown in Figure III-2. Four subunit polypeptide chains could be unambiguously placed in the averaged electron density. The final model, which includes the tetrameric endonuclease of residues 9-179, 53 water oxygen atoms, 2 partially occupied SO4- ions, and 4 gold atoms, was refined to 2.3 A with good sterochemistry to an R factor of 0.204 and a free R of 0.268 (Table 1). A Two Domain Monomer The endonuclease monomer folds into two distinct domains, the NH2-terminal domain (residues 9-84) and the COOH-terrninal domain (residues 85-179) (Figure III-3). 100 FIGURE III-2. A section of the averaged electron density map contoured at lcr shown together with the final refined structure model. The real space correlation coefficient of the final endonuclease structure model with the averaged map is 0.68 as calculated by 0 (32) . 101 Figure III-2 102 FIGURE 111-3. Ribbon representation of one subunit of M. jannaschii endonuclease as obtained with the program RJBBONS (37). Secondary structure elements as referred in text are indicated. 104 The NH2-domain consists of a mixed antiparallel/parallel 13-sheet and three a helices. The first three 13 strands (131-133) are antiparallel, and together with 134, are packed against two nearly perpendicular a-helices, al and a2. The third a-helix, a3, is associated with a2 via an antiparallel coiled-coil. The inclination angle between a2 and a3 is 20° such that apolar residues at the interface mesh together with a "knobs into holes"' packing (15). 47 Hydrophobic residues packed at the interface of the coiled-coil include Leu , Leu51, and Ile 54 71 74 from a2 and Phe , Leu , and Ala 78 . from a3 . The connect10n between 134 and a3 (residues 64-67) is disordered, implying that this loop is flexible. This feature is consistent with the proteolytic sensitivity of this region (16). A short loop consisting of three 81 82 charged residues (Glu , Glu , Arg8 3) links the N-terminal domain to the C-terminal domain. The C-terminal domain is the a/l3 type and consists of a central four-stranded mixed 13-sheet containing strands 135-136-137 _f38 (upperlines denote parallel strands) cradled between helices a4 and a5. Two basic folding motifs lead to this fold; the rarely seen al3l3 (a4l35l36) and the commonly seen l3al3 (l37a5l38). The last 13 strand, 139, is partially hydrogen bonded with !38 and has little involvement in the C-terminal domain folding itself; rather, it participates in dimerization interactions (see below). The interface between the NH2-domain and the COOR-domain consists mainly of aromatic and polar residues, all of which are from helices a2 and a4. One notable feature is a hydrogen bond 49 triad mediated by a single water molecule linking three conserved residues, Glu , Tyr92 , 6 and Arg9 . 105 The Homotetramer: A Dimer of Two Dimers When viewed along a non-crystallographic two-fold axis, the homotetrameric M. jannaschii endonuclease resembles a parallelogram with the four subunits occupying the four corners (Figure III-4A). The quaternary structure of the endonuclease is unusual in that it does not assume the D2 point group symmetry observed in most homotetrameric proteins. It therefore may be more precise to describe endonuclease as a dimer of dimers. Subunits Al and Bl, related by a true two-fold axis, are closer in space than subunits A2 and B2 which are related by the same two-fold axis. Subunits Al and A2 or B 1 and B2 are related by two non-orthogonal pseudo two-fold axes plus a 1.6 A translation along the rotation axis. The resulting dimers, Al-A2 and Bl-B2, lie on two opposite sides of the parallelogram. Each of the pseudo two-fold axes relating monomers within the Al-A2 and Bl-B2 dimers are at an angle of 32° with respect to the two-fold axis which relates the two dimers. In addition the three two-fold axes do not intersect. Each pseudo twofold axis is shifted~ 10 Ain the opposite direction relative to the two-fold axis that relates the two dimers. The Ca's of the Al-A2 dimer can be superimposed with a root-meansquare-difference (RMSD) of0.47 A with those of the Bl-B2 dimer. There are larger structural differences observed between the subunits within each dimer, their Ca carbons have a RMSD of 1.2 A. The largest deviation is observed on the loop L7 which harbors the conserved His 125 residue. The asymmetry observed in the organization of the four identical subunits suggests non-equivalent roles for the two subunits in a dimer. 106 FIGURE III-4. The tetrameric M. jannaschii endonuclease and subunit-subunit interactions. A. Each subunit in the tetramer is given a distinct color and a label. The tetramer is viewed along the true two-fold axis relating the Al-A2 and Bl-B2 dimers. B. The isologous dimer formed between subunits Al and A2 viewed approximately perpendicular to the pseudo two-fold axis between them. The main chain hydrogen bonds formed between f39 and f39' and between loops LS and LS' are drawn as black lines. Side chains of the hydrophobic residues enclosed at the dimer interface are shown as blue balland-stick models. C. The heterologous interaction between subunits Al and B2 (or B 1 and A2) via the acidic loops LlO and LS. An electrostatically positive cleft (blue mesh surface) on subunit Al (or Bl) formed between the NH2- and COOH-domains is complementary to the protruding LlO and LS (both in ball-and-stick model) from subunit B2 (or A2). Secondary elements that form and surround the cleft are shown in yellow ribbon and labeled. Negatively charged residue side chains are colored red for LlO and LS residues. Backbones and hydrophobic residues are colored green. Amino acid from LlO and LS buried in the cleft are numbered. This figure was generated using the program GRASP (38) . 107 A B C Figure 111-4 108 Subunit Interfaces The endonuclease dimers Al-A2 and Bl-B2 associate isologously in a tail-to-tail fashion (Figure III-4B). The dimerization interface buries extensive surface area (2422 A2) (17) and is formed by the interaction of the central COOR-terminal P-sheets from each monomer via main chain hydrogen bonding and hydrophobic interactions. The pseudo two-fold axis that relates the two monomers lies between the two P9 strands and between the two LS loops in each subunit. The NR2 portion of P9 (169-172) is hydrogen bonded to PS via the main chain atoms. The COOR-terminal half of P9 (residues 172177) bends away from the plane of the pleated central P sheet to form main chain hydrogen bonds with the symmetry related residues of P9 in the isologous subunit, leading to a two-stranded P-sheet spanning the subunit boundary. The symmetrically related LS loops form another layer on top of the two-stranded P9 sheet, and together with the P9 sheet enclose a hydrophobic core at the intersubunit surface. Residues in the hydrophobic core include Phel39, Leu 141 , Leu 144 , Gly146, Val148, Leu158, Jle160, Met 174, and Tyrl76_ These hydrophobic residues are important for stabilizing the dimer interface. The buried surface area at the dimer interface is 74 percent hydrophobic (18), slightly higher than the average 65 percent observed for 18 oligomeric proteins (19). The interface between the two dimers is between subunits Al and B2 and subunits 2 B 1 and A2, each pair contributing 40 percent of 5420 A total buried surface area ( 17). The remaining 20 percent buried surface area can be accounted for by the interface between subunits Al and B 1. The subunits A2 and B2 are not in contact. Interestingly, while the buried non-polar surface area at the tetramer interface is 64 percent ( 18), comparable to the observed average value of 65 percent (19) , some strikingly close 109 contacts between polar residues of opposite charges are observed between the two heterologously associated subunits (Figure III-4C). A positively charged cleft with a dimension of 9 Awide, 12 Along and 14 Adeep is formed between the two domains of subunit Al (or B 1). It accommodates a protruding negative surface on the subunit B2 (or A2) formed primarily by loop LIO and partially by L8. Residues from LIO and L8 that are deeply buried by this inter-domain cleft are mostly acidic and are comprised of Aspl64_ Ala 165_Asp 166_Glyl67_Aspl68 and Glul35_Aspl36_ Several positively charged residues (Lys88, Lysl03, Lysl07, Argll3) line the cleft and interact electrostatically with the acidic loop residues. The interactions via loops LIO and L8 to the inter-domain cleft contribute nearly all the buried surface area at the tetramerization interface. In addition, three completely buried water molecules at the base of the cleft form hydrogen bonds with polar residues from both subunits, indicating solvent accessibility of this cleft prior to tetramer association. Most interestingly, one of the water molecules at the base of the cleft is part of the hydrogen bond network formed among the conserved Glu4 9, Tyr92, and Arg96 residues that stabilizes the interaction between the NH2- and the COOH-domains (see above) . The isologous dimer interaction formed via the C-terminal interface is expected to be more stable than the heterologous polar interaction mediated by LIO and L8 , suggesting that the tetramer assembles through dimer intermediates. Thus the structure appears to be a dimer of two dimers. This interpretation is in agreement with the results obtained from protein cross-linking (16) . The prevalent cross-linked species seen in this study were dimers and tetramers. It is plausible that the observed dimers in solution resemble isologous dimers Al-A2 or B1-B2 in the crystal structure. 110 Active Sites The tRNA splicing endonucleases cleave pre-tRNA leaving 5'-OH and 2',3' cyclic phosphate termini. This is the same specificity as in ribonucleases such as RNaseA and Tl. The RNaseA mechanism has been studied extensively and a first guess would be that the endonuclease mechanism should be chemically similar. Figure 111-5 shows the accepted reaction pathway and mechanism for RNaseA The two step reaction is acidbase catalyzed. A general base abstracts a proton from the 2'-OH of ribose leading to an in-line attack on the adjacent phosphodiester bond and the formation of a pentacovalent intermediate (20). The general acid protonates the 5' leaving group leading to the 2',3' cyclic phosphate product. In a second step a proton is abstracted from a water molecule, the hydroxyl ion attacks the 2',3 cyclic phosphate which is then hydrolyzed to give the 3' phosphate. Figure 111-5 also diagrams the accepted structure of the transition state. His 12 is the general base in the first step, His 11 9 the general acid and the pentacovalent transition state is stabilized by Lys41. In the splicing endonuclease family there is a conserved histidine residue. There is strong evidence that this histidine is part of the active site. Changing this histidine at position 242 in Sen34 to alanine was shown to markedly decrease 3' splice site cleavage (9) . Daniels (personal communication) has shown that the equivalent histidine mutant in the H. volcanii enzyme greatly impairs cleavage activity as have Lykke-Anderson and Garrett (16) for a His125 to Ala125 mutant in the M. jannaschii endonuclease. For reasons outlined below we propose that His 125 together with the absolutely conserved 111 F1GURE III-5. RNaseA reaction mechanism and the accepted transition state. Adapted from reference 20. l l2 OH B --------- His119 U . 0 "- / P.• s> 'H •. b~ N / \ ·•oo· HvN···H···O NH ~ - His12 o ir 8' + \ ; ~Lys41 Figure III-5 113 FIGURE III-6. The spatial arrangement of three strictly conserved residues, Tyr11 5, Hi5125, Lys156 in relation to the density attributed to a bound negative ion. The NCSaveraged electron density at 3.0 A, shown around a modeled sulfate ion, is contoured at 3cr. 114 Figure III-6 115 Tyr 115 and Lys 156 constitute a catalytic triad for acid-base catalysis carried out by the tRNA endonucleases. The His 125 in M. jannaschii endonuclease is located on the boundary between L7 and ~7. Even though His 125 is well ordered in the structure in all four subunits, the rest of the loop L7 (residues 119-124) has weak electron density suggesting flexibility of this 2 loop. The average B factor of L7 is 66 A in comparison with the average value of 44 A2 for the protein as a whole. Several striking structural features are seen in this region. There are primarily positively charged or polar residues in the vicinity of His 125, including two other strictly conserved residues, Tyr115 and Lysl56_ The side chains of His 125, Tyr115 and Lysl56 are within 4-7 A of each other and form a His-Tyr-Lys triad (5.5 A between His 125 and Lysl56, 6.7 Abetween Lysl56 and Tyr 11 5, 6.1 A between Tyr 156 and His 125) (Figure IIl-6). A distinct electron density feature(> 5cr in the averaged F0 map) is found in the middle of the triad. This density cannot be attributed to the polypeptide chain but is compatible with the size of a sulfate ion (Figure IIl-6). A metal ion is ruled out due to the positive environment of this site and we have currently modeled this site with a sulfate ion which is present in the crystal stabilization buffer, although it could be a chloride ion also present in crystallization solutions. The side chains of the His-Tyr-Lys triad in M. jannaschii endonuclease can be spatially aligned with those of the catalytic triad in RNaseA (Figure III-7). The orientation of this alignment results in an overlap of His 125 with Hisl2 of RNase A, Tyr 11 5 with His119, and Lysl56 with Lys41 (Figure III-8). This leads to a model in which the M. jannaschii endonuclease employs Hisl25 as the general base, Tyr 115 as the general acid 116 FIGURE III-7. Spatial arrangement of the putative catalytic triad of M. jannaschii (red residues) in comparison with that of those in RNaseA (blue residues). The three catalytic residues from RNaseA were manually brought to best overlap with those in M. jannaschii endonuclease using the program O (32). M. jannaschii endonuclease is represented by orange ribbon and RNaseA (coordinates are from protein data bank code 5rsa) is shown in silver ribbon. 117 MJ RNaseA Figure III- 7 118 and Lys 156 functions to stabilize the transition state. Consistent with this model, single mutations in each of these three residues lead to impaired RNA cleavage activity (5). Two out of the three putative active site residues (Tyrl 15 and His125) are in the long loop L7. This loop is flexible, as commonly seen in proteins that bind to the phosphodiester backbone of nucleic acid. The intron-exon junctions in archaeal pre-tRNA substrates are all located in single stranded regions. The flexible L7 loop could potentially participate in correct positioning of the scissile phosphate for cleavage. Interestingly, the Gly 292 to Glu mutation in the sen2-3 mutant in yeast occurs within the aligned sequences of this loop and is impaired in cleavage of the 5' splice site (9). The M. jannaschii endonuclease tetramer contains four active sites but the following reasons make it likely that only two of these participate in cleavage of the pretRNA substrate. The yeast endonuclease tetramer contains two functionally independent active sites for cleavage of the 5' and 3' splice sites, Sen2 and Sen34, respectively. The H. volcanii endonuclease is a homodimer and thus contains only two functional active sites. The two intron-exon junctions of the archaeal tRNA substrate are related by a pseudo-two-fold symmetry axis in the consensus bulge-helix-bulge substrate. It is therefore expected that the M. jannaschii endonuclease contains two functional active sites related by two-fold symmetry. M jannaschii endonuclease is protected to a maximum of 50% from proteolytic cleavage upon RNA substrate binding, consistent with the notion that this enzyme contains only two functional active sites (16). There are four pairs of active sites in the M. jannaschii endonuclease that are related by two-fold symmetry. These are found in subunits Al and B 1, A2 and B2, Al and A2, and B 1 and B2 (Figure 111-8). We favor the choice of subunits Al and B 1 whose His-Tyr-Lys triads are separated by -27 A. The location of amino acid residues in subunits Al and B 1 119 protected from proteolytic cleavage upon RNA substrate binding (16) are shown in Figure III-8. The protected residues cluster around the two proposed catalytic sites and more significantly fall on a path between the two sites. This region of the enzyme, formed by L9 and the COOH portions of cx.5 and P9 in Al and B 1, has a positive electrostatic potential suitable for binding of the phosphodiester backbone of the pre-tRNA substrate. By contrast, the distance between the His-Tyr-Lys triads of subunits A2 and B2 is 75 A and there is a less appealing distribution of substrate protected residues between this pair of active sites (data not shown). We attempted to dock a model of the substrate with the tetramer, keeping in mind the possibility that both the protein and the RNA substrate may undergo significant conformational changes upon association. The consensus substrate has two unpaired loops of three nucleotides separated by a helix of four basepairs (Figure III-IA). A model for the substrate was constructed from a related structure solved by NMR, that of HIV TAR RNA (21, 22). The TAR RNA structure is a stem loop structure with a bulge of three bases in the stem (Figure III-9A). Calnan et al. (23) showed that the binding of Tat to TAR is primarily via one arginine residue. Puglisi et al. (21) and later Aboul-ela et al. (22) solved the structures of both free RNA and RNA bound to arginarnide. The two structures are different. In the free RN A the bulged bases are stacked within the helix. In the arginarnide complex one of the bulged bases forms a base triple with an adjacent basepair and the other two bases are rotated out of the helix. The substrate model was 0 constructed by rotating a section of Tar RNA 180 so as to create a pseudo-symmetric structure of two three base bulges separated by a four base pair helix conforming to the archaeal endonuclease consensus substrate (Figure III-9B). Interestingly, it is the model 120 FIGURE III-8. Electrostatic potential surface of the M. jannaschii tetramer generated by the program GRASP (38) and the location of proteolytically protected residues by an RNA substrate analog on subunits Al and B 1. The surface viewed along the strict twofold axis relating the two dimers, but in opposite direction of that in Figure III-4A. The catalytic His 125 for all four subunits, Tyr 115 and Lys 156 for subunits Al and Bl , are shown in red stick models. The locations of residues for subunits Al and B 1 protected from proteolytic cleavage by an RNA substrate analog as found by Lykke-Adnersen and Garrett (16) are indicated by yellow spheres. 121 Figure III-8 122 FIGURE III-9. A modeled endonuclease-RNA substrate complex. A. Sequence and secondary structure of HIV TAR RN A (PDB code 1arj) used to model M. jannaschii pre-tRNA minimum substrate (B). B. A two-fold symmetry operation as indicated by the filled ellipse is applied to coordinates of the TAR RNA NMR structure (light gray region) to obtain a structure model for the second three-base bulge and the acceptor stem (dark gray region). The corresponding splice sites on the final model are indicated by arrows. C. Modeled complex of M. jannaschii endonuclease with RNA substrate obtained by manually aligning the phosphate backbone of the four base-paired helix with the positively charged surface between subunits Al and B 1. This view is perpendicular to that of Figure III-9. The two putative active subunits Al and B 1 are shown in orange and green ribbons and catalytic triad residues are shown in magenta. The phosphodiester bonds at the splice sites on RNA are depicted in blue and fit nearly perfectly with the putative sulfate density shown by blue dots. The distance from the center of the catalytic triad to the surface where RNA binds is around 7 A. 123 C A G G C C B C C G G A U G C A U c~G l A U G C C G C A UGGG HIV TAR RNA Modeled archaeal substrate Figure III -9 124 derived from the arginamide complex of TAR RNA that best docks with the M. jannaschii endonuclease tetramer. Figure III-9C shows that the two phosphodiester bonds that are to be cleaved fit exactly into the active sites and superimpose with the putative sulfate ion found there. The four base-pair helix separating the two bulges docks on the basic side of the enzyme and could interact with a number of the residues that Lykke-Andersen and Garrett (16) found to be protected from protease cleavage upon binding (see Figure III-8) of a consensus substrate analog. Though this is a speculative exercise, it suggests that the splice sites must be bulged in order to fit into the~ 7 A deep active site cavity (Figure III9C). This enzyme must be specific for its substrate. The requirement for bulged bases, precisely separated by four base pairs, would insure that the enzyme could not cleave helical substrates. Work by Thompson and Daniels (8) has shown that the elimination of the bulges at either splice site of an archaeal RNA substrate blocks cleavage by H. volcanii endonuclease. bnplications for the Structure of Other Splicing Endonucleases As the H. volcanii endonuclease dimer and the M. jannaschii tetramer recognize the same consensus substrate, their active sites must be arrayed similarly in space. This was hard to understand until Lykke-Andersen and Garrett pointed out (16) that the H. volcanii endonuclease monomer is in fact a tandem repeat of the consensus sequence of the endonuclease gene family. The two repeats are more similar to the M. jannaschii endonuclease sequence than to each other. In particular the N-terrninal repeat does not contain a convincing NH2-domain and lacks two of the three putative active site residues (Figure III-lOB). It does, however, contain many of the features of the COOH-domain important for structural arrangement of the enzyme, in particular the loop LIO sequence. 125 FIGURE III-10. Structure models of endonuclease from M. jannaschii and H. volcanii. A. Schematic drawing shows the M. jannaschii endonuclease sequence features followed by structural arrangement model of the endonuclease homotetramer. Several important structural features as discussed in the text are indicated: the loop LIO, the C-terminus ~9 strands as indicated by arrows, and the conserved catalytic residue His125 shown as a pentagon. The two putative active sides that participate in cleavage reaction are those in subunits labeled Al and B 1. B. Schematic drawing shows the H. volcanii endonuclease sequence features followed by the proposed structural arrangement model. At top, a bar diagram shows the sequence features conserved between H. volcanii and M. jannaschii endonucleases. The N-terminal repeat lacks the catalytic residues but has kept LIO in order to mediate the dimer-dimer interaction. The C-terminal repeat contains all the sequence features as seen in M . jannaschii enzyme. At the bottom, a proposed structure of H. volcanii homodimer is schematized. Dashed lines represent the polypeptide chain connecting the C- and Nterminal repeats. C. Proposed yeast endonuclease structural model. Conserved amino acid sequences near the C-termini of archaeal enzymes, M.jann. (Methanococcus jannashcii), H.v. Nt (Haloferax volcanii N-terminal repeat) and yeast Sen54 (Sc Sen54p) and Senl5 (Sc Senl5p) subunits are aligned. Important hydrophobic residues that stabilize the isologous C-terminus interaction observed between M. jannaschii subunits Al and A2 or B 1 and B2 are highlighted in green and circled on the structure models of heterodimers derived from the two-hybrid matrix analysis (9). Sen54 and Senl5 are aligned with the loop LIO sequences in M. jannaschii and H. volcanii and are highlighted in red. 126 A M. jannaschii B - - Y- - -L10 NH2 H. volcanii 06 r • • • • • • • • ~ •• • •• C M.jann.- - - - ·~~~~~ ...:::J.!l,__~ ~ 1~~~~ H.vol Nt. Sc. Sen54 Sc. Sen15 Figure 111-10 •• • 127 We propose that the H. volcanii enzyme forms a pseudo-tetramer of two pseudo-dimers (Figure III- IOB). The structure of the pseudo-dimer is predicted to contain a two stranded ~9-~9' pleated sheet, an important structural feature of the M. jannaschii endonuclease dimer. The tandem repeat must be connected by a stretch of polypeptide (dotted line in Figure III-IOB). Within each dimer of the M. jannaschii endonuclease structure, the N-terminus of subunit A2 (or B2) and the C-terminus of subunit Al (or B 1) is separated by 28A. This would require a span of at least 10 amino acid residues of an extended polypeptide chain in order to connect the region between the N-terminal and the C-terminal tandem repeats in the H. volcanii endonuclease. The H. volcanii enzyme model contains only two of the proposed active site triads- in the C-terrninal repeat of each pseudo-dimer- and these are proposed to occupy an identical spatial configuration to those in the Al and B 1 subunits of the M. jannaschii endonuclease. The pseudo-dimers are proposed to interact via the conserved LIO loop sequences found in the N-terminal repeat-equivalent to those in the A2 and B2 subunits in the M. jannaschii enzyme. This H. volcanii endonuclease structural model reveals that only two of the active sites are necessary but to array these in space correctly one must retain important features of both the isologous dimer interactions (~9-~9') and the dimer-dimer interactions mediated by the loop LIO found in the M. jannashcii crystal structure. The yeast endonuclease contains two active site subunits, Sen2 and Sen34. The other two subunits did not initially appear to belong to the endonuclease gene family; however, Lykke-Andersen and Garrett (16) pointed out sequence similarity of both Sen15 and Sen54 near the C-termini to the M. jannaschii endonuclease. This region contains both an LIO loop sequence and the conserved hydrophobic residues near the C-terminus 128 as seen in P8 and P9 in the crystal structure of the M. jannaschii endonuclease. We propose that the strong Sen2-Sen54 and Sen34-Senl5 interactions seen in the two-hybrid experiment are mediated by C-terminal P9-P9'-like interactions (Figure III-lOC). These dimers are associated to form the tetramer via the LIO loop sequences of Senl5 and Sen54 (Figure III-lOC). We conclude that the Archaea and the Eucarya have inherited from a common ancestor an endonuclease active site and the means to array two sites in a precise and conserved spacial orientation. This conclusion is supported by the results of Fabbri et al., (39), which demonstrate that both the eucaryal and the archaeal endonucleases can accurately cleave a universal substrate containing the BHB motif. The eucaryal enzyme appears to dispense with the more complex ruler mechanism for tRNA substrate recognition when it cleaves the universal substrate. Thus the precise positioning of two active sites in endonuclease appears to have been conserved in evolution. Subunits Al and B 1 comprise the active site core of all tRNA splicing endonucleases, and subunits A2 and B2 position the two active sites precisely in space. The eucaryal enzyme has evolved a distinct measuring mechanism for splice site recognition by specialization of the A2 and B2 subunits, and it has retained the ability to recognize and cleave the primitive consensus substrate. Table 1. Statistics for MAD data collection and phase determination from an Au(CN)2 derivatized endonuclease crystal, flash cooled to 100K. Crystals of recombinant endonuclease were grown at 22°C by the vapor diffusion method at pH 5.5-6.0 using KCl as a salting-in precipitant with a starting concentration of IM. Crystals were soaked for 3 days in a stabilization/cryoprotection solution containing Au(CN)2 (20mM Na-cacodylate, 129 pH 6.0, 20mM (NH4)2SO4, 25% glucose, and 2rnM Au(CN)2) prior to data collection. Crystals belong to the space group P212121, with unit cell dimensions a= 61.8 A, b = 79.8 A, and c = 192.8 A. Three wavelengths near the LIII absorption edge of Au were collected from a single crystal at National Synchrotron Light Source bearnline X4A using Fuji image plates: A1, the absorption edge corresponding to a minimum of the real part (f) of the anomalous scattering factor for Au; A2, the absorption peak corresponding to a maximum of the imaginary part (f ') of the scattering factor for Au; and 11,3, a high energy remote to maximize f. The crystal was oriented with its c* axis inclined slightly from the spindle axis (~ 16 degrees) to avoid a blind region. The a* axis was placed initially within the plane of the spindle and the beam direction using the STRATEGY option in the program MOSFLM (version 5.4) (24). A total of 180 consecutive 2° images were collected with every 8° region repeated for all three wavelengths. All data were processed using the program DENZO and scaled with SCALEPACK (25). A previously collected data set at 1.5418 A on a crystal soaked with the same Au(CN)2 solutions was included as the fourth wavelength. This data set was measured with CuKa X-rays generated by a Rigaku RU200 rotating anode using a SIEMENS multiwire detector and was indexed and integrated using the program XDS (26) and reduced and scaled with ROTAVATA and AGROV ATA in CCP4 (27). All reflection data were then brought to the same scale as those of wavelength 1 using the scaling program SCALEIT followed by the Kraut scaling routine FHSCAL (26) as implemented in CCP4 (27). The wavelength 1 (edge) has minimum value off for Au and this data set was taken as the "native" data with intrinsic anomalous signals. The data sets collected at wavelengths 2 (peak) and 3 (remotel) were treated as individual "derivative" date sets. Anomalous signals are included in the data set 130 of wavelength 2 where the X-ray absorption of Au is expected to be maximum. Due to the absence of a sharp white line at the Au LIII edge, the observed diffraction ratio at wavelength 2 is small, 0.061 (Table 1), as compared to the error signal ratio of 0.036 computed from centric reflections. The data set at wavelength 4 (remote 2) was taken at different time and from a different crystal, and was included as another "derivative" as it is near the maximum f value. Two gold atoms in each asymmetric unit could be clearly located by both dispersive difference (A.3 or A.4 with respect to AI) and anomalous (A.2) Patterson functions using the program HASSP (28). Two additional sites were later located by difference Fourier. We treated MAD data as a special case of the multiple isomorphous replacement with the inclusion of anomalous signals (29). Initial phases were computed at 3.0 A using the program MLPHARE as implemented in the CCP4 program suite, which uses a maximum-likelihood algorithm to refine heavy atom parameters (30). We exploited the fact that there are four non-crystallographically related monomers per asymmetric unit and performed four-fold averaging in parallel with solvent flattening and histogram mapping using the program DM (31). The subsequent electron density map was markedly improved and the polypeptide chain could be unambiguously traced. A model representing the monomeric endonuclease containing all but the first 25 amino acids ( the 8 Histidine residues, the 9 FLAG-tag amino acids and the first 8 residues in the wild type endonuclease sequence) was built into the electron density map using the interactive graphics program O (32). The assignment of side-chains for the M. jannaschii endonuclease polypeptite was confirmed by a difference Fourier map between the Au(CN)2 derivatized protein and a selenomethionine substituted protein. The final model, which includes the tetrameric endonuclease of residues 9-179, 53 water oxygen atoms, 2 131 partially occupied SO4- atoms, and 4 gold atoms, was refined to 2.3 A against the observed anomalous data of wavelength 3 using X-PLOR version 3.8 (33). Noncrystallographic restraints were applied to atoms between dimers Al-A2 and Bl-B2 excluding those from loops L3 and L7. The structure was assessed with the structure verification programs PROCHECK (34) and VERIFY3D (35) and was shown to be consistent with a correct structure. There are more than 90 percent residues in the most favored and no residues in disallowed regions of a Ramachandran plot (34). table 1 continued Data. Statistics * A.l A.2 A.3 "-4 Item (edge) (peak) (remote 1) (remote 2) Wavelength (A) 1.03997 1.03795 1.02088 1.5418 Resolution (A) (last shell) 2.28 (2.36-2.28) 2.28 (2.36-2.28) 2.28 (2.36-2.28) 3.0 (3.16-3.0) measured reflections 165549 161691 164948 50259 Unique reflections 75778 75310 76385 16679 0.052 (0.153) 0.051 (0.151) 0.054 (0.167) 0.085 (0.359) <l>/«J(I)> 11.3 (2.6) 10.6 (2.6) 10.3 (2.4) 8.7 (2.2) Completeness(%) 89.2 (62.2) 89.0 (62.0) 90.4 (65.4) 83.3 (42.6) f (electrons) -20.21 -15.21 -10.76 -5.10 f' (electrons) 10.20 10.16 9.87 7.30 0.059 0.060 0.165 0.061 [0.036] 0.062 [0.036] Rmerge t Dispersive ratios (30-3A)t . + Anomalous ratio (30-3A)"• 0.061 [0.041] 132 Phasing and Symmetry Averaging Statistics Resolution range (A) Refinement Statistics§ Resolution range (A) 50-2.28 R value 0.204 0.11 Rfree value (10% reflections) 0.268 0.24 non-hydrogen protein atoms 5584 1.02 water molecules 53 so4- atoms 9 Au atoms 4 50-3.0 Phasing power for acentric reflections A2 ->AI A3 ->AI 11.4 ->AJ Figure of merit Initial 0.22 After solvent flattening and averaging 0.78 root-mean-square-deviations from ideal geometry (36) Average correlation coefficient of the four-fold Bond (.A.) 0.01 I NCS operations Angle ( 0 ) 1.586 Initial 0.369 Improper Dihedral (0 ) 0.733 After solvent flattening and averaging 0.861 Torsion (0 ) 27. 195 table I continued *Numbers in parentheses correspond to those in the last resolution shell. t Rmerge indicates agreement of individual reflection measurements over the set of unique averaged reflections. Rmerge= ':Rh,i II(h)i <l(h)>I I ':Rh,iI(h)i where his the Miller index, i indicates individually observed reflections, and <l(h)> is the mean of all i reflections of the Miller index h, Bijvoet mates are treated as independent reflections when computing Rmerge. :j:Diffraction ratios are defined as<• 1F12> 112t<IFl 2> 112 where• IFl 2 are taken between the current and the reference wavelength (AI) for dispersive ratios and between matched Bijvoet pairs for anomalous ratios. Ratios for centric reflections at each wavelength are in shown in square parentheses. §Bulk solvent correction is included in the final cycles of refinement. 133 References 1. D. Biniszkiewicz, E. Cesnaviciene, D. A Shub, EMBO J. 13, 4629-4635 (1994) ; B. Reinholdhurek, D. A Shub, Nature 357, 173-176 (1992). 2. H. G. Belford, S. K. Westaway, J. N. Abelson, C. L. Greer, J. Biol. Chem. 268, 2444-2450 (1993); S. K. Westaway, H. G. Belford, B. L. Apostol, J. N. Abelson, C. L. Greer, J. Biol. Chem. 268, 2435-2443 (1993); G. M. Culver, S. M. Mccraith, S. A Consaul, D. R. Stanford, E. M. Phizicky, J. Biol. Chem. 272, 13203-13210 (1997). 3. G. M. Culver, et al., Science 261, 206-208 (1993); S. M. McCraith, E. M. Phizicky, J. Biol. Chem. 266, 11986-11992 (1991). 4. J. R. Palmer, T. Baltrus, J. N. Reeve, C. J. Daniels, Biochimica Et Biophysica Acta 1132, 315-318 (1992); J. Kjems, J. Jensen, T. Olesen, R. A Garrett, Canadian J. Microbial. 35, 210-214 (1989). 5. C. R. Trotta, unpublished results. 6. I. Gomes, R. Gupta, Biochem. and Biophy. Rev. 237, 588-594 (1997). 7. V. M. Reyes, J. N. Abelson, Cell 55, 719-730 (1988). 8. L. D. Thompson, C. J. Daniels, J. Biol. Chem. 265, 18104-18111 (1990). 9. C. R. Trotta, et al., Cell 89, 849-858 (1997). 10. K. Kleman-Leyer, D. A Armbruster, C. J. Daniels, Cell 89, 839-848 (1997). 11. C. K. Ho, R. Rauhut, U. Vijayraghavan, J. N. Abelson, EMBO J. 9, 1245-1252 (1990). 12. C. J. Bult, et al., Sci. 273, 1058-1073 (1996). 13. The recombinant endonuclease was purified by a single Ni-NTA (Qiagen) affinity chromotography step following a heat-denaturation step to high purity. 14. W. A Hendrickson, Sci. 254, 51 (1991). 15. C. Chothia, AV. Finkelstein, Annu. Rev. Biochem. 59, 1007-1039 (1990). 16. J. Lykke-Andersen, R. A Garrett, EMBO 1.16, 6290-6300 (1997). 17. B. Lee, and F. M. Richards, J. Mol. Biol. 55, 379-400 (1971). 18. Computed using the program Biograf (Molecular Simulations Incorporated, San Diego, California). 19. J. Janin, S. Miller, C. Chothia, J Mol Biol 204, 155-64 (1988). 134 20. C. Walsh, Enzymatic reaction mechanisms (W.H. Freeman and Company, San Francisco, 1979). 21. J. D. Puglisi, L. Chen, A D. Frankel, J. R. Williamson, Proc. Nat. Acad. Sci. 90, 3680-3684 (1993). 22. F. Aboul-ela, J. Karn, G. Varani, Nuc. Acids Res. 24, 3974-3981 (1996). 23. B. J. Calnan, B. Tidor, S. Biancalana, D. Hudson, AD. Frankel, Sci. 252, 1167- 11 71 (1991) . 24. AG. W. Leslie, CCP4 AND ESF-EACMB Newsletter on Protein Crystallography No. 26, Daresbury Laboratory, Warrington, U. K. (1992). 25. Z. Otwinowski, in Proceedings of the CCP4 Study Weekend: "Data Collection and Processing," 29-30 January 1993 L. I. Sawyer, N. Bailey, S., Ed. (SERC Daresbury Laboratory, England) pp. 56-62; W. Minor, XDISPLAYF program, Purdue University (1993). 26. W. Kabsch, J. App. Crystallogr. 21, 916 (1988). 27. Collaborative Computational Project, Number 4. "The CCP4 Suite: Programs for Protein Crystallography," Acta Crystallogr. D 50, 760-763 (1994). 28. T. C. Terwilliger, S. H. Kim, D. E. Eisenberg, Acta Crystallogr. A 43, 34-38 (1987). 29. V. Ramakrishnan, V. Biou, Methods Enzymol 276, 538-57 (1997). 30. Z. Otwinowski, in Isomorphous Replacement and Anomalous Scattering W. E. Wolf, P.R. Leslie, AG. W., Ed. (Science and Engineering Research Council, Daresbury Laboratory, Daresbury, U. K., 1991) pp. 80. 31. K. Cowtan, Joint CCP4 and ESF-EACBM Newsletter on Protein Crystallography 31, 34-38 (1994). 32. T. A Jones, J. Y. Zou, S. W. Cowan, Kjeldgaard, Acta Crystallogr. A 47, 110-9 (1991). 33. AT. Brtinger, X-PLOR Version 3.1, A system for crystallography and NMR, (Yale Univ. Press, Department of Molecular Biophysics and Biochemistry, Yale University, New Haven, CT, 1996). 34. R. A Laskowski, M. W. MacArthur, D. S. Moss, J. M. Thornton, J. App. Crystallogr. 26, 283-291 (1993). 135 35. R. Luthy, J. U. Bowie, D. Eisenberg, Nature 356, 83-5 (1992). 36. R. Engh, R. Huber, Acta Crystallogr. A 47, 392-400 (1991). 37. M. Carson, J. Mol. Graphics 5, 103-106 (1987). 38. A. Nicholls, K. A. Sharp and B. Honig, Proteins: Structure, Function and Genetics 11, 282 (1991). 39. S. Fabbri et al., Science 280, 284-286 (1998). 40. We thank D. Graham and C. Woese for M.jannaschii genomic DNA, P. Bjorkman and D. Rees for their helpful discussions and generosity in sharing their X-ray diffraction and computation equipment, C. M. Ogata for X-ray beam time allocation at NSLS X4A beamline, S. Diana for assistance in data collection, R. Story and M. Sales for critical review of the manuscript and other members of the Abelson laboratory for their advice and support. This work was supported by American Cancer Society grant NP802 and NIH NRSA individual postdoctoral fellowship F32 GM188930-0l. 136 Chapter IV - Investigations into the Endonuclease Active Site and Enzyme Architecture Introduction With an understanding of the tRNA splicing endonuclease at the molecular level, we can now address two main lines of investigation. First, what is the reaction mechanism utilized by the tRNA splicing endonuclease for the cleavage reaction? Second, how does the eukaryotic enzyme recognize the pre-tRNA substrate in a mature domain dependent manner? The following chapter will present experimental designs which have been utilized to begin to answer these questions. Dissection of the Interaction Domains of the Archaeal Endonuclease As desribed in detail in Chapter III, the archaeal enzyme forms a homotetramer through the use of several evolutionary conserved interaction domains. Dimerization is mediated by an isologous interaction between two endonuclease monomers (see pg. 106). Dimerization of this dimeric unit leads to tetramer formation and correct orientation of the endonuclease active sites for the cleavage reaction. Tetramerization is accomplished through a heterologous interaction of loop LIO from one monomer with the groove formed between the N- and C- terminal domains of another monomer. This interaction is mediated by aspartic acid residues found on loop LIO contacting basic residues which line the interaction groove (Figure III-4C). In order to confirm that these interaction have functional relevance for the architecture of the active tRNA splicing endonuclease, we have undertaken a mutagenesis approach. As a first target we chose the aspartic acid residues found in loop LIO. Aspartic acids 164 and 166 were mutated to lysine either 137 singly or as a double mutant. The proteins were expressed in E.coli with His/Flag tag and the mutant enzymes were purified. The purified endonuclease was tested for cleavage activity and the percent activity was qualitatively compared to wild type enzyme. For the mutant enzyme with single aspartic acid mutations, the enzyme was approximately 30% as active as wild type. Activity goes down to 2% of wild type for the double aspartic acid to lysine mutant. There are several possible explanations for the drastic effects of such mutants. The simplest and most consistent with our idea of how the endonuclease functions is that the change in charge interferes with the loop LlO and basic groove interaction. Thus the enzyme would be unable to tetramerize. A second possibility is that the mutation effects the overall architecture of the enzyme by causing misfolding. In order to distiguish between these two possiblities, we performed gel filtration on the mutant enzyme. Figure IV-1 shows that the double aspartic acid mutant fractionates as a dimer compared to the wild type tetramer. There is a small fraction of tetrameric enzyme, perhaps explaining why this mutant enzyme retains some activity. Thus this experiment represents a powerful demonstration that the interactions described in the crystal structure are indeed functionally relevant. Trunctation mutants are also being utilized to probe the architecture of the enzyme. For example, the loop LlO sequence inserts into a basic groove formed between the N- and Cterminal domains of the monomer (see Figure III-4B). Truncation of the endonuclease which removes the N-terminal domain results in enzyme which is dead. However, upon addition of the N-terminal domain in trans, full activity can be restored (CRT unpub). This result again supports the importance of the loop LlO heterologous interaction for the architecture of the enzyme. 138 FIGURE IV-1 . Gel filtration of DD mutant endonuclease and wild type endonuclease. 139 f\Jative DD Mutant I ; ' Figure IV-1 140 Conservation of the interaction domains in the eukaryotic enzyme makes a direct comparison of the effects of mutants on the archaeal and eukaryotic enzymes possible. The goal of such comparisons is to gain some insight into the architecture of the yeast endonuclease for which a crystal structure is surely a long way off due the heterotetrarneric nature of the enzyme. The Endonuclease Active Site As discussed in Chapter III, the endonuclease active site, defined by the absolutely conserved histidine, appears to have a similar configuration as that of RNaseA. Thus the mechanism of cleavage by the tRNA splicing endonuclease is proposed to be acid-base catalysis. Experiments to determine the reaction mechanism have yet to be conducted in a rigorous fashion, but we have produced the reagents to allow such an examination. Again mutagenesis, of the crucial catalytic triad residues His 125, Lys156 and Tyrl 15, is the tool of choice. Table IV-1 lists the mutants that have been produced. Also shown is the qualitative results of the cleavage efficiency of the mutants compared to the wild type enzyme. As can be seen from the table, both His125 mutants and Lys156 mutants exhibit a drastic effect on catalysis. Surprisingly, mutants in Tyrl 15 seem to have no effect on catalysis under the conditions which have been utilized for the assay. In the model of the active site triad, Tyr 115 is proposed to be the general acid. Thus it is possible that under the reaction conditions utilized, we do not see an effect of the mutation due the ability of H20 to substitute for the loss of the functional moiety normally presented by tyrosine. Careful kinetic analysis, including pH dependence of enzyme activity, should shed light on this discrepancy. 141 Table IV-1 List of endonuclease mutants and their qualitative splicing activity Mutation WT H125-.N H125-.A K156-.R K156-.Q K156-.A y11s-.A y11s-+1= y11s+H y110-.A y11sy110+fF D166-.K 0166D164-.KK Splicing Phenotype +++ ++ +/+/+/+/+++ +++ +++ +++ +++ + +/- +++ is wi Id-type activity ++ is approximately half wild-type activity + is approximately 10% wild-type activity +/- very Iittle activity detected 142 The Molecular Ruler Based on the model from Figure III-IOC, we have constructed a working model for the interaction of the eukaryotic enzyme with the pre-tRNA substrate (Figure IV-2). We derived the pre-tRNA from modeling the yeast tRNAPhe with the archaeal substrate (see legend for Figure 1-8). Using the known constraints for pre-tRNA cleavage, namely SEN2 contains the active site for the 5' splice site and SEN34, the active site for the 3' splice site, we docked the molecule in a similar fashion to the archaeal enzyme docking. Upon cloning of the subunits of the yeast endonuclease, we noted that SEN54 was a very basic protein and could thus interact with the phosphate backbone of the tRNA molecule. In our working model for interaction, we find that the SEN54 protein is positioned quite nicely for interaction with the mature domain of the tRNA. Of course this is all highly speculative, but it definitely leads to a testable model. Since the domains for subunit interaction in SEN54 (and SEN15) were shown by homology to be present at the C-terminus, we propose that the N-terminal domain of the 54kD protein would embody the molecular ruler. Mutations in the N-terminal domain should not affect enzyme architecture, but should alter the ruler mechanism causing either abolishment or aberrant cleavage of yeast substrates by endonuclease containing mutant Sen54. Thus by employing a similar system as was used for purification of endonuclease containing mutant SEN34 H242A (see pg.71), enzyme decrepit for cleavage could be purified from yeast. The mutant enzyme could then be tested in vitro for cleavage ability. The first step of this screen has been conducted. Truncation mutants of SEN54 have been cloned and expressed in wild type haploid yeast. Interestingly, as was true for the catalytically decrepit endonuclease containing SEN34 H242A, expression of several 143 FIGURE IV-2. Working model for interaction of the yeast tRNA splicing endonuclease with pre-tRNA. The intron is shown in blue and the yeast endonuclease subunits are as labeled. The pre-tRNA is derived as described in the legend for Figure I-8. 144 Figure I V-2 145 SEN54 truncation mutants result in a slow growth phenotype suggesting they are somehow affecting the normal processing events in a negative fashion. Inability of the purified endonuclease mutants to cleave the yeast substrate is not conclusive proof that the ruler mechanism has been disabled. One possibility is that the mutations somehow affect the catalytic efficiency of the enyzme and not the substrate recognition. Careful kinetic analysis may allow an understanding of the defect. However, the model suggests that the mutant enzyme should contain two functional active sites and would thus be catalytically functional. In order to determine if this is the case, cleavage of pre-tRNAArch-Euka could be used to test the mutant enzymes. It has been demonstrated by Tocchini-Valentini and co-workers (1998) that this substrate is cleaved by both the archaeal and eukaryotic enzymes. Upon cleaving this substrate, the eukaryotic enzyme does not utilize the ruler mechanism, thus mutations in the mature domain as well as deletions in the anticodon stem do not affect intron removal. This substrate represents a unique opportunity to be able to test the catalytic ability of endonuclease that contains the mutant SEN54 protein. The prediction is that the enzyme should possess the ability to cleave this substrate while being unable to cleave the yeast substrates. Materials and Methods Mutagenesis of M. jannaschii Endonuclease: Petl la containing the wild type endonuclease gene was subjected to PCR utilizing KTLA under standard conditions (Barnes, 1996). PCR was carried out in two reactions. First the 5' and 3' half of the gene were amplified using a mutagenic oligonucleotide which consisted of one or two nucleotide changes flanked on either side by l O nucleotides. The products were gel 146 purified on a 0.8% LMP agarose gel. 50ng of each product, as determined by gel quantitation, was added to a second PCR reaction. Oligonucleotides flanking the open reading frame were used and the full length gene was isolated by gel electrophoresis. The products were then cut with the original cloning sites, Ndel 5' and BarnHI3', and were subcloned into petl la for expression in E. coli. All mutants were sequenced to confirm the fidelity of the entire open reading frame at the Caltech Sequencing Facility. Purification of Mutant Endonuclease: E. coli strain BL21 containing petlla and pRI952 (a plasmid containing two tRNAs for the rare arginine and isoleucine bacterial codons) was inoculated into 3 rn1s of LB for overnight growth to saturation. 300rnls of LB + 100ug ampicillin were then inoculated 1: 100 with the culture and the cells were grown to OD600 = 0.6 - 0.8. IPTG was added to lmM to induce expression from the plasmid and the cells were grown for 2 more hours. Cells were harvested by centrifugation and the pellet was resuspended in 8 rnls. breaking buffer (50mM Phosphate Buffer pH6.7, 500mM KCl, 5% Glycerol). Cells were sonicated 8X10sec and the lysate was spun 10,000xg for 20 minutes. The supernatant was then heated to 65°C for 20 minutes, followed by a 10,000xg spin for 20 minutes. The supernatant was removed and 0.5 m1s ofNi-NTA (Quiagen), equilibrated in breaking buffer, was added to the lysate. Incubation for 30 minutes at room temperature was followed by collection of the beads by filtration through a disposable column, and the beads were washed with breaking buffer. The beads were then washed with 10 vol. breaking buffer+ 15mM imidazole. Protein was eluted with breaking buffer +200mM imidazole. Enzyme was checked by SDS-PAGE gel electrophoresis and dialyzed against breaking buffer. 147 Appendix I THE J OURNAL OF B IOLOGICAL C HEM1ST RY Minireview Vol. 273, No. 21, Issue of May 22, pp. 12685--12688, 1998 © 1998 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in U.S .A. tRNA Splicing* John Abelson:j:, Christopher R. Trotta, and Hong Li From the California Institute of Technology, Division of Biology, Pasadena, California 91125 Introns interrupt the continuity of many eukaryal genes, and therefore their removal by splicing is a crucial step in gene expression. Interestingly, even within Eukarya there are at least four splicing mechanisms. mRNA splicing in the nucleus takes place in two phosphotransfer reactions on a complex and dynamic machine, the spliceosome. This reaction is related in mechanism to the two self-splicing mechanisms for Group 1 and Group 2 introns. In fact the Group 2 introns are spliced by an identical mechanism to mRNA splicing, although there is no general requirement for either proteins or co-factors. Thus it seems likely that the Group 2 and nuclear mRNA splicing reactions have diverged from a common ancestor. tRNA genes are also interrupted by introns, but here the splicing mechanism is quite different because it is catalyzed by three enzymes, all proteins and with an intrinsic requirement for ATP hydrolysis. tRNA splicing occurs in all three major lines of descent, the Bacteria, the Archaea, and the Eukarya. In bacteria the introns are self-splicing (1-3). Until recently it was thought that the mechanisms of tRNA splicing in Eukarya and Archaea were unrelated as well. In the past year, however, it has been found that the first enzyme in the tRNA splicing pathway, the tRNA endonuclease, has been conserved in evolution since the divergence of the Eukarya and the Archaea. Surprising insights have been obtained by comparison of the structures and mechanisms of tRNA endonuclease from these two divergent lines. tRNA Precursors in Eukarya and Archaea The earliest studies of tRNA splicing were in the yeast Saccharomyces cerevisiae where tRNA introns were first discovered (4, 5). With the completion of the S. cerevisiae genome sequence it is now known that yeast contains 272 tRNA genes of which 59, encoding 10 different tRNAs, are interrupted by introns (6). The introns are 14-60 nucleotides in length and interrupt the anticodon loop immediately 3' to the anticodon (7). Among the 10 different yeast pre-tRNAs there is no conservation of sequence at the splice junctions although the 3' -splice junction is invariably in a bulged loop (8). Early studies on the structure of yeast tRNA precursors showed that the conformation of the mature domain is retained suggesting the model of the tertiary structure of eukaryal pre-tRNA shown in Fig. IA (9, 10). In the Archaea the introns are also small and often interrupt the anticodon loop, but they are found elsewhere as well, for example interrupting the dihydro U stem (11). In several of the Archaea, tRNA genes have been found that contain two introns. The splice sites are found in an absolutely conserved structural motif consisting of two loops of three bases separated by a four-base pair helix, the bulge-helix-bulge (BHB)1 motif(12). This structure, modeled in Fig. 1B from the related TAR RNA structure (13), allows the archaeal splicing mechanism to be extended to introns in rRNA that also retain this motif. Thus, early on it was suggested that the eukaryal and archaeal splicing systems operate by a different mechanism. * This minireview will be reprinted in the 1998 Minireview Compendium, which will be available in December, 1998. :j: To whom correspondence should be addressed. Fax: 626-796-7066; Email: abelsonj@cco.caltech.edu. 1 The abbreviations used are: BHB, bulge-helix-bulge; pre-tRNA, precursor tRNA; TAR, trans-activating response region. This paper is available on line at http://www.jbc.org The Pathway of tRNA Splicing in Eukarya The early discovery by Hopper and co-workers (14) that pretRNAs accumulate in the yeast mutant rnal-1 provided a source of pre-tRNA substrates, which allowed the development of the first in vitro RNA splicing system (15, 16). Using this system the pathway of tRNA splicing was deduced (17-20). The tRNA splicing reaction in yeast occurs in three steps; each step is catalyzed by a distinct enzyme, which can function interchangeably on all of the substrates (Fig. 2). In the first step the pre-tRNA is cleaved at its two splice sites by an endonuclease. The products of the endonuclease reaction are the two tRNA halfmolecules and the linear intron with 5' -OH and 3' -cyclic PO 4 ends. The endonuclease has been purified to homogeneity (6, 21). The enzyme behaves as an integral membrane protein, and since splicing takes place in the nucleus, it may be an inner nuclear envelope protein. The two tRNA half-molecules, in essence a nicked tRNA, are the substrate for the ensuing ligase reaction. This baroque reaction, catalyzed by the 90-kDa tRNA ligase (22, 23), takes place in three steps. In the first step the cyclic PO 4 is opened to give a 2'-PO 4 and 3'-OH. In the second step the 5'-OH is phosphorylated with the y-PO 4 of GTP (24, 25). tRNA ligase is adenylated at an active site lysine (26), and then the AMP is transferred to the 5'-PO 4 of the substrate. Formation of the 5'-3'-phosphodiester bond proceeds and AMP is released. The phosphate at the spliced junction is derived from the y-phosphate of GTP, and the phosphate originally at the 5'-splice site remains at tp.e spliced junction as a 2'-phosphate and must be removed to complete the splicing reaction. Phizicky and co-workers (27) have characterized the enzymology of the removal of the 2'-PO 4 from the spliced tRNA. A nicotinamide adenine dinucleotide (NAD)-dependent phosphotransferase catalyzes the transfer of the 2'-PO 4 to NAD (28, 29). Surprisingly the structure of the transfer product is ADP-ribose l"-2"-cyclic phosphate (30). The nicotinamide moiety is displaced, apparently supplying the energy for cyclization. It is tempting to speculate that this unique and hitherto unknown compound goes on to play some crucial regulatory role in the cell. Specificity in tRNA Splicing: Recognition of pre-tRNA by the Endonuclease The eukaryal endonuclease is solely responsible for the recognition of the splice sites contained in the pre-tRNA. Since only the mature domain in the pre-tRNA is conserved, it was postulated that endonuclease recognizes the splice sites by measuring the distance from the mature domain to the splice sites (3 1). This hypothesis was confirmed by experiments in which insertion mutations in the pre-tRNA that changed the distance between the mature domain and the anticodon resulted in a predictable shift of the splice sites (31). The intron is not completely passive in the recognition process. The Xenopus oocyte tRNA endonuclease has been shown to recognize a crucial element involving the intron (8). Yeast tRNA introns contain a conserved purine residue three nucleotides upstream of the 3' -splice site. This base must be able to pair with a pyrimidine at position 32 in the anticodon loop in order for the intron to be recognized by either yeast or Xenopus endonuclease (8, 32). These experiments suggested complexities in the structure of the pretRNA effecting the recognition by the eukaryal endonuclease, which had not been previously appreciated. It has also been demonstrated that there are different requirements for the recognition of the 5 ' - and 3 '-splice sites (32). Recognition of archaeal tRNA splice sites by the archaeal tRNA endonuclease relies solely on the BHB motif (Fig. lB). Yeast pretRNAs are not substrates for the Haloferax volcanii endonuclease (33), and unlike the eukaryal endonuclease, the mature domain of the pre-tRNA is not required for intron excision by the archaeal enzyme (12). Despite the differences in both substrate and the mechanism for substrate recognition between the archaeal and 12685 Minireview: tRNA Splicing 12686 FIG. 1. Proposed tertiary structure of the tRNA splicing substrates. Model of the eukaryal tRNA splicing endonuclease (AJ as proposed by Lee and Knapp (9) based on the crystal structure of tRNA he and the substrate for the archaeal tRNA splicing endonuclease (B ) modeled from the NMR structure of the HIV-1 TAR RNA (42, 43). The intron is shown in bold, dashed lines. The proposed tertiary structure of archaeal tRNA is shown in gray. pre-tRNA tRNA half-molecules Ligase (RLG1) CPDase Kinase + 4_ppG (6). Lig~se-ll,I-A ~ 1ASTase L1gase . / Alllpp c mature tRNA L Phosphotransferase (TPT1) Ligase ~ ~ Ji;~-- \I~r :~~ Iw -cp,~ OH OH ADP-ribose 1"-2" cyclic phosphate (App>P) gAJ;A_- _ M 1' -~p,~NJ w OH purifying the enzyme was the construction of a modified gene for the 44-kDa subunit, SEN2, containing His 8 and Flag epitope affinity tags. The SEN2 gene had been found earlier in genetic screens (34, 35), and the sen2-3 allele was shown to specifically block 5' -splice site cleavage (35). The enzyme turned out to be an a/3-yS heterotetramer whose subunits have molecular masses of 54 (SEN54), 44 (SEN2), 34 (SEN34), and 15 kDa (SEN15). Each of these genes proved to be essential for cell viability. The amino acid sequence of each of the four subunits contains a canonical nuclear localization sequence. Sen2 contains the only plausible transmembrane sequence, suggesting that it anchors the endonuclease complex to the nuclear membrane. Two subunits of endonuclease, Sen2 and Sen34, contain a homologous domain approximately 130 amino acids in length, suggesting that they perform a similar function. The excitement came when we learned that apparent homologs of this domain are encoded by the gene for the archaeal tRNA splicing endonuclease of H. volcanii cloned by Daniels and co-workers (36) and in endonuclease homologs found in the sequenced genomes of Methanococcus jannaschii , Methanobacterium thermoautotrophicum, and Archaeoglobus fulgidis. The homology between Sen2, Sen34, and the archaeal endonucleases immediately suggested a model in which the yeast endonuclease contains two distinct active sites, one for each splice site (6, 37). The fact that sen2-3 is defective in cleavage of the 5'-splice site suggested that Sen2 contains the active site for 5'-splice site cleavage and led to the prediction that Sen34 cleaves the 3'-splice site. This hypothesis was strongly supported by the observation that a Sen34 mutant enzyme is defective in 3'-splice site cleavage OH NAO+ (AppN) FIG. 2. The tRNA splicing pathway of yeast. Gene names are given in parentheses for the proteins of the pathway. See text for details. CPDase, cyclic phosphodiesterase; ASTase, adenylyl synthetase. Adapted with permission from Refs. 24 and 30. eukaryal systems, as we shall discuss below, the endonuclease that catalyzes the first step in splicing has been conserved between Eukarya and Archaea. Different mechanisms for substrate recognition have evolved since the divergence from their common ancestor. The Yeast and Archaeal tRNA Splicing Endonucleases: Related Enzymes The characterization of the yeast tRNA endonuclease was extremely difficult for two reasons. The enzyme was present at very low levels, approximately 150 molecules per cell, and it appeared to be an integral membrane protein. However, after many years of work the enzyme was successfully purified, and the genes for all four of its subunits were cloned (6, 21). Of particular advantage in Our biochemical experience with the endonuclease had suggested strong interactions between the subunits. To probe the nature of these interactions a two-hybrid experiment · was performed in which all possible pairwise combinations of the four subunits were probed (6). It turned out that strong interactions were only seen between Sen2 and Sen54 and between Sen34 and Sen15. These results together with those described above lead us to a model for the yeast endonuclease in which Sen2 contains the active site for 5' -splice site cleavage and Sen34 the active site for 3'-splice site cleavage (see below; Fig. 4C). There is as yet no evidence as to how the ruler mechanism works, although we propose that it could be via the interaction of Sen54, a very basic protein, and Sen2. The tRNA splicing endonuclease of the archaeon H. volcanii was shown to behave as a homodimer in solution (36). Since its substrate, the consensus BHB motif, has pseudo 2-fold symmetry, it seems very likely that the 2-fold symmetric dimer recognizes its substrate such that each splice site is cleaved by a separate active site (see below). Thus we are led to a unified model of tRNA splicing in which the two splice sites are cleaved by separate protein subunits, each containing an active site. The Three-dimensional Structure of an Archaeal tRNA Splicing Endonuclease M. jannaschii contains a gene that encodes an endonuclease homologous to the H. volcanii enzyme but is about half the size ( 179 amino acids in the case of M. jannaschii). We believed that a high resolution structure of the simpler archaeal enzyme would shed light on the mechanism of the more complicated but related eukaryal endonuclease and thus embarked on a structural study of the M. jannaschii endonuclease, obtaining an x-ray structure refined to a resolution of 2.3 A (13). The M. jannaschii endonuclease is an a 4 tetramer different from the dimeric H. volcanii enzyme (13, 36, 38). Fig. 3A shows that the M. jannaschii endonuclease monomer consists of two distinct domains: the N-terminal domain (residues 9-84) and the C-terminal domain (residues 85-179). The N-terminal domain consists of three a-helices and a mixed antiparallel/parallel /3-pleated sheet of four strands. The C-terminal domain contains two a-helices flanking a five-stranded mixed /3-sheet. The last strand /39 is partially hydrogen bonded with /38, but its main interactions mediate the isologous pairing seen in the tetramer (see below). Understanding the interactions between monomers has turned out to be crucial to under- Minireview: tRNA Splicing A 12687 A M. Jannaschll B H. volcanll NH2 y L,o NH2 Y H KL10 - . . - - .zi""11!'1!1, s' ·•: •"p 1 C FIG. 3. Crystal structure of the M. jannaschii tRNA splicing endo- nuclease. A, ribbon representation of the M. jannaschii tindonuclease monomer. The proposed catalytic triad residues are within 7 A of each other and are shown in red ball-and-stick models (see text). The electron density is drawn near the putative catalytic triad to represent a putative so~- bound in the active site pocket. B, M. jannaschii endonuclease tetramer. Each subunit is represented by a distinct color and a label. The tetramer is viewed along the true 2-fold axis relating the Al-A2 and Bl-B2 dimers. The main chain hydrogen bonds formed between /39 and /39' and between loops LS and LS ' for isologous dimers are drawn as thin lines. Side chains of the hydrophobic residues enclosed at the dimer interface are shown as blue ball-andstick models. The heterologous interaction between subunits Al and B2 (or Bl and A2) through the acidic loops LlO and LS are highlighted by dotted surfaces. C, model of the M. jannaschii endonuclease active site subunits Al and Bl docked with a model substrate based on the TAR RNA structure (intron in blue ) (see also Fig. lB). Adapted with permission from Ref. 13. standing the structure and evolution of the members of the tRNA splicing endonuclease family. Two pairs of subunits (Al and A2, Bl and B2) associate to form isologous dimers via extensive interactions between their {39 strands (Fig. 3B). The carboxyl half of {39 from one subunit forms main chain hydrogen bonds with the symmetry-related residues of {39 from the other subunit ({39'), leading to a two-stranded {3-sheet spanning the subunit boundary. The two LS loops, also hydrogen bonded and related by the same symmetry, form another layer on top of the two-stranded {39 sheet. The {39-{39' sheet together with L8-L8' encloses a hydrophobic core at the intersubunit surface. These hydrophobic residues are important for stabilizing the dimer interface and lead to an extremely stable dimeric unit, which we believe has been conserved in evolution (see below). The tetramer is formed via heterologous interaction between the two dimers. The main interaction between the two dimers is via the insertion of loop Ll0 from subunits A2 and B2 into a cleft in subunits Bl and Al between the N- and C-terminal domains of each monomer. The interaction is primarily polar between acidic residues in loop Ll0 and basic residues in the cleft. This arrangement causes the two isologous dimers to be translated relative to each other by about 20 A. This brings subunits Al and Bl much closer together than A2 and B2, which do not interact at all, and results, as we shall see below, in an arrangement of subunits in which only one symmetrically disposed pair of active sites can recognize the substrate. These interactions, though probably less stable, are also likely to have been conserved in evolution because they lead to a distinctive interaction between the dimers, which in turn leads to the required positioning of the two active sites. The tRNA splicing endonucleases cleave pre-tRNA leaving 5'-OH and 2'-3'-cyclic PO 4 termini. This is the same specificity seen in the ribonucleases, RNase A, Tl, etc. The RNase A mechanism has been studied extensively, and a first guess would be that chemically the endonuclease mechanism should be similar (39, 40). The reaction pathway for RNase A is a two-step acid-base-catalyzed reaction. A general base abstracts a proton from the 2'-OH of ribose leading to an in-line attack on the adjacent phosphodiester bond and the formation of a pentacovalent intermediate. The general acid protonates the 5'-leaving group leading to the 2'-3' cyclic M.jann. H.vol.Nt. Sc. Sen54 Sc. Sen15 FIG. 4. Model of the tRNA splicing endonucleases of M. jannaschii, H. volcanii, and S. cerevisiae. A, model of the M. jannaschii homotetramer. Several important structural features discussed in the text are indicated: loop Ll0, the COOR-terminal /39 strands (arrows) and the conserved catalytic residue Ris 125 (pentagon). B, proposed subunit arrangement of the H. volcanii endonuclease. The two tandem repeats are more similar to the M. jannaschii endonuclease sequence than to each other. The NR 2 -terminal repeat lacks two of the three putative active site residues (white bars). It does, however, contain many of the features of the COOR domain, which are important for structural arrangement of the enzyme, in particular the Ll0 sequence (yellow bars). The COOR-terminal repeat contains all the sequence features of the M. jannaschii enzyme. Dashed lines represent the polypeptide chain connecting the COOR- and NR2 -terminal repeats. C, a proposed structural model for the yeast endonuclease. Conserved amino acid sequences near the COOR termini of archaeal enzymes, M. jannaschii (M. Jann.), H. volcanii NR~terminal repeat (H. vol. Nt.), and yeast Sen54 (Sc. Sen54) and Sen15 (Sc. !::ienl5) subunits are aligned. Important hydrophobic residues that stabilize the isologous COOR terminus interaction between M. jannaschii subunits Al and A2 or Bl and B2 are highlighted in green and circled on the structural models ofheterodimers. The sequences ofSen54 and Senl5 aligned with LIO sequences in M. jannaschii and H. volcanii are highlighted in red. Loops Ll0 on both Sen54 and Sen15 are labeled on the proposed heterotetramer model of the yeast endonuclease. Reprinted with permission from Ref. 13. PO 4 product. In a second step a proton is abstracted from H 2 O, OH- attacks, and the 2'-3- cyclic PO 4 is hydrolyzed to give the 3' -PO 4 . In RNase A, His 1 2 is the general base in the first step, His 119 is the general acid, and the pentacovalent transition state is stabilized by Lys 41 . In the endonuclease family there is a conserved histidine residue at position 125 in the M. jannaschii enzyme. There is strong evidence that this histidine is part of the active site. A change to alanine in the equivalent histidine at position 242 in Sen34 was shown to impair 3'-splice site cleavage (6). Daniels 2 has shown that the equivalent histidine mutant in the H. volcanii enzyme impairs cleavage as has Garrett (38) for a His 125 to Ala mutant in the M. jannaschii endonuclease. His 125 (in L 7) is found in a cluster with conserved residues Tyr 115 (in L 7) and Lys 156 (in a5) on the surface of the monomer and forms a pocket into which the scissile phosphate to be cleaved is proposed to fit (Fig. 3A). Significantly these three residues can be spatially superimposed with the catalytic triad of RN ase A. In the superposition, His 125 is equivalent to His 12 of RNase A and should be the 2 C. J. Daniels, personal communication. 12688 Minireview: tRNA Splicing general base; Tyr 115 should be the general acid, and Lys 156 stabilizes the transition state. This would appear to be a case of convergent evolution, because it is clear that the tRNA endonucleases and the RNases do not share a common ancestor. We expect two of the M. jannaschii endonuclease active sites to recognize and cleave the symmetric tRNA substrate. We favor the choice of the symmetrically disposed subunits Al and Bl to function as active subunits. The active sites on subunits Al and Bl are at one side of the tetramer that is shown to be basic from an electrostatic potential calculation. This side of the surface could therefore bind the phosphodiester backbone of the tRNA substrate. A model of the substrate derived from the TAR RNA NMR structure docks well with the proposed active subunits Al and Bl (Fig. 3C). The two phosphodiester bonds fit exactly into the Al and Bl active site pocket and superimpose with the putative so~density found in each site. The distance between other pairs of active sites in the tetramer is too long to fit with this model substrate. This is particularly so of the other symmetrically related pair, in A2 and B2, which are so far apart that it is unlikely that any change in substrate geometry could allow for a fit. Obviously it is of high priority to solve a structure of the enzyme-substrate complex. The Structure of the M. jannaschii Endonuclease Provides Insight into the Evolution of tRNA Splicing The H. volcanii dimer and the M. jannaschii tetramer recognize the same consensus substrate so their active sites must ultimately be arrayed similarly in space. This was difficult to understand until Garrett pointed out that the H. volcanii monomer is actually a tandem repeat of the consensus sequence of the endonuclease gene family (38). The N-terminal repeat does not contain the N-terminal domain, and it lacks 2 of the 3 putative active site residues. It does, however, contain the structural features of the C-terminal domain, in particular the Loop LIO sequence. Fig. 4 shows a proposed model of the H. volcanii enzyme, which is best described as a pseudotetramer of two pseudo-dimers. The structure of the pseudo-dimer is predicted to contain a two stranded /39-/39 ' pleated sheet, an important structural feature of the M. jannaschii dimer. The H. volcanii enzyme only contains two active sites (found in the Cterminal repeats), and these are proposed to occupy an identical spatial configuration to those in the Al and Bl subunits of the M. jannaschii enzyme. The pseudo-dimers are proposed to interact via the conserved loop LIO sequences in the N-terminal repeats, equivalent to those in the A2 and B2 subunits in the M. jannaschii enzyme. The H. volcanii enzyme tells us that only two of the active sites are necessary, but to array these in space correctly one must retain important features of both the isologous dimer interactions (/39-/39' ) and the dimer-dimer interactions mediated by Loop LIO. The yeast endonuclease contains two active site subunits, Sen2 and Sen34. The other two subunits do not appear to belong to the endonuclease gene family; however, Garrett pointed out that both Sen15 and Sen54 contain a stretch of sequence similarity to the endonuclease family near their COOH termini (38). Upon inspection of the crystal structure of the M. jannaschii endonuclease, this sequence conservation appears to contain the Loop LIO and hydrophobic core interactions crucial to the association of the monomers to form the tetrameric M. jannaschii endonuclease. We have proposed that the strong Sen2-Sen54 and Sen34-Sen15 interactions seen in two-hybrid experiments (13) are mediated by the C-terminal {39-/39' -like interactions. These two heterodimers are proposed to interact to form the heterotetramer via the conserved Loop 10 sequences of Sen15 and Sen54. Thus it is likely that what has been conserved since the divergence of the Eukarya and the Archaea is the endonuclease active site and the means to array two of them in a precise and conserved spatial orientation. Further support for this evolutionary pathway is supported by the results of Tocchini-Valentini and co-workers (41), where it is demonstrated that both the eukaryal and archaeal endonucleases can accurately cleave a universal substrate contain- ing the BHB motif. The eukaryal enzyme seems to dispense with the ruler mechanism for tRNA substrate recognition when cleaving the universal substrate. This leads to the conclusion that the precise positioning of two active sites in endonuclease has been conserved. Thus, subunits Al and Bl comprise the active site core of all tRNA splicing endonucleases, and subunits A2 and B2 position the two active sites precisely in space. The eukaryal enzyme has evolved a distinct measuring mechanism for splice site recognition via the specialization of the A2 and B2 subunits while retaining the ability to recognize and cleave the primitive consensus substrate. REFERENCES 1. Biniszkiewicz, D. , Cesnaviciene, E., and Shub, D. A. (1994) EMBO J. 13, 4629-4635 2. Reinhold-Hurek, B. , and Shub, D. A. (1992) Nature 357, 173-176 3. Kuhse!, M. G., Strickland, R., and Palmer, J. D. (1990) Science 250, 1570 -1573 4. Goodman, H. M., Olson, M. V. , and Hall, B. D. (1977) Proc. Natl. Acad. Sci. U.S.A. 74,5453- 5457 5. Valenzuela, P. , Venegas, A., Weinberg, F., Bishop, R. , and Rutter, W. J . (1978) Proc. Natl. Acad. Sci. U.S. A. 75, 190- 194 6. Trotta, C. R., Miao, F., Arn, E. A., Stevens , S. W., Ho, C. K., Rauhut, R. , and Abelson, J. N. (1997) Cell 89, 849 - 858 7. Ogden, R. C., Lee, M. C., and Knapp, G. (1984) Nucleic Acids Res. 12, 9367-9382 8. Baldi, M. I., Mattoccia, E., Bufardeci, E., Fabbri, S., and Tocchini-Valentini, G. P. (1992) Science 255, 1404- 1408 9. Lee, M. C., and Knapp, G. (1985) J. Biol. Chem. 260, 3108-3115 10. Swerdlow, H. , and Guthrie, C. (1984) J. Biol. Chem. 259, 5197- 5207 11. Thompson, L. D., Brandon, L. D., Nieuwlandt, D. T., and Daniels, C. J. (1989) Can. J. Microbial. 35, 36 - 42 12. Thompson, L. D., and Daniels , C. J. (1988) J. Biol. Chem. 263, 17951-17959 13. Li, H., Trotta, C. R., and Abelson, J. N. (1998) Science 280, 279 - 284 14. Hopper, A. K., Banks, F. , and Evangelidis, V. (1978) Cell 14, 211- 219 15. Knapp, G., Beckmann, J . S., Johnson, P. F., Fuhrman, S. A., and Abelson, J. N. (1978) Cell 14, 221-236 16. O'Farrell, P. Z. , Cordell, B. , Valenzuela, P. , Rutter, W. J., and Goodman, H . M. (1978) Nature 438-445 17. Greer, C. L., Peebles, C. L. , Gegenheimer, P., and Abelson, J. (1983) Cell 32, 537-546 18. Peebles, C. L., Ogden, R. C., Knapp, G. , and Abelson, J. (1979) Ce// 18, 27-35 19. Knapp, G., Ogden, R. C. , Peebles, C. L., and Abelson, J. (1979) Cell 18, 37-45 20. Peebles, C. L., Gegenheimer, P., and Abelson, J. (1983) Cell 32, 525- 536 21. Rauhut, R., Green, P. R., and Abelson, J. (1990) J. Biol. Chem. 265, 18180 - 18184 22. Phizicky, E. M., Schwartz, R. C., and Abelson, J. (1986) J. Biol. Chem. 261, 2978 - 2986 23. Westaway, S. K., Phizicky, E. M., and Abelson, J. (1988) J . Biol. Chem. 263, 3171- 3176 24. Westaway, S. K., Belford, H. G., Apostol, B. L. , Abelson, J. , and Greer, C. L. (1993) J. Biol. Chem. 268, 2435-2443 25. Belford, H. G., Westaway, S. K., Abelson, J., and Greer, C. L. (1993) J. Biol. Chem. 268, 2444 - 2450 26. Xu, Q., Teplow, D., Lee, T. D. , and Abelson, J. (1990) Biochemistry 29, 6132-6138 27. Culver, G. M., McCraith, S. M., Consaul, S. A., Stanford, D.R. , and Phizicky, E. M. (1997) J. Biol. Chem. 272, 13203- 13210 28. McCraith, S. M ., and Phizicky, E. M. (1990) Mo/. Cell. Biol. 10, 1049- 1055 29. McCraith, S. M., and Phizicky, E. M. (1991) J. Biol. Chem. 266, 11986- 11992 30. Culver, G. M. , McCraith, S. M., Zillmann, M. , Kierzek, R. , Michaud, N. , Lareau, R. D., Turner, D. H., and Phizicky, E. M. (1993) Science 261 , 206-208 31. Reyes , V. M., and Abelson, J . (1988) Cell 55, 719 - 730 32. Di Nicola Negri , E., Fabbri, S., Bufardeci, E., Baldi, M. I., Gandini Attardi, D., Mattoccia, E., and Tocchini-Valentini, G. P. (1997) Cell 89, 859-866 33. Palmer, J. R. , Nieuwlandt, D. T. , and Daniels, C. J. (1994) J. Bacterial. 176, 3820 - 3823 34. Winey, M. , and Culbertson, M. R. (1988) Genetics 118, 49 - 63 35. Ho, C. K., Rauhut, R., Vijayraghavan, U., and Abelson, J. (1990) EMBO J. 9, 1245- 1252 36. Kleman-Leyer, K., Armbruster, D. A., and Daniels, C. J. (1997) Cell 89, 839 - 848 37. Belfort, M., and Weiner, A. (1997) Cell 89, 1003- 1006 38. Lykke-Andersen, J., and Garrett, R. A. (1997) EMBO J. 16, 6290-6300 39. Thompson, J. E., and Raines, R. T. (1994) J. Am. Chem. Soc. 116, 5467- 5468 40. Walsh, C. (1979) Enzymatic Reaction Mechanisms, pp. 202-207 , W. H. Freeman and Co., San Francisco 41. Fabbri, S., Fruscoloni, P., Bufardeci, E., Di Nicola Negri, E. , Baldi, M. I. , Gandini Attardi, D., Mattoccia, E., and Tocchini-Valentini, G. P. (1998) Science 280, 284 - 286 42. Calnan, B. J. , Tidor, B., Biancalana, S. , Hudson, D., and Frankel, A. D. (1991) Science 252, 1167- 1171 43. Aboul-ela, F., Karn, J., and Varani , G. (1996)NucleicAcids Res. 24, 3974-3981 152 Appendix II Cell , Vol. 89, 849-858, June 13, 1997, Copyright ©1997 by Cell Press The Yeast tRNA Splicing Endonuclease: A Tetrameric Enzyme with Two Active Site Subunits Homologous to the Archaeal tRNA Endonucleases Christopher R. Trotta, Feng Miao,* Eric A. Arn,t Scott W. Stevens, Calvin K. Ho, Reinhard Rauhut,:t and John N. Abelson Division of Biology California Institute of Technology Pasadena, California 91125 Summary The splicing of tRNA precursors is essential for the production of mature tRNA in organisms from all major phyla. In yeast, the tRNA splicing endonuclease is responsible for identification and cleavage of the splice sites in pre-tRNA. We have cloned the genes encoding all four protein subunits of endonuclease. Each gene is essential. Two subunits, Sen2p and Sen34p, contain a homologous domain of approximately 130 amino acids. This domain is found in the gene encoding the archaeal tRNA splicing endonuclease of H. volcanii and in other Archaea. Our results demonstrate that the eucaryal tRNA splicing endonuclease contains two functionally independent active sites for cleavage of the 5' and 3' splice sites, encoded by SEN2 and SEN34, respectively. The presence of endonuclease in Eucarya and Archaea suggests an ancient origin for the tRNA splicing reaction. Introduction Although RNA splicing is a required step for gene expression in all three taxonomic kingdoms, at least four different mechanisms of splicing have evolved. The most primitive mechanism, autocatalytic self splicing of Group I and Group II introns, involves two phosphotransfer reactions (Cech, 1993). Pre-mRNA splicing requires ATP hydrolysis and more than 100 proteins, but it proceeds via the same chemical mechanism used by Group II introns, suggesting that it is also RNA catalyzed and that it shares a common origin with Group II intron splicing (Moore et al., 1993; Madhani and Guthrie, 1994). In contrast to these RNA-catalyzed reactions, splicing of eucaryotic tRNA introns occurs by a mechanism involving three protein enzymes (Winey and Culbertson, 1989; Abelson, 1991; McCraith and Phizicky, 1991 ). In the archaeal kingdom, tRNA splicing proceeds by a similar protein-based enzymatic mechanism, but this splicing is distinct by a number of functional criteria (Thompson et al., 1989). Splicing of all known prokaryotic tRNA introns occurs via the primitive autocatalytic mechanism (Reinholdhurek and Shub, 1992; Biniszkiewicz et al., *Present Address: Beckman Research Institute of the City of Hope, 1450 E. Duarte Rd., Duarte, California 9101 0. tPresent Address: Department of Cell Biology, Harvard Medical School, Boston, Massachusetts 02115. :tPresent Address: Justus Liebig Universitat Giessen, lnstitut tor Mikrobiologie und Molekularbiologie, Frankfurter Strasse 17, 35392 Giessen, Germany. 1994). So while tRNA splicing occurs in all three king doms, the mechanism for this process is not conserved. Major questions regarding the origin of tRNA introns and the relationship between the disparate enzymatic machineries required for their removal therefore remain. Do the different splicing mechanisms reflect independent origins or a single origin and later divergence? The results of this work together with those in the accompanying papers of Di Nicola Negri et al. (1997 [this issue of Ce//]) and Kleman-Leyer et al. (1997 [this issue of Ce//]) provide new insight into the origin of the tRNA splicing machinery. Much of our knowledge of the mechanism of eucaryotic tRNA splicing comes from studies in Saccharomyces cerevisiae. The S. cerevisiae genome contains 272 tRNA genes of which 59, encoding ten different tRNAs, contain introns (C. R. T., unpublished data). These introns are 14-60 bases in length, and all interrupt the anticodon loop immediately 3' to the anticodon (Ogden et al. , 1984). Splicing of introns from precursor tRNA (pre-tRNA) is accomplished through the action of a sitespecific endonuclease, an RNA ligase, and a phosphotransferase. The tRNA splicing endonuclease (referred to herein as "endonuclease") cleaves pre-tRNA at the 5 ' and 3 ' splice sites to release the intron. The products of the endonuclease reaction are an intron and two tRNA half-molecules bearing 2' ,3 ' cyclic phosphate and 5' -OH termini (Peebles et al., 1983). tRNA ligase covalently joins the half-molecules through a complex series of ATP- and GTP- dependent reactions (Westaway et al. , 1988; Belford et al., 1993; Westaway et al., 1993). After ligation, a 2' phosphate remains at the splice junction and is removed by a specific phosphotransferase (McCraith and Phizicky, 1991; Culver et al. , 1993). Figure 1A summarizes the conserved features of the ten yeast pre-tRNAs containing introns (as depicted by Lee and Knapp, 1985). There are no conserved sequences at the splice sites, but the intron is invariably located at the same site in the gene, placing the splice sites an invariant distance from the constant structural features of the tRNA body. This conserved placement implies that recognition of the splice sites may be accomplished via interaction with the mature tRNA domain. The hypothesis that the endonuclease measures the distance from the mature domain to the splice sites to correctly place cleavages (Figure 18) was confirmed by experiments in which insertions in the anticodon stem of pre-tRNA that changed the distance between the mature domain and the splice sites resulted in a predictable shift of the cleavage sites (Reyes and Abelson, 1988). Further evidence for a strict measuring mechanism was provided by the fact that an artificial tRNA intron composed almost entirely of uridine residues was spliced correctly in the yeast system (Reyes and Abelson, 1988). The intron is not, however, completely passive in the recognition of the splice sites. The Xenopus tRNA endonuC:ease has been shown to recognize a crucial structural element in tRNA introns (Baldi et al. , 1992). Yeast tRNA introns contain a conserved purine residue three Cell 850 nucleotides upstream of the 3' splice site. This base must be able to pair with a pyrimidine at position 32 in the anticodon loop for the intron to be recognized by either yeast or Xenopus endonuclease (Figure 1A) (Baldi et al., 1992; Di Nicola Negri et al. , 1997). These data suggested that there could be different requirements for the recognition of the 5 ' and 3' splice sites. More recent evidence in this paper and the accompanying paper by Di Nicola Negri et al. (1997) supports the hypothesis that the endonuclease contains two independent active sites with different substrate recognition mechanisms. The mechanism of intron removal in tRNA splicing has also been investigated through purification of the endonuclease enzyme from yeast (Rauhut et al., 1990). Endonuclease behaves as a membrane-bound protein present in a low abundance of ~ 100 molecules per cell (Rauhut et al., 1990). The purified enzyme was seen to contain three subunits of 54, 44, and 34 kDa and, as we report here, a fourth subunit of 15 kDa. The molecular weight of the holoenzyme as judged by gel exclusion chromatography implies that the enzyme is an ai3-y8 tetramer. In this paper, we report the primary sequences of all four of the endonuclease subunits. Analysis of the peptide sequence of the endonuclease subunits presented here sheds light on the architecture of the enzyme and on its possible interactions with other components of the nucleus. The pre-tRNAs of Archaea are similar to those of Eucarya in that they contain introns that must be removed by protein enzymes in trans. Superficially, these introns are similar to yeast tRNA introns in that they are often (though not always) located in the anticodon loop. Removal of tRNA introns by either eucaryal or archaeal endonucleases yield 2' ,3' cyclic PO 4 and 5' -OH termini. Studies of removal of tRNA introns in Archaea have, however, shown notable differences between the yeast and archaeal tRNA endonucleases. Yeast pre-tRNAs are not substrates for the Haloferax volcanii endonuclease (Palmer et al., 1994), and the mature domain of a Halobacterium volcanii pre-tRNA is not required for correct excision of the intron (Thompson and Daniels, 1988). Rather, archaeal tRNA splice sites must be located in bulged loops separated by a helix of four base pairs (Figure 1C). This evidence would seem to suggest a different origin of the splicing process in these two kingdoms. The results presented here, however, together with those in the accompanying paper by Kleman-Leyer et al. (1997) , establish that two of the yeast endonuclease subunits define a gene family that includes the single homodimeric 40 kDa H. volcanii endonuclease as well as related genes from other Archaea. We discuss the evolutionary and mechanistic implications of these findings. Results The Gene for the 44 kDa Subunit, SEN2 Our primary goal in the characterization of the yeast tRNA splicing endonuclease has been to clone the genes for each of the subunits of this enzyme. We previously reported the isolation and characterization of the cold-sensitive sen2-3 mutation, an allele of the SEN2 (§plicing endonuclease 2) gene found by Winey and Culbertson (1988) in their search for yeast tRNA mutants defective in tRNA splicing in vitro. At the nonpermissive temperature, sen2-3 accumulates " 2/3" pre-tRNA molecules cleaved at the 3' splice site only (Ho et al., 1990), strongly suggesting that Sen2p is a component of endonuclease. In this paper, we report the cloning of the SEN2 gene by complementation of the sen2-3 mutant and the determination of its nucleotide sequence (GenBank accession #M32336). Analysis of the amino acid sequence of Sen2p reveals the presence of a probable transmembrane sequence between residues 225 and 243 (Figure 2B). During purification, endonuclease behaves as a membrane-bound enzyme (Rauhut et al., 1990). None of the other subunits (see below) contain a convincing transmembrane sequence, so it is possible that Sen2p is the subunit that anchors the endonuclease complex to the nuclear membrane. Sen2p also contains a domain predicted to form a canonical antiparallel coiled-coil structure between residues 122 and 184 (Figure 2B). Sequencing of the mutant sen2-3 allele reveals a single point mutation causing the change of glycine 292 to glutamate (Figure 2B). A New Purification Strategy for the tRNA Splicing Endonuclease We had previously developed a chromatographic procedure for the purification of endonuclease that resulted in homogenous enzyme (Rauhut et al., 1990) and provided sufficient material for the characterization of two of the endonuclease subunits. However, this procedure was C Yeast Splicing Substrate Consensus Yeast Endonuclease Ruler Model Archaeal Splicing Substrate Consensus Figure 1. Conserved Sequence and Structure Elements of Yeast Pre-tRNA and Archaeal Minimal Pre-tRNA (A) A composite, modified from Lee and Knapp (1985), of yeast pretRNA sequences (Ogden et al., 1984). {X) = unconserved base in a region of variable length or secondary structure; (0) = an unconserved base in a region of conserved secondary structure. {G), {C) , and (A) indicate conserved bases conserved among pre-tRNAs. M and {R) = conserved pyrimidines and purines, respectively. The conserved intron-anticodon tertiary interaction is depicted by a line between the conserved R and Y bases. (B) Taken from Reyes and Abelson (1988) . Presumed tertiary structure of pre-tRNAP"• {from Lee and Knapp , 1985). Endonuclease is proposed to interact with the mature domain and measure the distance to the two splice sites. {C) A composite of archaeal pre-tRNA sequences, modified from Thompson and Daniels (1988). Position designations are as in {A). This represents the minimal substrate for cleavage by the archaeal endonuclease. tRNA Splicing Endonuclease of Yeast 851 A 8 Sen54p Crude Exi'act t 144 TnionX-100 I100,oooxg 1hr Pellet s uJnalant ~ E-Sephacryl ~~w~h Hitalalt Sen34p HI Sen15p 11 24 aa ~ a r in Hyper-O Flow- O tM..-0.8M NaCl Gradient rtvough ~ T A Agarose Flowttv-ough • Ii FIi Sen2p H I 1467 aa I377aa Jl3oo aa t50m M IE1,=e [,t-FLAG Mab Agarose ♦ Flowthrough ♦ FLAG Pephde Elut10n ~ C MW 200k0 A R 116k0 97k0 66k0 55k0 36k0 31k0 21k0 14k0 11-Sen54 11-Sen2 11-Sen34 A A A R R R 11-Sen15 A R --- - • 6k0 Figure 2. Affinity Purification of Endonuclease from Yeast Allows Isolation of the Genes Encoding the Four Subunits of the Endonuclease {A) Purification procedure for isolation of affinity-tagged endonuclease from yeast strain CRT44 (see Experimental Procedures for details). (B) Primary amino acid sequence of each of the endonuclease subunits depicted graphically. Closed boxes represent bipartite or basic nuclear localization sequences. Hatched box in Sen2p and Sen34p represents the homologous region between these two subunits and endonuclease from Archaea {see text for details). Boxed (C) in Sen2p represents putatative coiled-coil as predicted by COILS progam of Lu pas et al. (1991 ). Dot above Sen2p represents G292E mutation. Shaded box in Sen2p represents putative transmembrane domain as predicted using the KyteDoolittle algorithm and a window of 19 amino acids. (C) Left panel: silver stained gel of lane A, 0.5 µg of affinity tagged endonuclease purified as in Figure 2A; lane R, 0.5 µg of endonuclease purified by Rauhut et al. (1990). {MW) represents molecular weight standards (Novex), which are labeled in kilodaltons. Duplicate blots of gel from left panel were probed with 1:500 dilution of antibodies to Sen54p, Sen2p, Sen34p, and Sen15p, respectively. extremely time consuming and gave a low yield of enzyme. We have developed a highly efficient affinity purification procedure for large-scale preparation of the enzyme based on the cloned SEN2 gene. The SEN2 gene was modified by addition of an aminoterminal peptide tag consisting of 8 histidine residues followed by the 4 residue FLAG epitope. This gene, cloned in a yeast 2µ vector, complemented the haplolethal phenotype of a SEN2 deletion, proving that the affinity tag did not affect the activity of Sen2p in vivo. The complemented sen2 deletion strain , CRT44, was used as a source of tagged enzyme for purification by a procedure utilizing standard chromatographic methods and two specific affinity steps, as diagrammed in Figure 2A. This rapid procedure resulted in an efficient (10 %20%) yield of 5 µg of pure enzyme per 150 grams of cells. A Fourth Subunit in the tRNA Splicing Endonuclease Pure endonuclease preparations had previously been characterized by SDS-PAGE and silver staining, which detected three subunits with apparent molecular masses of 51 kDa, 42 kDa, and 31 kDa (Rauhut et al., 1990). To compare endonuclease purified by the affinity method to that obtained in our original purification, we employed a high resolution gradient gel procedure. This procedure revealed that both preparations contain proteins of 54, 44, and 34 kDa (cf. lanes A and R, Figure 2C), although the 44 kDa band in lane A has a lower mobility due to its affinity tag (the molecular weights of the proteins are designated here as determined by comparison with markers in gradient gel electrophoresis). Both preparations contained, in addition, a fourth band with an apparent molecular weight of 15 kDa. This band would have run with the dye front in our original procedure. The fact that all four proteins are present in stoichiometric amounts in endonuclease purified by different purification procedures and that all four proteins cosediment in glycerol density gradients (data not shown) strongly suggests that all four polypeptides are subunits of the enzyme. The molecular mass of the enzyme as measured by gel filtration chromatography in our earlier study was 140 kDa (Rauhut et al. , 1990). With the fourth subunit, the predicted mass of a 1 :1 :1 :1 endonuclease complex is 146 kDa, very close to the observed molecular mass. It is therefore likely that the subunit constitution of this enzyme is cxj3-yo. The Genes for the 34 kDa, 54 kDa, and 15 kDa Subunits Purified endonuclease (50 µg) produced by the original Rauhut procedure was fractionated by SOS-PAGE, and individual bands were excised from the gel and digested Cell 852 GAL4 Activation Domain Fusion SEN 15 SEN34 SEN2 No Insert GAL4 DNA Binding Domain Fusion Figure 3. Two-Hybrid Matrix Analysis of Endonuclease Subunits Activation of histidine reporter gene by various hybrid combinations of Gal4 binding domain and Gal4 activation domain to the subunits of endonuclease. Reporter strain was transformed, and then duplicate colonies from each were picked to a new selective plate. This plate was then replica plated to plates lacking histidine and containing 5 mM 3-aminotriazole (to suppress background growth) to determine activation of the reporter gene. with lysyl-protease. A peptide resulting from digestion of the 34 kDa subunit yielded the sequence FIAYPGDPL RFX(R?)XLTIQ, uniquely found in the open reading frame (ORF) YAR008w on S. cerevisiae chromosome I. This ORF encodes a protein with a predicted molecular weight of 34 kDa. We have renamed this gene SEN34. Further evidence described below confirms that this protein is the 34 kDa subunit of the endonuclease. Endonuclease (5 µg) purified by the affinity method described above was also fractionated by SDS-PAGE, and the 54 kDa band was digested with lysyl-protease. One unambiguous peptide sequence was obtained, NDDLQHFPTYK, which matches the sequence encoded by ORF YPL083c of S. cerevisiae chromosome XVI. This 467 amino acid protein has a predicted molecular weight of 54 kDa. We have renamed this gene SEN54. Sen54p is overall a very basic protein (pl = 9.5) that also contains several highly acidic regions. Further evidence described below confirms that this protein is a subunit of endonuclease. Using a variation of the affinity purification method (see Experimental Procedures), several micrograms of pure 15 kDa subunit were prepared from SDS-PAGEfractionated endonuclease and analyzed by trypsin digestion, followed by LC-MS/MS analysis resulting in identification of three peptides, two of which were identical (except for a possible acetylation modification of the methionine) to the peptide sequence MATTDIISLVK, suggesting that this peptide is the amino terminus of the protein (K. Swiderek, personal communication). The two unique peptides are found in the protein encoded by ORF YMR059w on chromosome XIII. This ORF potentially encodes a protein of 17 kDa. However, if the methionine codon at position 20 in this ORF is the translation start site for this protein, as suggested by the MS/MS analysis, the resulting protein would be predicted to have a length of 124 amino acids with a molecular weight of 15 kDa, in agreement with the size determined by SOS-PAGE. We have renamed this gene SEN15. Further evidence described below confirms that this protein is a subunit of endonuclease. Sequence analysis of all four subunits revealed the presence of basic and bipartite nuclear localization sequences of the type described by Kalderon et al. (1984) and Robbins et al. (1991) (Figure 28, closed boxes). Chromosomal disruptions of each of the four endonuclease subunit genes are lethal in haploid yeast cells, indicating that all of their gene products are essential for vegetative growth , and most likely for the tRNA splicing process (data not shown). Antibodies to the Subunits of Endonuclease To unambiguously determine that each of these genes encodes a subunit of endonuclease, we prepared polyclonal antisera specific for each of the four gene products expressed in E. coli. A Western blot experiment was performed in which purified endonuclease was fractionated by SOS-PAGE and individually probed with antisera specific for each of the four gene products. As shown in Figure 2C, probing of endonuclease purified by the affinity procedure (lane A) , and endonuclease purified by the Rauhut procedure (lane R), reveals that each of the antisera reacts specifically with the appropriate polypeptide. The antibody to Sen15p was produced to a histidine/FLAG-tagged antigen, and as can be seen , the polyclonal antisera also contains antibodies to the affinity tag present on Sen2p and an apparent degradation product of Sen2p (Figure 2C, lane A; compare a-Sen15 to a-Sen2). Since the two preparations of pure endonuclease were prepared by different purification protocols, these results strongly support the conclusion that each of these four genes encodes a subunit of endonuclease. Two-Hybrid Analysis of the Four Subunits of Endonuclease To begin to understand the functional interactions of the endonuclease subunits, we have tested the four subunit genes in a two-hybrid matrix experiment. Each of the four genes was translationally linked to either a DNA-binding or transcriptional activation domain for analysis in the Fields two-hybrid system (Fields and Song, 1989; Bartel and Fields, 1995). Figure 3 shows the results of combinatorial expression of subunit gene pairs using rescue of histidine auxotrophy as an assay of interaction. Two interactions were detected in vivo: interaction between Sen2p and Sen54p, and between Sen34p and Sen15p. The reciprocal interaction between Sen2p and Sen54p did not activate expression, indicating that this interaction is sensitive to the orientation of the subunits relative to the binding domain/activation domains. Equivalent results were obtained with a 13-galactosidase reporter gene (data not shown). Thus, we have observed interactions between Sen34p and Sen1 5p and between Sen2p and Sen54p, but the other possible pairwise interactions were not observed. Sen2p and Sen34p Are Members of a Gene Family of Endonucleases A search of the GenBank database (version 97) for sequences similar to the four endonuclease subunit genes did not reveal any potential homologs. However, similarity was detected between the Sen2 and Sen34 proteins themselves (see Figure 28). The similarity occurs in a region of approximately 130 amino acids (see below). A sequence profile of the region of similarity between tRNA Splicing Endonuclease of Yeast 853 Figure 4. Sequence Alignment between the Archaeal Endonucleases and Two Subunits of the Yeast Endonuclease Identifies a Family of tRNA Splicing Endonucleases Sequence alignment between endonuclease identified by Daniels and coworkers from H. volcanii and M. thermoautotrophicum, endonuclease from M. jannaschii, putative endonuclease from P. aerophilum , putative endonuclease from Z. mays, and two subunits, Sen34p and Sen 2p, of the yeast tRNA splicing endonuclease. Sequence alignment was performed using the GCG program Pileup. these proteins was used to search the translated GenBank database and was found to match one sequence, a 205 amino acid ORF contained in a presumed intron of an HMG-like gene of Zea mays (Yanagisawa and lzui, 1993). C. Daniels and colleagues have determined the nucleotide sequence of the gene for the tRNA splicing endonuclease of the Archaea H. volcanii and Methanobacterium thermoautotrophicum (Kleman-Leyer et al., 1997). Homologs of this gene occur in other Archaeon, Pyrobaculum aerophilum and Methanococcus jannaschii (Bult et al., 1996). To determine if the M. jannaschii homolog is a functional homolog of the H. volcanii endonuclease, the gene product was expressed in E. coli, purified, and shown to cleave intron-containing archaeal pre-tRNAs (C. R. T. , unpublished data). Interestingly, the endonucleases of the Archaea align with the yeast endonuclease subunits, Sen2p and Sen34p, and the ORF of Z. mays (Figure 4). In all, with sequences for seven members of this gene family available, a convincing homologous core domain of 50 amino acids is now apparent with lesser regions of homology extending over 130 amino acids. Two Active Sites in Endonuclease The sequence similarity between Sen2p and Sen34p suggests that these proteins have a similar function. It seems that this function is very likely to be the catalysis of tRNA cleavage, since these proteins belong to a family that includes the homodimeric tRNA splicing endonuclease of H. volcanii. Interestingly, members of the endonuclease family found in the genomes of M. thermoautotrophicum and M. jannaschii are both small proteins made up entirely of the sequence domain found in the yeast and H. volcanii endonucleases. This further supports the notion that the sequence alignment reveals the catalytic core of the endonuclease enzyme. Previous results have indicated that Sen2p carries the active site for 5' splice site cleavage. Ho et al. (1990) showed that the sen2-3 mutation results in endonuclease that is defective in cleavage only at the 5' splice site of tRNA precursors in vivo (Ho et al. , 1990). As reported here, the glycineto-glutamate change in this allele is located at position 292, within the domain conserved between the eucaryal and archaeal endonucleases. This residue itself is not conserved in the endonuclease family, as perhaps expected for a conditional lethal mutation. Given that Sen34p is homologous to Sen2p and to the archaeal endonucleases we predicted that the Sen34p contains the active site for 3' splice site cleavage. This prediction led to the model shown in Figure 5 in which the 5' splice site is cleaved by Sen2p and the 3' splice site by Sen34p. This model clearly predicts that endonuclease containing a mutant Sen34p subunit should fail to cleave the 3' splice site of pre-tRNA while retaining full activity in cleavage of the 5' splice site. To test this prediction, we changed the conserved histidine at position 242 of SEN34 to an alanine residue (H242A). Extrachromosomal expression of a HIS/FLAG-tagged mutant Sen34p H242A in wild-type haploid cells had a dominant deleterious effect, increasing the doubling time from 2.5 hr to 6 hr in minimal media. Extrachromosomal expression of wild-type Sen34p in the same strain had little or no effect on growth. Endonuclease was purified from each of these strains by the affinity purification method described in the Experimental Procedures to select only endonuclease complexes containing the tagged Sen34p subunit. tRNA splicing assays were performed with endonuclease containing wild-type Sen34p and Sen34p H242A. Endonuclease containing affinity-tagged wildtype Sen34p cleaves pre-tRNNh• normally, indicating Figure 5. Two Active Site Model for Cleavage of Pre-tRNA by Yeast tRNA Splicing Endonuclease Yeast pre-tRNNh• tertiary structure is depicted with a schematic diagram of the yeast endonuc lease. In this model, Sen2p and Sen34p each contain an active site for cleavage of the 5' and 3 ' splice sites, respectively. The Sen2p-Sen54p and Sen34p-Sen15p interactions, detected in the two-hybrid analysis, are depicted as well as the potential tRNA recognition mediated by Sen54p {see text for details). Cell 854 ..--.r ~ - lntron-3'exon • - lntron Figure 6. tRNA Splicing Activity of Sen34p H242A-Containing Endonuclease Yeast pre-tRNAe,, was incubated without endonuclease (pre) or with endonuclease purifi ed from HIS/FLAG-tagged Sen2p (WT) , HIS/ FLAG-tagged wild-type Sen34p (Sen34p), or HIS/FLAG-tagged Sen34p H242A (Sen34p H242A) expressing yeast. Pre-tRNA was incubated for 5, 10, and 15 minutes, and the resulting cleavage products (shown to the right) were fractionated on a 10% polyacrylamide gel. The amount of endonuclease used in each time course was similar based on silver staining (wild-type enzyme was used at 1 oo x concentration to allow for complete digestion of pre-tRNA). that the affinity-tagged enzyme functions as wild-type endonuclease (Figure 6). At the low enzyme concentrations used in this experiment, a slight accumulation of a 2/3 molecule consisting of the intron joined to the 3' exon , derived from cleavage at the 5' splice site only, is normally seen (Miao and Abelson , 1993). In contrast, endonuclease containing Sen34p H242A shows a marked accumulation of 5' exon and the intron-3 ' exon 2/3 molecule. There is a nominal amount of 3' exon and intron product, but we cannot determine if this is derived from retention of some 3' splice site cleavage activity in Sen34p H242A or from the presence of some wildtype endonuclease in the purified mutant extract. However, the assay clearly demonstrates that endonuclease containing Sen34p H242A is, as predicted by the model, impaired in 3' splice site cleavage. Discussion Yeast tRNA Endonuclease Contains Two Active Sites The sequence similarity between Sen2p and Sen34p and their homology to the archaeal tRNA splicing endonucleases implied that each of these two subunits contains an active site for tRNA cleavage. As described above, endonuclease containing Sen34p H242A appears to be significantly impaired for 3' splice site cleavage, but 5' splice site cleavage occurs normally. A complete blockage of 5' splice site cleavage was seen in endonuclease containing sen2-3 (Ho et. al., 1990). These results taken together provide strong support for the model in Figure 5, indicating that the yeast tRNA splicing endonuclease contains two distinct, functionally independent active sites. Sen2p contains the active site for cleavage at the 5' splice site and Sen34p contains the active site for cleavage at the 3' splice site. We are currently undertaking a more extensive mutagenesis of SEN34 and SEN2 to generalize these results. Additional support for the model shown in Figure 5 is derived from previous work on the Xenopus endonuclease, in which it has been shown that the two cleavage sites in yeast tRNA precursors have different properties. The Xenopus tRNA splicing endonuclease recognition of the 3' splice site requires a base pair between a purine located 3 bases upstream of the 3' splice site and the conserved pyrimidine at position 31 in the anticodon loop (the A-I base pair) (Baldi et al., 1992). Cleavage at the 5' splice site requires only a purine as the adjacent 5' nucleotide. It is conceivable that different specificities in cleavage could be obtained by a conformational change in a single active site occurring between cleavage reactions. For this to occur, one might expect an obligatory reaction pathway, i.e., cleavage at a primary site followed by cleavage at the second site. However, we have shown that the yeast endonuclease can and does cleave the 5' and 3' splice sites in a random order (Miao and Abelson, 1993). More recently, the accompanying paper by TocchiniValentini 's laboratory has shown that accurate 3' splice site cleavage does not require the mature tRNA domain of the pre-tRNA (Di Nicola Negri et al., 1997). A small substrate containing only an intron and anticodon stem is cleaved by the Xenopus endonuclease at the 3' splice site but not at the 5' splice site. While both splice sites in an intact pre-tRNA are subject to the measuring mechanism, it appears that only the 5' splice site absolutely requires this mechanism. The H. volcanii tRNA splicing endonuclease cleaves with the same general mechanism as the eucaryal endonuclease, producing 2' ,3 ' cyclic PO 4 and 5' -OH termini, but has a different splice site recognition mechanism and fails to cleave yeast pre-tRNAs (Thompson and Daniels, 1988, 1990). The mature tRNA domain is not a requirement for the archaeal endonuclease. Archaeal tRNA splice sites are located in bulged regions separated by four helical base pairs (Figure 1 C). In the accompanying paper by Kleman-Leyer et al. (1997), the H. volcanii endonuclease is shown to exist as a homodimer, and thus contains two active sites. The view that each of these active sites is functional is strengthened by the evidence presented here for two distinct active sites in the yeast endonuclease. However, because of the near symmetry of the archaeal splice site structure, we cannot rule out that a single active site could act interchangeably. Interestingly, most archaeal pre-tRNAs appear to contain the conserved A-I base pair of eucaryotic pre-tRNAs in the context of the four base pair stem separating the bulged loops of the splice sites. In this regard , the requirements for 3' splice cleavage are more similar between Archaea and Eucarya than are those for 5' splice site cleavage. In both kingdoms the splice site must be in a single-strand bulge near a base-paired element (in the eucaryotic pre-tRNAs, it is the A-I base pair). Subunit Interactions in Endonuclease In a complete matrix of two-hybrid interactions between the four subunits of endonuclease, strong interactions were seen only between the Sen2p and Sen54p pair and the Sen34p and Sen1 Sp pair. Since all four subunits tRNA Splicing Endonuclease of Yeast 855 remain tightly associated through two different purification regimens, it seems likely that there are also interactions between the two presumed heterodimers. These interactions, however, would not be detected in a twohybrid experiment if they exist only within the context of the intact enzyme, or if three of the subunits are required for the interaction. The involvement of each of the active site subunits in a specific heterodimeric interaction suggests that the different specificities of the two active sites may be determined by these interactions. For example, the measuring mechanism, seen primarily in the recognition of the 5' splice site, may be determined through the interaction between Sen2p and Sen54p, as indicated in Figure 5. Sen54p is a very basic protein and as such might be expected to interact with the phosphate backbone of the pre-tRNA substrate, presumably in the mature tRNA domain. Sen15p could be primarily responsible for the recognition of the A-I basepair. Evolutionary Implications of the tRNA Splicing Endonuclease Gene Family The existence of homologous tRNA splicing endonuclease genes in Eucarya and Archaea suggest that this enzyme was present in their common ancestor. Since endonuclease participates in tRNA splicing in both Eucarya and Archaea, it seems likely that the original function of this enzyme was the same. The discussion of the time of appearance of introns in evolution has been a lively one, and it is safe to say that a consensus has not emerged. The simplest hypothesis consistent with the data presented here and in the accompanying papers is that the function of endonuclease in the common eucaryal-archaeal ancestor was the removal of introns from tRNA. A less parsimonious hypothesis is that the tRNA splicing function of this enzyme evolved independently in both kingdoms from an earlier, more generalized RNA processing function. In some present-day Archaea, the tRNA splicing endonuclease can apparently also function to remove introns from ribosomal RNA (Kjems and Garrett, 1985; Kjems et al., 1989; Burggraf et al., 1993). Recently, Walter and coworkers have detected an intron in a yeast transcription factor gene that is not processed by the normal mRNA spliceosomal machinery (Cox and Walter, 1996; Sidrauski et al., 1996). They have demonstrated that tRNA ligase is involved in the splicing of this intron, but the tRNA splicing endonuclease has not yet been implicated. Clearly, members of this family of proteins have a number of functions, but it is not yet possible to determine which are original and which have been co-opted later in evolution. Although the endonuclease active site is apparently a very primitive protein domain, it is clear that its substrate specificity has been altered in evolution because the recognition of the splice sites is different in Eucarya and Archaea. We speculate that the peculiarities of the eucaryal system are in part explained by the heterodimeric interaction of the divergent subunits-Sen2p with Sen54p and Sen34p with Sen15p. Many fascinating questions remain concerning this evolutionary specialization. Since Sen54p and Sen15p have no known homologs, where did these proteins come from? Do Sen54p and Sen15p play a role in splice site selection? Can the active site subunits function in conjunction with other protein subunits, thereby modifying endonuclease specificity for other RNA processing reactions? We are now able to begin to address these questions. A further complexity of the evolution of the tRNA splicing process is indicated by the lack of a yeast tRNA ligase (or T4 RNA ligase) homolog in the entire M. jannaschii genome (Bult et al., 1996; C. R. T. and E. A. A., unpublished data). Interestingly, several archaeal genomes contain a clear homolog of a bacterial RNA ligase that can form a 2' -5 ' linkage when joining tRNA halfmolecules derived from endonucleolytic cleavage of yeast pre-tRNAs (Greer et al., 1983; Arn and Abelson, 1996). The complete ligation mechanism for archaeal tRNA splicing has not yet been determined, but it has been demonstrated that an RNA splicing ligase activity present in H. volcanii extracts does not require ATP or exogenous phosphate (R. Gupta, personal communication). The E. coli 2' -5' RNA ligase is the only known RNA ligase activity that does not require a nucleotide triphosphate cofactor (Arn and Abelson, 1996). We are presently attempting to ascertain whether the archaeal 2' -5 ' RNA ligase homolog functions in tRNA splicing in vivo. Experimental Procedures Yeast Methods Yeast were transformed by the lithium acetate procedure (Gietz and Schiestl, 1995). Yeast minimal media (SD) contained Bacto yeast nitrogen base without amino acids (YNB - aa, 6.7g/I), dextrose (2% w/v), and a supplemental amino acid mix (Bio101) lacking various amino acids for auxotroph selection . All other yeast media were prepared according to Sherman et al. (1986). For fermentation growth of CRT44, SD-histidine-tryptophan, prepared as described above, was used and supplemented with 25 I-Lg/ml Kanamycin. The fermentation was fed using SD + 10 X -histidine-tryptophan amino acid mix with constant flow for 24 hr. Two-hybrid strains were obtained by a generous gift from P. Bartel of S. Fields laboratory and are described in Table 1. All two-hybrid techniques used are described by Hannon et al. (1995). Plates used for analysis contained 5 mM 3-aminotriazole that suppressed background growth associated with histidine reporter assay (Hannon and Bartel, 1995). Strains E. coli strains utilized include HB101 and XL 1-Blue (Novagen) and DH5 (GIBCO-BRL). S. cerevisiae strains are described in Table 1. Knockout strains YPH274LI.SEN2 and CRT54 were created utilizing methods described in Guthrie and Fink (1991 ). Knockout strains of SEN15 (CRT1 5) and SEN34 (CRT34) were created utilizing methods and plasmids generously obtained from and described by Wach (1996). All knockouts were confirmed by PCR analysis, and each knockout was complemented with the corresponding ORF only. Strain CRT44 was prepared by transforming YPH274LI.SEN2pC10 with YpH8/FLAG SEN2, plating on selective media (SD-His-Trp-Ura), and streaking out isolates on selective media (SD-His-Trp) + 5 FOA to select for loss of pC10 (URA3 CEN/ARS). Plasmid Preparation All restriction enzymes used were obtained from New England BioLabs and Boehringer Mannheim and used in accordance with the manufacturers' instructions. Restriction digests, DNA ligations, and electrophoresis of DNA were performed by standard methods of Sambrook et al. (1990). pC10 was prepared by subcloning an EcoRV/EcoRI fragment of DNA derived from the YcP50 sen2-3 complementing clone, into the vector pRS416 (URA3)(Sikorski and Hieter, 1989). This fragment Cell 856 Table 1. Yeast Strains Strain Genotype Source/Reference J051 Mat ala; ura3-52/ura3-52; leu2-3, 112/leu2-3, 112; trp1t.63/trp1t.63; lys2-801 /lys2-801 ; his3t:,.200/his3t.200 Mat ala; ura3-52/ura3-52; leu2t.1 /leu2t.1 ; lys2-801 /lys2-801trp 1t.1 ltrp tt.1 ;; his3t.200/his3t.200; ade2-101' /ade2-101 ' Mat a; ura3-52; leu2 tH; /ys2-801 ; trp 1t.1; his3t.200; ade2-101' ; SEN2::HIS3 and YpH8/FLAG SEN2 Same as YPH274t.SEN2 except contains YpHB/FLAG SEN2 Same as J051 except SEN54::HIS3/SEN54 Same as J051 except SEN34::kanMX6/SEN34 Same as J051 except SEN15::S. pombe H/S3/SEN15 Mat a; ura3-52; /eu2-3, 112; /ys2-801 ; trp1-901 ; his3t.200; ade2-101' ; gal4-9542; galB0-538; LYS2::GAL1u•s-GALw.-HJS3; URA3::GAL4 17""""rCYC1 w. -lacZ Mat a; ura3-52; /eu2-3, 112; /ys2-801; trp1-901 ; his36.200; ade2-101' ; gal4-9542; galB0-538; URA3::GAL4 17me~rCYC1 r•r•-/acZ A. Varshavsky YPH274 YPH274t.SEN2 CRT44 CRT54 CRT34 CRT15 HF7c Y527 contains an open reading frame coding for the 377 amino acid SEN2 gene plus 250 base pairs in the promoter region and 345 base pairs of downstream region. pC10 was capable of complementing a knockout of the SEN2 ORF. YpHB/FLAG was prepared by PCR of the CUP1 promoter, from J051 , to allow cloning of the CUP1 promoter into the Hindlll/EcoRI site of pBluescript (Stratgene). This plasmid was then digested with Xbal, filled in, and cut with Pstl , and the Pstl-Nael fragment from pGBT9 (from S. Fields). containing the ADH1 terminator sequence, was cloned into this site, yielding pBSCUP/TERM. This plasmid was then cut with Hindlll/Sacl , and the resulting fragment was subcloned into pRS424 (Sikorski and Hieter, 1989). to yield YpCUP/TERM. This plasmid was cut with Ndel and EcoRI , and the HIS/FLAG linker was cloned into this site, generating YpHB/FLAG . YpHB/FLAG SEN2 was prepared by digestion of pGBT9-SEN2 with EcoRI and Pstl , gel isolation of the fragment containing SEN2, and subcloning into EcoRI/Pstl-digested YpH8/FLAG . YpHBFLAG SEN34 was prepared by digestion of pGBT9-SEN34 with EcoRI and Pstl , gel isolation of the fragment containing SEN34, and subcloning into EcoRI/Pstldigested YpH8/FLAG. Oligonucleotide-directed mutagenesis was performed on single-stranded YpH8/FLAG-SEN34 to effect an H242A mutation according to the protocols provided by the manufacturer (Amersham). Two-hybrid vectors were prepared by cloning PCR products of SEN15 (oligos SEN15-N 5' -GGAATTCGCAACGACAGATATCATATC-3 ' and -C 5'-TTCTGCAGATCTTCAATTTCTTTTCGGTTTTCG-3'), SEN54 (oligos SEN54-N 5'-GGAATTCCAATTCGCTGGGAAG-3' and-C 5'-TTC TGCAGTTAATGCTTCTTCCATCTTTG-3 '), SEN2 (oligos SEN2-N 5' -GGA ATTCATGTCTAAAGGGAGGG and -C 5'-CGGGATCCCAATGAAA AGTTTAGAGC) and SEN34 (oligos SEN34-N 5'-CGGAATTCGGG ATCCATATGCCACCGCTAGTATTTGAC-3 ' and -C 5' -CGGGATCCC GGGAAGCTTAACCAAATCCAGCCC-3') digested with EcoRI and BamHI and cloned into EcoRI/BamHl-digested pGBT9 and pGA0424 (Hannon and Bartel, 1995) (these constructs were sequenced). Antibodies and Western Blots Polyclonal antibodies to Sen34p were produced by injection of rabbit (CoCalico Biosciences) with peptides NH,-OKTTAEELQRLOK SSC-COOH and NH,-DIRSOSOSLSROOIC-COOH. Polyclonal antibodies to Sen2p were produced by injection of rabbits with peptide derived from the C terminus of Sen2p (NH ,-MSKGRVNQKRYPLCCOOH). All peptides were produced by the Caltech Peptide Facility. The antibodies were used directly in rabbit serum at the indicated dilution. Polyclonal antibodies to both Sen54p and Sen1 5p were prepared in the following manner: antigen was derived from purification of HISS/FLAG-tagged protein from E. coli. Each gene was cloned into the EcoRI/Pstl site of YpH8/FLAG by subcloning from pGBT9 SEN54 and pGBT9 SEN15. Each construct was then cut with Ndel/Pstl and subcloned into Pet11 a. Purification was carried out under denaturing conditions by Ni-NTA chromotography as suggested by Qiagen . Protein eluted from the Ni-NTA resin was gel purified by Prep-Cell Chromatography (Biorad) per the manufacturer's suggested protocol with elution into buffer containing 0.05% P. Heiter This study This study This study This study This study Feilotter et al. , 1994 Bartel et al., 1993 SOS. The eluate was directly injected into Swiss Webster mice per the protocol of Susan Ou of the Caltech Monoclonal Antibody Facility. SOS-PAGE was performed on 10%-20 % gradient gels (Biorad). Proteins were transferred to PVOF (lmmobilon) by semidry transfer (Biorad) . Western blots were performed as described by Harlow and Lane (1988) and were visualized by enhanced chemiluminescence (EGL, Amersham) (Harlow and Lane, 1988). Endonuclease Assay Endonuclease was assayed by incubation of 2 µ I of extract for a specified amount of time at 30' C in a 10 µI reaction containing 20 mM Na-HEPES (pH 7.5), 5 mM MgCI,, 2.4 mM spermidine-HCI (pH 7 .5), 0.1 mM OTT, 0.4 % Triton X-100, and 10 fmol pre-tRNA""' bodylabeled with [a- 32 P]UTP (specific activity ~ 500 cpm per fmol tRNA) by T7 polymerase transcription according to Sampson and Saks (1993) . RNA products were phenol/chloroform extracted , precipitated, and fractionated on a 10% polyacrylamide gel containing BM urea. Gels were exposed to phosphorimager for 8 hr. Purification of tRNA Endonuclease Purification of endonuclease was carried out utilizing three separate purification strategies. The first is described by Rauhut et al. (1990). This protein was used to obtain the gene for the 34 kOa subunit. Affinity purification was carried out utilizing yeast strain CRT44 grown by fermentation at 30' C. The cells were harvested and stored at - 70' C. All subsequent steps were performed at 4' C. Yeast cells (150 g) resuspended in 1.5 x breaking buffer (37.5 mM Tris [pH 8.0]. 15% (v/v) Glycerol, 5 mM B-ME, 1 mM EOTA, 0.1 % (v/v) Triton X-100, 405 mM NH,SO,) to 300 ml was ground in a Bead-Beater (BioSpec Corp) for 20 X1 minute intervals, and Triton X-100 (BoehringerMannheim) was added to 1 %. The extract was incubated for 1 hr and then centrifuged (40,000 rpm x 1.5 hr). The supernatant was loaded onto a 300 ml OEAE-Sepharose CL-6B equilibrated in 1 x breaking buffer + 1 % Triton X-100. OEAE effluent was diluted approximately 3-fold and loaded onto a 100 ml Heparin-HyperO (BioSepra) column equilibrated in buffer A (25 mM MOPS (pH 7.1], 10% Glycerol, 1 mM EOTA, 5 mM BME, 0.9% Triton X-100, 100 mM NaCl), at 1 ml/minute. The column was washed with 500 ml buffer B (25mM Tris [pH 8.0], 10% Glycerol, 5mM BME, 0.3% Triton X-100, 150 mM NaCl). A 2 x 200 ml gradient of 0.1-0.8 mM NaCl in buffer B was applied . Active fractions were pooled and bound in batch to 5 ml of Ni-NTA agarose (Quiagen, equilibrated in buffer B) for 1.5 hr. The suspension was poured into a 2 x 10 mm column and washed extensively with 5 ml of each of the following buffers: buffer N (25mM Tris-HCI [pH 7.5]. 150 mM NaCl, 10% Glycerol, 5 mM B-mercaptoethanol, 0.3% Triton X-100) + 650 mM NaCl + 15 mM imidazole; buffer N + 650 mM NaCl + 20mM imidazole; buffer N + 20 mM imidazole; buffer N + 40 mM imidazole. Endonuclease was eluted in Buffer N + 150 mM imidazole. The eluate was loaded onto a 1 ml M2 monoclonal anti-FLAG antibody column (Kodak) equilibrated in Buffer N. The column was washed in buffer N and tRNA Splicing Endonuclease of Yeast 857 eluted with Buffer N + 150 µ g/ml FLAG peptide. This endonuclease was used to prepare the 54 kDa subunit. A modified purification protocol was developed to prepare the 15 kDa subunit and purification of mutant containing endonuclease. Yeast cells {strain CRT44 or as indicated in the text) were resuspended and broken as above. The extract was then spun at 40,000 rpm x 2 hr to pellet the nuclear membranes. The pellet was resuspended in 50 ml of Buffer R (50 mM Tris [pH 7.5], 150 mM NaCl, 5% Glycerol, 1 mM EDTA, 5 mM BME, 0.9% Triton X-100) and incubated for 1 hr. The extract was subjected to centrifugation as before and the supernatant was loaded directly onto a 2.5 ml M2 monoclonal anti-FLAG antibody column. The column was washed with buffer R, and endonuclease was eluted with Buffer S (30 mM Tris [pH 8.0], 500 mM NaCl, 10% Glycerol, 5 mM BME, 0.3% Triton X-1 00, 5 mM lmidazole) containing 150 µg/ml FLAG peptide. The elution was incubated with 1 ml of Ni-NTA agarose, equilibrated in Buffer S, for 1.5 hr and poured into a 3 ml column . The column was washed with Buffer S + 40 mM imidazole, and endonuclease was eluted in Buffer S + 150 mM imidazole. Protein Sequencing Three separate procedures were utilized for sequencing the subunits of endonuclease. In all three cases, protein was submitted as a gel slice with various amounts of protein in each. For the 34 kDa subunit, 25 µg of protein was submitted to the protein sequencing facility at Cold Spring Harbor and was digested in-gel with lysylendoprotease, followed by extraction and HPLC purification of the resulting peptides following the protocols of the facility {Dan Marshak, personal communication). Fractions containing single peaks were sequenced by Edman degradation. For the 54 kDa subunit, 5 µ g of protein was digested in-gel with trypsin as described by Hellman et al. (1995) {Hellman et al., 1995), except that 0.3 µ g of Promega modified pig trypsin and speedvac drying of the gel slice was used. Peptides were HPLC purified, and fractions containing single mass peaks, as determined by mass spectrometry on a Voyager Mass Spec, were submitted for Edman Degradation sequencing on an Applied Biosystems sequencer {Gary Hathaway, personal communication) . For the 15 kDa subunit, 5 µg of protein was submitted to the protein sequencing facility at the City of Hope {Duarte, CA) for ingel trypsin digestion, followed by peptide sequencing utilizing a Beckman LC/MS-MS protein sequencer {K. Swiderek, personal communication). Data obtained was used to search the yeast genome database. Acknowledgments Correspondence regarding this paper should be addressed to J . N. A. {abelsonj@starbase1 .caltech .edu). We wish to acknowledge first and foremost the protein sequencing work done by Gary Hathaway and Dirk Krapf at Caltech Protein Analysis Laboratory, Dr. Terry Lee and Dr. Kristine Swiderek at the City of Hope Protein Sequencing Facility, and Dr. Dan Marshak at the Cold Spring Harbor Protein Sequencing Facility. Special thanks to Paul Bartel and Stan Fields for two-hybrid strains and plasmids. We also want to thank Chuck Daniels for communicating the endonuclease sequences of H. volcanii and M. thermoautotrophicum prior to publication and Ors. Mel Simon and Ung-Jin Kim for sequence of the P. aerophilum homolog. We wish to thank Ors. Christine Guthrie, Eric Phizicky, Chris Greer, and Anita Hopper for their insight and critical review of the manuscript and members of the Abelson laboratory for their advice and support. This work was supported by generous grants from the American Cancer Society and National Research Service Award {to C. R. T.). Received February 10, 1997; revised May 5, 1997. References Abelson , J. (1991 ). RNA splicing in yeast. Harvey Leet. 85, 1-42. Arn , E.A., and Abelson , J .N. (1996). The 2'-5 ' ligase from Escherichia coli . J. Biol. Chem. 271 , 31145-31153. Baldi, M.I., Mattoccia, E., Bufardeci, E. , Fabbri , S., and TocchiniValentini , G.P. (1992). Participation of the intron in the reaction catalyzed by the Xenopus tRNA splicing endonuclease. Science 255, 1404-1408. Bartel, P.L., and Fields, S. (1995). Analyzing protein-protein interactions using 2-hybrid system. Meth. Enzymol. 254 , 241-263. Belford, H.G., Westaway, S.K., Abelson, J., and Greer, C.L. (1993). Multiple nucleotide cofactor use by yeast ligase in tRNA splicing : evidence for independent ATP- and GTP-binding sites. J. Biol. Chem . 268 , 2444-2450. Biniszkiewicz, 0., Cesnaviciene, E. , and Shub, D.A. (1994). Selfsplicing group-I intron in cyanobacterial initiator methionine tRNA: evidence for lateral transfer of introns in bacteria. EMBO J. 13, 4629-4635. Bult, C.J., White, 0 ., Olsen , G.J., Zhou, L.X., Fleischmann, R.D., Sutton, G.G., Blake, J .A., Fitzgerald , L.M ., Clayton, R.A. , Gocayne, J.D., et al. (1996). Complete genome sequence of the methanogenic archaeon, Methanococcus jannaschii. Science 273, 1058-1073. Burggraf, S., Larsen, N., Woese, C.R., and Stetter, K.O. (1993). An intron within the 16S rRNA gene of the archaeon Pyrobaculum aerophilum. Proc. Natl. Acad. Sci. USA 90, 2547-2550. Cech, T. (1993). Structure and mechanism of the large catalytic RNAs: group I and group II introns and ribonuclease P. In RNA World, R.F. Gesteland and J.F. Atkins, eds. {Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press), pp. 239-270. Cox, J.S., and Walter, P. (1996) . A novel mechanism for regulating activity of a transcription factor that controls the unfolded protein response. Cell 86, 391-404. Culver, G.M., McCraith, S.M., Zillmann , M., Kierzek, R. , Michaud, N., Lareau, R.D., Turner, D.H., and Phizicky, E.M . (1993) . An NAO derivative produced during tRNA splicing : ADP-ribose 1 " -2" cyclic phosphate. Science 261, 206-208. Di Nicola Negri, E., Fabbri , S., Bufardeci, E., Baldi, M.I., Gandini Attardi , 0 ., Mattoccia, E., and Tocchini-Valentini , G.P. (1997). The eucaryal tRNA splicing endonuclease recognizes a tripartite set of RNA elements. Cell, this issue. Fields, S., and Song, O.K. (1989). A novel genetic system to detect protein-protein interactions. Nature 340, 245-246. Gietz, R.D., and Schiestl, R.H. (1995). Transforming yeast with DNA. Meth. Mol. Cell. Biol. 5, 255-269. Greer, C., Javor, G. , and Abelson, J .N. (1983). RNA ligase in bacteria: formation of a 2' ,5' linkage by an E. coli extract. Cell 33, 899-906. Guthrie, C., amd Fink, G.R. (1991 ). Methods in Enzymology. Guide to Yeast Genetics and Molecular Biology. {San Diego, CA: Academic Press, Inc.). Hannon, G. , and Bartel, P. (1995) . Identification of interacting proteins using the 2-hybrid system . Meth. Mol. Cell. Biol. 5, 289-297. Harlow, E., and Lane, D. (1988) . Antibodies: A Laboratory Manual {Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press). Hellman, U., Wernstedt, C., Gonez, J., and Heldin, C.H . (1995). Improvement of an in-gel digestion procedure for the micropreparation of internal protein fragments for amino acid sequencing . Anal. Biochem . 224 , 451-455. Ho, C.K., Rau hut, R. , Vijayraghavan, U., and Abelson, J. (1990). Accumulation of pre-tRNA splicing 2/3 intermediates in a Saccharomyces cerevisiae mutant. EMBO J. 9, 1245-1252. Kalderon , 0 ., Roberts, B.L., Richardson, W.D., and Smith, A.E. (1984). A short amino-acid sequence able to specify nuclear location. Cell 39, 499-509. Kjems, J., and Garrett, R.A. (1985). An intron in the 23S rRNA gene of the Archaebacterium Desulfurococcus mobilis. Nature 318, 675--677. Kjems, J., Jensen, J., Olesen , T., and Garrett, R.A. (1989). Comparison Of tRNA and rRNA intron splicing in the extreme thermophile and Archaebacterium Desulfurococcus mobilis. Can. J. Microbiol. 35, 21 0-214. Kleman-Leyer, K., Armbruster, D.A., and Daniels, C.J. (1997) . Properties of the H. volcanii tRNA intron endonuclease reveal a relationship between the archaeal and eucaryal tRNA intron processing systems. Cell , this issue. Cell 858 Lee, M.C., and Knapp, G. (1985). tRNA splicing in Saccharomyces cerevisiae: secondary and tertiary structures of the substrates. J. Biol. Chem. 260, 3108-3115. Winey, M., and Culbertson, M.R. (1989). Split tRNA genes and thei r products: a paradigm for the study of cell-function and evolution. Yeast 5, 405-427. Lupas, A., Vandyke, M., and Stock, J. (1991). Predicting coiled coils from protein sequences. Science 252, 1162-1164. Yanagisawa, S., and lzui, K. (1993). Molecular cloning of 2 DNAbinding proteins of maize that are structurally different but interact with the same sequence motif. J. Biol. Chem . 268, 16028-16036. Madhani, H.D., and Guthrie, C. (1994) . Dynamic RNA-RNA interactions in the spliceosome. Annu . Rev. Genet. 28, 1-26. McCraith, S.M., and Phizicky, E.M. (1991). An enzyme from Saccharomyces cerevisiae uses NAO+ to transfer the splice junction 2' phosphate from ligated tRNA to an acceptor molecule. J. Biol. Chem. 266, 11986-11992. Miao, F., and Abelson, J. (1993). Yeast tRNA-splicing endonuclease cleaves precursor tRNA in a random pathway. J. Biol. Chem. 268, 672-677. Moore, M., Query, C.C., and Sharp, P. (1993). Splicing of precursors to mRNAs by the spliceosome. In RNA World , R.F. Gesteland and J .F. Atkins, eds. (Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press), pp. 303-349. Ogden , R.C., Lee, M.C., and Knapp, G. (1984). Transfer RNA splicing in Saccharomyces cerevisiae: defining the substrates. Nucl. Acids Res. 12, 9367-9382. Palmer, J.R., Nieuwlandt, D.T., and Daniels, C.J . (1994). Expression of a yeast intron-containing tRNA in the Archaeon Haloferax volcanii. J. Bacterial. 176, 3820-3823. Peebles, C.L. , Gegenheimer, P., and Abelson, J. (1983) . Precise excision of intervening sequences from precursor tRNAs by a membrane-associated yeast endonuclease. Cell 32, 525-536. Rau hut, R., Green, P.R., and Abelson, J. {1990). Yeast tRNA-splicing endonuclease is a heterotrimeric enzyme. J. Biol. Chem . 265, 1818018184. Reinholdhurek, B., and Shub, D.A. (1992). Self-splicing intrans in tRNA genes of widely divergent bacteria. Nature 357, 173-176. Reyes, V.M ., and Abelson, J. (1988). Substrate recognition and splice site determination in yeast tRNA splicing. Cell 55, 719-730. Robbins, J., Dilworth, S.M., Laskey, R.A., and Dingwall, C. (1991). 2 interdependent basic domains in nucleoplasmin nuclear targeting sequence: identification of a class of bipartite nuclear targeting sequence. Cell 64 , 615-623. Sherman, F., Fink, G.R., and Hicks, J.B. (1986). Methods in Yeast Genetics. (Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press). Sidrauski, C., Cox, J.S., and Walter, P. (1996). tRNA ligase is required for regulated mRNA splicing in the unfolded protein response. Cell 86, 405-41 3. Sikorski, R.S., and Hieter, P. (1989). A system of shuttle vectors and yeast host strains designed for efficient manipulation of DNA in Saccharomyces cerevisiae. Genetics 122, 19-27. Thompson, L.D. , Brandon, L.D., Nieuwlandt, D.T., and Daniels, C.J. (1989). tRNA intron processing in the halophilic Archaebacteria. Can . J. Microbial. 35, 3~2. Thompson, L.D. , and Daniels, C.J . (1990) . Recognition of exonintron boundaries by the Halobacterium volcanii tRNA intron endonuclease. J . Biol. Chem. 265, 18104-18111 . Thompson , L.D ., and Daniels, C.J. (1988) . A tRNA'"' intron endonuclease from Halobacterium volcanii: unique substrate recognition properties. J. Biol. Chem. 263, 17951-17959. Wach, A. (1996) . PCR-synthesis of marker cassettes with long flanking homology regions for gene disruptions in Saccharomyces cerevisiae. Yeast 12, 259-265. Westaway, S.K., Belford, H.G., Apostol, B.L., Abelson, J., and Greer, C.L. (1993). Novel activity of a yeast ligase deletion polypeptide: evidence for GTP-dependent tRNA splicing. J. Biol. Chem. 268, 2435-2443. Westaway, S.K., Phizicky, E.M., and Abelson, J. (1988). Structure and function of the yeast tRNA ligase gene. J. Biol. Chem. 263, 3171-3176. Winey, M., and Culbertson, M.R. (1988). Mutations affecting the tRNA-splicing endonuclease activity of Saccharomyces cerevisiae. Genetics 118, 609-617. GenBank Accession Number The accession number for the SEN2 nucleotide sequence reported in this paper is M32336. 163 Appendix III Reprint Series 10 Apri l 1998, Volume 280, pp. 279-284 SciENCE Crystal Structure and Evolution of a Transfer RNA Splicing Enzyme Hong Li, Christopher R. Trotta, and John Abelson* Copyright © 1998 by the American Association for the Advancement of Sc ience Crystal Structure and Evolution of a Transfer RNA Splicing Enzyme Hong Li , Christopher R. Trotta, John Abelson* The spl icing of transfer RNA precursors is simil ar in Eucarya and Archaea. In both kingdoms an endonuclease recognizes the splice sites and releases the intron , but the mechanism of spl ice site recognition is different in each kingdom. The crystal structure of the endonuclease from the archaeon Methanococcus jannaschii was determined to a resolution of 2.3 angstroms. The structure indi cates that the cleavage reaction is similar to that of ribonuclease A and the arrangement of the active sites is conserved between the archaeal and eucaryal enzymes. These results suggest an evolutionary pathway for splice site recognition . l ntrons are found in the tRNA genes of organisms in all three of the great li nes of descent: the Eucarya, the A rchaea, and the Bacteria. In Bac teria, tRNA intrans are se lf-sp licing gro up I intrans and the splicing mechanism is autocatalytic ( J ). In Eucarya, tRNA intrans are sma ll and inva riably inte rrupt the anticodon loop 1 base 3' to the anticodon. T hey are removed by the stepwise action of an endonuclease , a ligase, and a phosphotransferase (2). In Archaea, the in trons are also small and often reside in the same loca tion as eucaryal tRNA intrans (3 ). Splicing in Archaea is catalyzed by an endonuclease, but the mechanism of ligation is likely differe nt from that in Eucarya as t here is no homo log of the eucarya l tRNA splicing ligase in the complete genome sequence of several members of the Archaea (4). The tRNA splicing endonucleases of both the Eucarya and the Archaea cleave the pre-tRNA substrate leaving 5' -hydroxyl and 2',3' cyclic phosphate termini , but these enzymes recogn ize their substrates differen tly. The eucarya l enzyme uses a measuring mechanism to determine the position of the un iversally pos itioned splice sites relative to the conserved doma in of pre-tRNA (Fig. lA) (5). In Archaea, the enzyme recognizes a pseudosymmetric substrate in wh ich two bulged loops of 3 bases are separa ted by a stem of 4 base pairs (bp) [bu lge-helix-bu lge (BHB)] (Fig. lA) (6). T hese observations suggested that the tRNA splicing mechan isms of the three ma jor kingdoms evolved independently. Recently, however, characterization of eucaryal and archaeal endonucleases has revealed a possible common origin for these enzymes . The yeast endonuclease contains four distinct sub un its of 15, 34, 44, and 54 kD (7). The endonuclease from the arDivision of Biology, Mail Code 14 7-75, California Institute of Technology, Pasadena, CA 91125, USA. 'To whom correspondence should be addressed. E-mail: abelsonj@cco.caltech.edu chaeon Haloferax volcanii is a dimer of identical 37-kD subunits (8 ). The H. volcanii enzyme is homologous to both the 34- and 44-kD subunits of the yeas t enzyme, which suggests tha t these distinct subunits contain ac ti ve sites for tRNA splicing. This h ypothes is was strongly supported by the findin g that sen2 -3, a mutant in the 44-kD subunit, is se lective ly defective in 5' sp lice site cleavage (9), whereas a mutant in the 34-kD subunit, H242A, is se lective ly defective in 3' sp lice site cleavage (7). Thus the 44-kD subunit cleaves t he 5' sp li ce site and the 34-kD subun it is proposed to cleave the 3' spl ice site (7). The funct io n of the two non homo logous subun its ( the 15- and the 54-kD subuni ts) remains unclear, altho ugh the basic 54-kD subunit was suggested to embody th e molecular ru ler (7). In H. volcanii, the h omod imeric enzy me is proposed to interact with the pseudosymmetric substrate so that each sp lice site is cleaved by a symmetrically A O- OH 0 0-0 0-0 G-C R-Y A/ R C-GA U-R C- G v,P-0 /~R-Y XR- \ XX XX Eucaryal substrate consensus Eucaryal endonuclease Archaeal substrate consensus Archaeal endonuclease Fig. 1. (A) Consensus sequence and secondary structure of precursor tRNA substrate for yeast and archaeal endonucleases. Splice sites are indicated by short arrows. 0 and X = nonconserved bases in regions of conserved and variable secondary structure, respectively; Y = pyrimidines; R = purines. Yeast endonuclease is proposed to interact with the mature domain of the tRNA and measure the distance to the splice sites. In contrast, the archaeal enzyme recognizes two 3-nucleotide loops that are separated by a helix of 4 bp (shaded area). (B) Comparison of endonuclease models in Eucarya (yeast) (7) and Archaea (H. volcanii) (8) (see text). www.sciencemag.org • SCIENCE • VOL. 280 • LO APRIL 1998 279 disposed acti ve site monomer (Fig. 1B). Thus it is likely that the spatial disposition of the ac ti ve subunits has been conserved throughout evolution. We believed that a high-resolution structure of the arch aea l tRNA endonuclease wou ld shed light on the mechanism of the m ore complicated but related eucarya l endonuclease . Therefore, we determined the cryst al structure of the endonuclease from the archaeon Methanococcus jannaschii , which is a h omotet ra mer of 21 kD (179 amino acids) ( IO). The M. jannaschii endonuclease gen e with His8 and FLAG epitope tag seq uences (7) at its 5' end was cloned by polymerase ch ain reaction from genomic DNA into the pETl la vector. Soluble protein was obtained by overexpression in Escherichia coli ( 11). The structure was determined by the multi wavelength an omalous diffraction (MAD) method on crysta ls derivatized by Au(CN) 2. Th e final m ode l, which includes the tetrameric endonuclease of residues 9 to 179, 53 water oxygen atom s, 2 partially occupied SO 4 - ion s, and 4 go ld atoms, was refined to 2.3 A with good stereoch emistry to an R factor of 0 .204 and a free R value of 0.268 (Table 1). The endonuclease m on omer fo lds into 0 Table 1. Statistics for MAD data collection and phase determination from a Au(CN) 2 -derivatized endonuclease crystal flash cooled to 100 K (11). Crystals belong to the space group P2 1 2 1 2 1 , with unit cell dimensions a = 61.8 A, b = 79.8 A, and c = 192.8 A. Three wavelengths near the L 111 absorption edge of Au were collected from a single crystal at National Synchrotron Light Source beamline X4A using Fuji image plates: 1\ 1 , absorption edge corresponding to a minimum of the real part (f') of the anomalous scattering factor for Au; 1\ 2 , absorption peak corresponding to a maximum of the imaginary part (f") of the scattering factor for Au; 1\3 , a high-energy remote to maximize f'. The crystal was oriented with its c* axis inclined slightly from the spindle axis (- 16°) to avoid a blind region. The a* axis initially was placed within the plane of the spindle and the beam direction using the STRATEGY option in the program MOSFLM (version 5.4) (21). A total of 180 consecutive 2° images were collected with every 8° region repeated for all three wavelengths. All data were processed with the DENZO and SCALEPACK programs (22) . A previously collected data set at 1.5418 Aon a crystal soaked with the same Au(C N)2 solutions was included as the fourth wavelength. This data set was measured with CuKa x-rays generated by a Rigaku RU200 rotating anode using a SIEMENS multiwire detector and was processed with the program XDS (23) followed by scaling and reducing with ROTAVATA and AGROVATA in CCP4 (24). All reflection data were then brought to the same scale as those of wavelength 1 with the scaling program SCALEIT followed by the Kraut scaling routine FHSCAL (25) as implemented in CCP4. We treated MAD data as a special case of the multiple isomorphous replacement with the inclusion of anomalous signals (26). The wavelength 1 (edge) was taken as the "native" data with intrinsic anomalous signals. The data col lected at wavelengths 2 (peak), 3 (remote 1), and 4 (remote 2) were treated as individual "derivative" data sets. Anomalous signals are included in the data two distinct domains, the NHz-termina l domain ( residues 9 to 84) and the COOHterminal domain (residues 85 to 179) (Fig. 2A). The N H z-dom ain co n sists of a mixed antiparallel/parallel [3 sheet and three o. helices. The first three [3 strands ([31 to [33) are antiparallel and, together with [34 , are packed against two n early perpendicular o. helices, o.l and o.2. The third o. helix, o.3, is associated with o.2 via an antiparallel co iled -co il. The connec tion between [34 and o.3 (residues 64 to 67) is disordered, which im plies that this loop is flex ible. This feature is con sist ent with the proteolytic sens itiv ity of this region ( JO). A sh ort loop set of wavelength 2 where the x-ray absorption of Au is expected to be maximum. Because of the absence of a sharp white line at the Au L 11 1 edge, the observed diffraction ratio at wavelength 2 is small, 0.061, compared with the error signal ratio of 0.036 computed from centric reflections. Two gold atoms in each asymmetric unit could be clearly located by both dispersive difference (/\ 3 or 1\4 with respect to 1\ 1) and anomalous (/\ 2 ) Patterson functions with the program HASSP (27). Two partially occupied sites were later located by difference Fourier. Initial phases were computed at 3.0 Awith the program MLPHARE as implemented in the CCP4 program suite, which uses a maximum-likelihood algorithm to refine heavy atom parameters (28). There are four noncrystallographically related monomers per asymmetric unit. We performed fourfold averaging in parallel with solvent flattening and histogram mapping with the program OM (29). The subsequent electron density map was improved markedly and the polypeptide chain could be traced unambiguously. A model representing the monomeric endonuclease containing all but the first 25 amino acids (the 8 His, the 9 FLAG epitope tag amino acids, and the first 8 amino acids in the endonuclease) was built into the electron density map with the interactive graphics program O (30). The assignment of side chains for the M. jannaschii endonuclease was confirmed by a difference Fourier map between the Au(CNkderivatized protein and a selenomethionine-substituted protein. The final model was refined to 2.3 A against the observed anomalous data of wavelength 3 with X-PLOR version 3.8 (31). Noncrystallographic restraints were applied to atoms between dimers A 1-A2 and B1 -B2 excluding those from loops L3 and L7. The structure was assessed with PROCHECK (32) and VERIFY3D (33) and was found to be consistent with a correct structure. More than 90% of the residues are in the most-favored regions , and no residues are in disallowed regions of a Ramachandran plot (32) . Data collection statistics* 1\ 1 (edge) 1\ 2 (peak) 1\3 (remote 1) 1\4 (remote 2) Wave!ength (A) Resolution (A) Unique Measured reflections reflections 1.03997 1.03795 1.02088 1.54180 2.3 (2.4-2.3) 2.3 (2.4-2.3) 2.3 (2.4-2.3) 3.0 (3.1-3.0) 165,549 161,691 164,948 50,259 75,778 75,3 10 76,385 16,679 Rmerge'f (/}/(a(/)} Completeness (%) 0.052 (0. 153) 11.3 (2.6) 0.051 (0.151) 10.6 (2.6) 0.054 (0. 167) 10.3 (2 .4) 0.085 (0.359) 8.7 (2.2) 89.2 (62.2) 89.0 (62.0) 90.4 (65.4) 83.3 (42.6) Phasing and symmetry averaging statistics Phasing power Figure of merit f' f" (e·) (e·) -20.21 10.16 -15 .21 10.20 - 10.76 9.87 -5.10 7.30 Dispersive ratio (30 to3A)t 0.059 0.060 0.1 65 Anomalous ratio (30 to3A}t 0.061 [0.041 l 0.061 [0.036] 0.062 [0.036] Refinement statistics Corr. coef. of NCS§ operations 1'2 A3 A4 Initial After OM Initial After OM 0.11 0.24 1.02 0.22 0.78 0.37 0.86 Resolution (A) 50.0-23 R Rt,ee value (all data) value (10% data) 0.204 0.268 No. of atoms in refined model rms deviationsll Bond (A) Angle (0) Protein H2 0 so 4 - Au 0.011 1.586 5584 53 9 4 ·Numbers in parentheses correspond to those in the last resolution shell. tRmerge indicates agreement of individual reflection measurements over the set of unique averaged reflections. Rmerge = :i.h., Ii(h), - (/(h)) I / :i.h)(h), where h is the Miller index, i indicates individually observed reflections, and (/(h)) is the mean of all reflections of the Miller index h. Bijvoet mates are treated as independent reflections when computing Rmerge· +Diffraction ratios are defined as (c, IFl 2 ) 112/( IFl 2 ) 112 where t. IFl 2 are taken between the current and the reference wavelength (1\ 1) for dispersive ratios and between matched Bijvoet pairs for anomalous ratios. Ratios for centric reflections at each wavelength are shown in brackets. §Noncrystallographic symmetry. IIGeometry values of the final protein model were compared with ideal values from Engh and Huber (34). 280 SCIENCE • VOL. 280 • 10 APRIL 1998 • www. sciencemag.org consisting of three charged residues (Glu 81, Glu 82 , Arg 83 ) links the NHz- to the COOH-terminal domain. The COOH-terminal domain is of the a/[3 type and consists of a centra l four-stranded mixed 13 sheet 135-136-13 7-138 ( overbars denote parallel strands) cradled between helices a4 and a5. Two basic folding motifs lead to this fo ld: the rare al3l3(a4-135-l36) and the more common l3al3(13 7-a5-138). The last 13 strand, 139, is partially hydrogen-bonded with 138 and has little involvement in the COOH-terminal domain folding; rather, it participates in dimerization (see below). The quaternary structure of the endonuclease is unusual in that it does not assume the 0 2 point group symmetry observed in most homotetrameric proteins. When viewed along a noncrystallographic twofold axis, the homotetrameric M. jannaschii endonuclease resembles a parallelogram with the four subunits occupying the four corners (Fig. 2B). Subunits Al and Bl, related by a true twofold ax is, are closer in space than subunits A2 and B2, which are related by the same twofold axis. Subunits Al and A2 or Bl and B2 are related by two nonorthogona l pseudo-twofold axes plus a 1.6-A. translation along the rotation axes. The resulting dimers, Al-A2 and Bl-B2, lie on two opposite sides of the parallelogram. Each of the pseudo-twofold axes relating monomers within the Al-A2 and Bl-B2 dimers is at an angle of 32° with respect to the twofold axis that relates the two dimers. In addition, the three twofold axes do not intersect. Each pseudo-twofold axis is shifted about 10 A in the opposite direction relative to the twofold ax is that relates the two dimers. It is therefore more precise to describe the endonuclease as a dimer of dimers. The a carbons of the A l-A2 dimer can be superimposed with those of the BlB2 dimer with a root-mean-square difference of 0.4 7 A. The a carbons of the subunits within each dimer have a root mean square of 1.2 A. The largest deviation is observed on loop L 7, which harbors the conserved His 125 residue. The asymmetrical organization of the four identical subunits suggests that the two subunits in the dimers have nonequivalent roles. The endonuclease dimers Al-A2 and Bl-B2 associate isologously in a tail-to-tai l fashion (Fig. 2B). The dimerization interface buries extensive surface area (2422 A. 2) ( 12) and is formed by the interaction of the central COOH-terminal 13 sheets from each monomer by main chain hydrogen bonding and hydrophobic interactions. The pseudotwofold ax is that relates the two monomers lies between the two 139 strands and between the two LS loops in each subunit. The NH 2 portion of 139 (residues l69 to 172) is hydrogen-bonded to 138 by the main chain atoms. The COOH-terminal half of 139 (residues 172 to 177) bends away from the plane of the pleated central 13 sheet to form main chain hydrogen bonds with the symmetry-related residues of 139 in the isologous subunit (139' ); as a result, a twostranded 13 sheet is produced that spans the subunit boundary. The symmetrically related LS loops form another layer on top of the two-stranded 139 sheet and, together with the 139 sheet, enclose a hydrophobic core at the intersubunit surface. Prealbumin and enterotoxin have similar modes of subun it association (13). The hydrophobic core includes Phe 139 , Leu 141 , Leu 14 4, Gly 146, Va!148, Leu1 ss, Ile1 60, Metl74, and Tyrl76 (Fig. 2B, blue ball-and-stick model). These hydrophobic residues are important for stabilizing the dimer interface. The interface between the two dimers is between subunits A l and B2 and subunits Bl and A2, with each pair contributing 40% of 5420 A2 of total buried surface area ( 12). The remaining 20% of buried surface area can be accounted for by the interface between subunits Al and Bl. Subunits A2 and B2 are not in contact. The principal interaction between dimers is mediated by close contacts between polar residues of opposite charges in the two heterologously associated subunits. A positively charged cleft (9 A wide, 12 A long, 14 A deep) is formed between the two domains of subunit Al (or Bl). It accommodates a protruding negative surface on the subunit B2 (or A2) formed primarily by LIO and partially by LS (Fig. 2B, dotted surfaces). Residues from LIO and LS that are deeply buried by this interdomain cleft are mostly acidic (Asp 164 A lat65 -Aspl 66_Gly1 67 -Asp168 and G[ul35 Asp 136 ). Several positively charged residues (Lys 88 , Lys 103 , Lys 107 , Argll 3) line the cleft and interact electrostatically with the acidic loop residues. The interactions between LIO and LS and the interdomain cleft contribute nearly all the buried surface area at the tetramerization interface. In addition, three completely buried water molecules at A COOH Fig. 2. (A) Ribbon representation of one subunit of M. jannaschii endonuclease obtained with the RIBBONS program (20). The proposed catalytic triad residues are within 7 A of each other and are shown in red ball-and-stick models (see text). The electron density in the averaged F0 map is contoured at 5u and is drawn only near the putative catalytic triad. (B) Subunit arrangement and interactions in the M. jannaschii endonuclease. Each subunit is represented by a distinct color and a label. The tetramer is viewed along the true twofold axis relating the A1-A2 and B1-B2 dimers. The main chain hydrogen bonds formed between (39 and (39' and between loops L8 and L8 ' for isologous dimers are drawn as thin lines. Side chains of the hydrophobic residues enclosed at the dimer interface are shown as blue ball-and-stick models. The heterologous interaction between subunits A1 and B2 (or B1 and A2) through the acidic loops L1O and L8 are highlighted by dotted surfaces. www.sciencemag.org • SCIENCE • VOL. 280 • 10 APRIL 1998 281 the base of the cleft form hydrogen bonds with polar residues from both subunits, indicating that this cleft is so lvent accessib le before tetramerization. The isologous dimer interaction fo rmed by the COOH-terminal interface is expected to be more stable than the heterologous polar interaction med iated by LIO and L8, which suggests that the tetramer asse mbles through dimer intermediates. This interpretation is in agreement with protein crosslinking data (JO). It is likely that the dimers observed in solution are the isologous dimers Al-A2 or Bl-B2 in the crystal structure. The tRNA splicing endonucleases generate 5 '-hydroxyl and 2' ,3' cyclic phosphate termini, as do other ribonucleases (RNases) such as RNase A and Tl. C leavage by RNase A is a two-step reaction that is acid-base catalyzed (/4). Three residues in RNase A have been identified as the cata lytic triad for this reaction: His 12 is the general base in the first step, His 119 is the general acid , and Lys 41 stabilizes the pentacovalent transition state. Spli cing endonucleases contain a conserved histidine residue that has been suggested to be part of the active site on the basis of mutational studies (7 , 10, 15) . In the M. jannaschii endonuclease structure, His 125 is at the boundary between L7 and 137 (Fig. 2A). A lthough His 125 is well ordered in the structure in all four subuni ts, the rest of L7 (residues 119 to 124) has A B G G C C weak electron density, which is suggestive of flexibility. The average B factor of L7 is 66 A2 compared with the average value of 44 A2 for the protein as a whole. Two striking structural features are observed in this region. First, there are primarily positively charged or polar residues in the vicinity of His 125 , including two other strictly conserved residues, T yr 1I5 and Lys 156 . The side chains ofHis 125 , T yr 1I5 , and Lys 156 are within 4 to 7 A of each other and form a His-T yr-Lys triad (5.5 A between His 125 and Lys 156 , 6.7 A between Lys 156 and T yr115 , 6.1 A between T yr 156 and His I25 ) (Fig. 2A). Second, a distinct electron density feature ( > Ser in th e averaged F map ) is found in the middle of the triad . This density is not part of the polypeptide chain and is compatible with the size of a sulfate ion that is present in crystallization and stab ilization so lutions. W e propose that the side chains of the His-Tyr-Lys triad in M. jannaschii endonuclease form the cata lytic triad for the cleavage reaction. This His-T yr-Lys triad is spatially superimposable with the catalytic triad in RNase A (16) and is consistent with a model in which the endonuclease His 125 functions as the general base, T yr 115 functions as the general acid, and Lys 156 functions as a stab ilizer of the transition state. A lthough mutagenesis studies do not provide conclusive support for the essentiality of T yr 11 5 , they do indicate an essential ro le for both Hi s I25 and Lys 156 (4). More studies O C C C G G A U G C A U I cu UG C A U G C C G C A UGGG HIV TAR RNA Modeled archaeal substrate Fig. 3. (A) Sequence and secondary structure of HIV TAR RNA (PDB code 1arj) used to model M. jannaschii pre-tRNA minimum substrate. (B) A symmetry operation around the twofold axis as indicated is applied to coordinates of the TAR RNA nuclear magnetic resonance structure to obtain a structural model for the second threebase bulge and the stem extension (boxed region). The corresponding splice sites on the final model are indicated by arrows . (C) Modeled complex of M. jannaschii endonuclease with RNA substrate obtained by manually aligning the phosphate backbone of the 4-bp helix with the positively charged surface between subunits A1 and B1. This view is perpendicular to that of Fig . 2 B. Only subunits A1 and B1, which are proposed to participate in the cleavage reaction, are shown. The proposed catalytic triad residues are shown in magenta. The phosphodiester bonds at the splice sites on RNA are depicted in blue and fit nearly perfectly with the putative sulfate density shown by blue dots. The distance from the center of the catalytic triad to the surface where RNA binds is about 7 A. 282 are necessary to establish the exact catalytic mechanism and the function of these residues in catalysis. Interestingly, the inactivating G ly292 to Glu mutation in the yeast enzyme occurs within the aligned sequences of the flexible loop L7, which contains both His 125 and T yr 115 (7). Although the M. jannaschii endonuclease tetramer conta ins fou r active sites, it is likely that only two of these participate in cleavage. The yeast endonuclease tetramer contains two functionally independent active sites for cleavage . The H. volcanii endonuclease is a homodimer and thus also contains two functional acti ve sites. The two intron-exon junctions of the archaeal tRNA substrate are related by a pseudo-twofold symmetry ax is in the consensus BHB substrate. The M. jannaschii endonuclease is therefore predicted to contain o nly two fun ction al active sites related by twofold symmetry. Consistent with this notion, upon substrate binding the endonuclease is protected from proteolytic cleavage to a max imum of 50% (J O). Four pairs of enzyme ac tive sites are related by twofold symmetry; these occur in subunits Al and Bl, A2 and B2, Al and A2, and Bl and B2 (Fig. 2B). We postulate that the ac ti ve sites reside in Al and Bl; their His-T yr-Lys triads are separated by abou t 27 A. Th e res idues protected upon RNA binding cluster on Al and Bl around the two proposed ca talytic sites, falling on a path between the two sites. This region of the enzyme, fo rmed by L9 and the carboxyl portions of aS and 139 in Al and Bl, has a positive electrostatic potential suitable for binding the phosphodiester backbone of the pre-tRNA substrate. By contrast, th e distance between the His-Tyr-Lys triads of A2 and B2 is 75 A and there is a more scattered distribution of substrate protected residues between this pair of active sites (17). We atte mpted to dock a model of the substrate with the tetramer. We created the mode l from a related structure solved by the nuclear magnetic resonance method , that of the human immunodeficiency virus t rans-act ivation response element (TAR) RNA complexed with arginamide (J 8). TAR RNA is a stem-loop structure with a bulge of three bases in the stem (Fig. 3A). W e obtained the substrate model by rotating a section of TAR RNA by 180° to create the pseudosymmetric BHB structure conforming to the archaeal endonuclease consensus substrate (Fig. 3B). This substrate model derived fro m the arginam id e complex of T AR RNA docks well with subunits Al and Bl in the M. jannaschii endonuclease tetramer (Fig. 3C). The two phosphodiester bonds that are cleaved fit precisely into the active SCIENCE • VOL. 280 • 10 APRIL 1998 • www.sciencemag.org sites of A l and Bl and superim pose with th e putati ve sulfa te ion . The 4-bp h elix separating the two bulges interacts with t he basic face of the enzyme and could contac t a number of res idues that are protected fro m protease cleavage upon RNA binding ( JO ). As the H . volcanii endonuclease dimer and the M. jannaschii tetramer recognize the same consensus substrate, their active sites must be arrayed simi larly in space. This proposal is stron gly supported by the fact that the H. volcanii endonuclease monomer is in fac t a tandem repeat of the consensus sequence of the endonuclease gene family ( 10 ) (Fig. 4B) . W e propose that the H . volcanii enzyme fo rms a pseudotetramer of two pseudodimers (Fig. 4B). The structure of the pseudodimer is predicted to contain a twostranded 139-139 ' pleated sheet, an important structu ra l feature of the M. jannaschii endonuclease dimer (Fig. 4A ). The tandem repeat must be connected by a stretch of polypeptide (Fig. 4B, dotted line ). Within each dimer of the M. jannaschii endonuclease structure, the N H z- terminus of subunit AZ (or B2) and the COOH-terminus of subunit A l (or Bl) is separa ted by 28 A. This requires a span of at least 10 amino acids of an extended polypeptide chain to connect the region between the NH z- terminal and the COOH-terminal tandem repeats in the H . volcanii endonuclease. The H . volcanii enzyme model con ta ins only two of the proposed acti ve site triads in the COOH-termin al repeat of each pseudod imer, and these triads are proposed to occupy a spatial configuration identical to those in the A l and Bl subunits of the M. jannaschii endonuclease. The pseudod imers are proposed to interact via the conserved Ll 0 sequences. This H. volcanii endonuclease structu ra l model reveals chat only two of the acti ve sites are necessary, but to array these in space correctly one must retain features of both the isologous dimer interactions (139-139' ) and the dimer-dimer interactions mediated by Ll 0 in the M. jannaschii enzyme. The heterotetrameric yeast endonuclease is proposed to contain two active subunits, Sen 2 and Sen34. The other two subunits, Sen 15 and Sen54, have a sequence to the M. jannaschii endonuclease at the COOH-terminus ( JO ). This conserved region conta ins the Ll 0 residues and the h ydrophobic res idues at the COOH-terminus as observed in 138 and 139 in the crystal structure of the M. jannaschii endonuclease. W e propose that the stron g Sen 2-Sen54 and Sen34-Senl 5 interactions observed in the two-hybrid ex periment (7) are mediated by COOH-terminal 139- 139' - like interactions (Fig. 4C ) and that these dimers are associated to fo rm the tet ramer via the Ll 0 sequences of Senl 5 and Sen54 (Fig. 4C). We conclude that the A rchaea and the Eucarya h ave inherited from a common ancestor an endonuclease active site and the means to array two sites in a precise and conserved spatial orientation . This conclusion is supported by the res ults of Fabbri et al. ( 19), which de monstrate that both the eucaryal and the archaeal endonucleases can accura tely cleave a uni ve rsal substrate conta ining the BHB mo tif. The eucaryal en zyme appears to dispense with the more complex ruler mechanism fo r tRNA substrate recogni t ion when it cleaves the uni- versa! su bstrate. Thus the precise positioning of two active sites in endonuclease appears to have been conserved in evolu tion . S ubuni ts A l and Bl comprise the active site core of all tRNA splicing endonucleases, and subuni ts AZ and B2 pos ition the two active sites precisely in space. The eucaryal enzyme has evo lved a distinct measuring mechanism fo r splice site recogni t ion by specialization of the AZ and B2 subunits, and it has reta ined the ability to recogniz and cleave the primitive consensus substrate. Fig. 4. (A) Sequence fea- A Y115 H125 K156 L10 tures and subunit arrangement of the M. jannaschii M. jannaschii ■■ II homotetramer. Several important structural features discussed in the text are indicated: loop L10, the COOH-terminal [39 strands (arrows), and the conserved catalytic residue His 125 (pentagon). (B) Sequence features conserved between H. volcanii and M. Jannaschii endonucleases and the proposed su bunit arrangement of the former. The two tan- B NH 2 Y115 L10 Y115 H125 K156 L10 dem repeats are more similar to the M. jannaschii en- H.volcanii - ■■ I donuclease sequence than to each other. The NH 2 -terminal repeat lacks two of the three putative active site residues (white bars). It does, however, contain many of the features of the COOHterminal domain that are important for structural arrangement of the enzyme, in particular the L1O sequence (yellow bars). The COOHterminal repeat contains all the sequence features of C the M. jannaschii enzyme. ps Dashed lines represent the polypeptide chain connectM. jann . ing the COOH-and NH 2 -terminal repeats. (C) Proposed 467 structural model for the yeast endonuclease. Conserved amino acid sequences near the COOH-termini of archaeal enzymes M. Jannaschii (M . jann.), H. volcanii NH 2 -terminal repeat (H. vol. Nt), and yeast Sen54 (Sc. Sen 54p) and Sen15 (Sc. Sen15p) subunits are aligned . Important hydrophobic residues that stabilize the isologous COOHterminus interaction between M. Jannaschii subunits A1 and A2 (or B1 and B2) are highlighted in green and circled on the structural models of heterodimers derived from the two-hybrid matrix analysis (7). The sequences of Sen54 and Sen 15 aligned with L1O sequences in M. Jannaschii and H. volcanii are highlighted in red. Loops L1O on both Sen54 and Sen1 5 are labeled on the proposed heterotetramer model of the yeast endonuclease. Abbreviation for amino acids are as fo llows: L, Leu; I, lie; A, Ala; V, Val; D, Asp; G, Gly; Y, Tyr; N, Asn; M, Met; E, Glu; T, Thr; F, Phe; W, Trp; S, Ser; K, Lys. www.sciencemag.org • SCIENCE • VOL. 280 • IO APRIL 1998 I 283 REFERENCES AND NOTES 1. D. Biniszkiewicz, E. Cesnaviciene, D. A. Shub, EMBO J. 13, 4629 (1994); B. Reinholdhurek and D. A. Shub, Nature 357 , 173 (1992). 2. S. K. Westway and J. Abelson, in tRNA: StrJcture, Biosynthesis, and Function, D. Soll and U. Raj Bhandary, Eds. (American Society for Microbiology, Washington, DC, 1995), pp. 79-92; M. R. Culbertson and M. Winey, Yeast 5,405 (1989); E. M. Phizickyand C. L. Greer, Trends Biochem. Sci. 18, 31 (1993). 3. J. R. Palmer, T. Baltrus, J. N. Reeve, C. J. Daniels, B1ochim. Biophys. Acta 1132, 315 (1992); J. Kjems, J. Jensen, T. Olesen, R. A. Garrett, Can. J. Microbial. 35 , 210 (1989). 4. C. R. Trotta, unpublished results. 5. V. M. Reyes and J. N. Abelson, Cell 55 , 719 (1988). 6. L. D. Thompson and C. J. Daniels, J. Biol. Chem. 265, 18104 (1990). 7. C.R. Trotta et al., Cell 89 ,849 (1997). 8. K. Kleman-Leyer, D. A. Armbruster, C. J. Daniels, ibid., p. 839. 9. C. K. Ho, R. Rauhut, U. Vijayraghavan, J. N. Abelson, EMBO J. 9, 1245 (1990). 10. J. Lykke-Andersen and R. A. Garrett, ibid. 16, 6290 (1997). 11. Recombinant endonuclease was purified from E. coli by a Ni-NTA (Qiagen) affinity chromatography step after heat denaturation. Crystals were grown at 22'C by vapor diffusion using KCI as a salting-in precipitant with a starting concentration of 1 M. The well solutions were at pH 5.5 to 6.0 and contained 20 to 80 mM (NH 4 ),SO 4 . Crystals were soaked for 3 days in 20 mM sodium cacodylate (pH 5.5 to 6.0), 20 mM (NH 4 ),SO4 , 25% glucose, 2 mM Au(CN), before data collection. 12. B. Lee and F. M. Richards, J. Mot. Biol. 55 , 379 (1971). 13. C. C. F. Blake et al., ibid. 88, 1 (197 4); T. K. Sixma et al., Nature 351 ,371 (1991). 14. C. Walsh, Enzymatic Reaction Mechanisms (Freeman, San Francisco, 1979). 15. C. J. Daniels, personal communication. 16. The catalytic triad (residues His 12 , His 1 19 , and Lys 41 ) of RNase A was brought manually to superimpose with Tyr 11 5 , His 1 25 , Lys 15 6 of M. jannaschii endonuclease. The best superimposition was obtained when Tyr 11 5 aligns with His 11 9 , His 1 25 with His1 2 , and Lys 156 with Lys41 . 17. H. Li, C.R. Trotta, J. N. Abelson, data not shown. 18. J. D. Puglisi, L. Chen, A. D. Frankel, J. R. Williamson , Proc. Nat. Acad. Sci. U.S.A. 90 , 3680 (1993); F. Aboul -ela, J. Karn, G. Varani, Nucleic Acids Res. 24 , 3974 (1996). 19. S. Fabbri et al. , (1998) Science, in press. 20. M. Carson, J. Mot. Graphics 5, 103 (1987). 21. A. G. W. Leslie, CCP4 AND ESF-EACMB News/. Protein Crystallogr. 32, 2 (1996). 22 . Z. Otwinowski , in Proceedings of the CCP4 Study Weekend: Data Collection and Processing, 29-30 January 1993, L. I. Sawyer, N. Isaac, S. Bailey, Eds. (Science and Engineering Research Council, Daresbury Laboratory, Warrington , UK, 1993), pp. 56-62; W. Minor, XDISPLAYF Program (Purdue University, West Lafayette, IN, 1993). 23. W. Kabseh, J. Appl. Crystallogr. 21 , 916 (1988). 24. Collaborative Computational Project, Number 4, "The CCP4 Suite: Programs for Protein Crystallography," Acta Crystallogr. D 50 , 760 (1994). 25. J. Kraut et al., Proc. Natl. Acad. Sci. U.S.A. 48, 1417 (1962). 26. V. Ramakrishnan and V. Biou, Methods Enzymof. 276, 538 (1997). 27. T. C. Terwilliger, S. H. Kim, D. E. Eisenberg, Acta Crystallogr. A 43 , 34 (1987). 28. Z. Otwinowski, in fsomorphous Replacement and Anomalous Scattering, W. E. Wolf, P. R. Leslie, A. G. W. Leslie, Eds. (Science and Engineering Research Council, Daresbury Laboratory, Warrington, UK, 1991), p. 80. 29. K. Cowtan, Joint CCP4 ESF-EACBM News/. Protein Crystallogr. 31 , 34 (1994). 30. T. A. Jones, J. Y Zou, S. W. Cowan, M. Kjeldgaard, Acta Crystallogr. A 47 , 110 (1991). 31. A. T. Brunger, X-PLOR Version 3.1, A System for 284 Crystallography and NMR (Yale Univ. Press, Department of Molecular Biophysics and Biochemistry, New Haven, CT, 1996). 32. R. A. Laskowski, M. W. MacArthur, D.S. Moss, J.M. Thornton , J. Appl. Crystallogr. 26 , 283 (1993). 33. R. Luthy, J. U. Bowie, 0. Eisenberg, Nature 356 , 83 (1992). 34. R. Engh and R. Huber, Acta Crystallogr. A 47 , 392 (1991) 35. We thank D. Graham and C. Woese for M. jannaschii genomic DNA, P. Bjorkman and D. Rees for their helpful discussions and generosity in sharing their equipment, C. M. Ogata for x-ray beam time allocation, F. T. Burling and S. Diana for assistance in data collection, R. Story and M. Saks for critical review of the manuscript, and other members of the Abelson laboratory for their advice and support. This work was supported by American Cancer Society grant NP802 and NIH Individual National Research Service Award F32 GM 188930-01 . PDB code for the coordinates is 1a79. 10 December 1997; accepted 13 February 1998 SCIENCE • VOL. 280 • l 0 APRIL 1998 • www.sciencemag.o rg