Advertisement
JBC

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Originally published In Press as doi:10.1074/jbc.M412594200 on December 7, 2004

J. Biol. Chem., Vol. 280, Issue 10, 9555-9566, March 11, 2005
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplemental Data
Right arrow All Versions of this Article:
280/10/9555    most recent
M412594200v1
Right arrow Submit a Letter to Editor
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowRequest Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Cho, H. D.
Right arrow Articles by Weiner, A. M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Cho, H. D.
Right arrow Articles by Weiner, A. M.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?

Archaeal CCA-adding Enzymes

CENTRAL ROLE OF A HIGHLY CONSERVED {beta}-TURN MOTIF IN RNA POLYMERIZATION WITHOUT TRANSLOCATION*{boxs}

HyunDae D. Cho, Christophe L. Verlinde, and Alan M. Weiner{ddagger}

From the Department of Biochemistry, School of Medicine, University of Washington, Seattle, Washington 98195-7350

Received for publication, November 8, 2004


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS AND DISCUSSION
 REFERENCES
 
The CCA-adding enzyme (tRNA nucleotidyltransferase) builds and repairs the 3' end of tRNA. A single active site adds both CTP and ATP, but the enzyme has no nucleic acid template, and tRNA does not translocate or rotate during C75 and A76 addition. We modeled the structure of the class I archaeal Sulfolobus shibatae CCA-adding enzyme on eukaryotic poly(A) polymerase and mutated residues in the vicinity of the active site. We found mutations that specifically affected C74, C75, or A76 addition, as well as mutations that progressively impaired addition of CCA. Many of these mutations clustered in an evolutionarily versatile {beta}-turn located between strands 3 and 4 of the nucleotidyltransferase domain. Our mutational analysis confirms and extends recent crystallographic studies of the highly homologous Archaeoglobus fulgidus enzyme. We suggest that the unusual phenotypes of the {beta}-turn mutants reflect the consecutive conformations assumed by the {beta}-turn as it presents the discriminator base N73, then C74, and finally C75 to the active site without translocation or rotation of the tRNA acceptor stem. We also suggest that {beta}-turn mutants can affect nucleotide selection because the growing 3' end of tRNA must be properly positioned to serve as part of the ribonucleoprotein template that selects the incoming nucleotide.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
EXPERIMENTAL PROCEDURES
 RESULTS AND DISCUSSION
 REFERENCES
 
The CCA-adding enzyme (ATP(CTP):tRNA nucleotidyltransferase) is the only RNA polymerase that can build or rebuild a specific nucleic acid sequence without using a nucleic acid template (1, 2). CCA-adding enzymes belong to the nucleotidyltransferase superfamily (3, 4), which can be subdivided into class I and class II according to highly conserved features of the nucleotidyltransferase motif (5) that positions two metal ions for the nearly universal phosphoester bond transfer mechanism (6, 7). Archaeal CCA-adding enzymes belong to class I, and eukaryotic and eubacterial CCA-adding enzymes belong to class II (5). Most surprisingly, the class I and class II CCA-adding enzymes exhibit strong core homology within each class (about 40 kDa in class I and 25 kDa in class II) but little obvious homology between the two classes outside the immediate vicinity of the nucleotidyltransferase motif. A further complication is that CCA-adding activity is the joint responsibility of related class II CC- and A-adding enzymes in certain deeply rooted eubacteria (8, 9). The CCA-adding enzyme is a nonessential repair enzyme in those eubacteria whose tRNA genes all encode CCA (10, 11), but is an essential enzyme in eukaryotes (12) and presumably also in those Archaea and eubacteria whose tRNA genes do not all encode CCA.

How does the CCA-adding enzyme build CCA without using a nucleic acid template? We found previously that tRNA does not translocate along the CCA-adding enzyme during addition of C75 and A76 (addition of C74 was not examined) (13) and that a single nucleotidyltransferase motif is responsible for adding all three nucleotides (14). (The 3'-terminal tRNA sequence NCCA is designated positions 73, 74, 75, and 76 in the universal tRNA numbering convention, where N is the discriminator base.) How then is the 3'-terminal hydroxyl group of the three different tRNA substrates (tRNA-N, tRNA-NC, and tRNA-NCC) identically positioned with respect to the single catalytic nucleotidyltransferase motif? In the simplest scenario, the catalytic motif would move dramatically with respect to the tRNA-binding site after each nucleotide addition; however, such large molecular motions seemed unlikely in enzymes of <50 kDa. In the "collaborative templating" model (13), the growing 3' end of the tRNA would be sequestered within a pocket near the active site; progressive packing of this pocket would create a binding site for the incoming nucleotide and correctly position the attacking 3'-hydroxyl of the tRNA substrate; CCA synthesis would then cease when the pocket was full. In the "scrunching-shuttling" model (15, 16), two identical subunits of the tRNA-induced tetrameric enzyme would function nonequivalently. One subunit would add C74 and then "scrunch" or bulge C74 to allow addition of C75. After addition of C75, the CC terminus would be long enough to "shuttle" over to a nonequivalent subunit, which would add A76. The scrunching-shuttling model, however, turned out to be inconsistent with the ability of a single active subunit in the multimeric enzyme to catalyze all three steps of CCA addition (17). In a fourth model (18) that did not address the absence of tRNA translocation, the CCA-adding enzyme would function as a poly(C) polymerase that undergoes a significant conformational change after two rounds of CTP addition, thus enabling addition of ATP (perhaps prebound at another site (19)) to terminate poly(C) synthesis. As discussed below and recently reviewed (20), new cocrystal structures (21, 22) suggest that the actual mechanism of CCA addition is a hybrid between collaborative templating (13) and "double scrunching" (16), in which both C74 and C75, rather than C74 alone, are bulged within a single subunit.

We began this mutational analysis of the class I Sulfolobus shibatae enzyme before the structures of any CCA-adding enzymes had been solved. We modeled the S. shibatae enzyme on the crystal structures of two other class I nucleotidyltransferases, yeast (23) and bovine poly(A) polymerase (24), using the 3D-PSSM threader (25). We then generated an array of mutations in residues surrounding the active site of the archaeal enzyme, with an emphasis on an evolutionarily versatile {beta}-turn, and we tested the mutant enzymes for the ability to add CTP and ATP to the three possible substrates tRNA-N, tRNA-NC, and tRNA-NCC. We found two striking phenotypes: mutations that specifically blocked addition of C74, C75, or A76; and mutations that progressively inhibited addition of C74 > C75 > A76 or A76 > C75 > C74. Most surprisingly, only one of the mutations we tested caused nucleotide misincorporation, although all were located close to the active site where nucleotide selection must take place.

Crystal structures have now been determined for the class I Archaeoglobus fulgidus apoenzyme (26, 27) and, more recently, for the Archaeoglobus enzyme in complex with mature tRNA, as well as with oligonucleotide mimics of two different tRNA substrates (tRNA-NC and tRNA-NCC) and incoming nucleotide (21). The Archaeoglobus structures validate our predicted structure for the highly homologous Sulfolobus enzyme; conversely, our mutational analysis of the Sulfolobus enzyme provides an independent test of conclusions based on the Archaeoglobus crystal structures. Our data also focus attention on an evolutionarily versatile {beta}-turn, located between strands 3 and 4 of the nucleotidyltransferase domain, which serves fundamentally different roles in different polymerases belonging to the nucleotidyltransferase superfamily. Our data highlight how consecutive conformations of the {beta}-turn present the growing 3' end of the tRNA substrate to the catalytic site, circumventing the need for rotation or translocation of the tRNA acceptor stem as polymerization proceeds.


   EXPERIMENTAL PROCEDURES
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
RESULTS AND DISCUSSION
 REFERENCES
 
Threading the S. shibatae CCA-adding Sequence—The Sulfolobus protein sequence (5) was submitted to the following threaders for fold recognition: 3D-PSSM (28), BIOINBGU (29), SAM-T02 (30), FUGUE (31), and MGENTHREADER (32). All five threaders predicted with high confidence (Table I) that the CCA-adding enzyme adopts the same fold as yeast (PDB1 file 1FA0 [PDB] (23)) and bovine poly(A) polymerase (PDB file 1F5A [PDB] (24)). Our mutational analysis was based on the model of the Sulfolobus CCA-adding enzyme generated by 3D-PSSM using the yeast poly(A) polymerase coordinates. As a negative control, the Escherichia coli class II CCA-adding enzyme was submitted to the five threaders, but no significant hits were found until release of the coordinates for the Bacillus stearothermophilus CCA-adding structure (16) or the archaeal A. fulgidus CCA-adding structure (21, 26, 27). Site-directed Mutagenesis—Constructs expressing the recombinant Sulfolobus CCA-adding enzyme (14, 17) were mutated using the Quick-ChangeTM kit from Stratagene. Mutations were generated by PCR using sense and antisense mutagenic oligonucleotide primers and Pfu DNA polymerase. The methylated wild type template was digested by DpnI, which cuts only dam-methylated DNA. The surviving nicked and double-stranded plasmids were transformed into E. coli XL1-Blue supercompetent cells, and the presence of each mutation was verified by sequencing.


View this table:
[in this window]
[in a new window]
  		TABLE I
Search for structural homologs of the class I S. shibatae CCA-adding enzyme

 

To construct deletion (D4, amino acids 92–95; D5, amino acids 90–94; and D6, amino acids 90–95) and substitution (amino acids 89–96 of Sulfolobus to 150–157 of bovine poly(A) polymerase) of the {beta}-turn mutants (Fig. 3B), the following four primers were used: two nonmutagenic outside primers (T-up, 5'-GCGAAATTAATACGACTCACTATAG-3', and T-down, 5'-GGTTATGCTAGTTATTGCTCAGCGG-3'), and two mutagenic inside primers (D4-up, 5'-AAAAGATCTAGATTATACTTTAGCATACGCTGTGATAGTCAATATAAA-3', and D4-down, 5'-TTTATATTGACTATCACAGCGTATGCTAAAGTATAATCTAGATCTTTT-3'; D5-up, 5'-AAAAGATCTAGATTATACTTTAGCATATGTGATAGTCAATATAAA-3', and D5-down, 5'-TTTATATTGACTATCACATATGCTAAAGTATAATCTAGATCTTTT-3'; D6-up, 5'-AAAAGATCTAGATTATACTTTAGCAGTGATAGTCAATATAAATAACG-3', and D6-down, 5'-CGTTATTTATATTGACTATCACTGCTAAAGTATAATCTAGATCTTTT-3'; and S8-up, 5'-ATCTAGATTATACTTTAGAAGAAGCTTTTGTTCCAGTTATGATAGTCAAT-3', and S8-down, 5'-ATTGACTATCATAACTGGAACAAAAGCTTCTTCTAAAGTATAATCTAGAT-3', where mutations are underlined). The PCR1 product was generated by using nonmutagenic upstream and mutagenic downstream primers, and the PCR2 product was generated by using mutagenic upstream and nonmutagenic downstream primers. The PCR1 and PCR2 products were gel-purified and used as templates with the two nonmutagenic outside primers to generate the recombinant PCR3 product. The PCR3 products were gel-purified, digested with NdeI and XhoI, ligated into pET22b, and transformed into DH5-electrocompetent cells. Constructs conferring ampicillin resistance were sequenced to confirm each mutation.



View larger version (72K):
[in this window]
[in a new window]
  		FIG. 3.
Substitution or deletion of the {beta}-turn can affect CCA addition. Sequences surrounding the {beta}-turn in class I archaeal CCA-adding enzymes and eukaryotic poly(A) polymerases were aligned using ClustalW (43, 52). Conserved {beta}-turn sequences are highlighted in green, and conserved {beta}-strand sequences in orange. A, class I archaeal CCA-adding enzymes. GenBankTM accession numbers are in parentheses as follows: S. shibatae (U66004 [GenBank] ); Sulfolobus solfataricus (NP_342511 [GenBank] ); Sulfolobus tokodaii (NP_376857 [GenBank] ); Pyrococcus horikoshii (NP_142115 [GenBank] ); Pyrococcus abyssi (NP_125796 [GenBank] ); Pyrococcus furiosus (NP_577755 [GenBank] ); Archaeoglobus fulgidus (NC_000917 [GenBank] ); Aeropyrum pernix (NP_148168 [GenBank] ); Methanococcus jannaschii (NP_248104 [GenBank] ); M. maripaludis (NP_988069 [GenBank] ); Methanobacterium thermoautotricum (NP_275727 [GenBank] ); Methanopyrus kandleri (NP_614649 [GenBank] ); Picrophilus torridus (NC_005877 [GenBank] .1); Methanococcoides burtonii (ZP_00148440); Ferroplasma acidarmanus (ZP_00307239); Thermoplasma acidophilum (NP_394900 [GenBank] ); Thermoplasma volcanium (NP_110636 [GenBank] ); Halobacterium sp. NRC-1 (NP_279277 [GenBank] ); Methanosarcina barkeri (ZP_ 00295845); and Methanosarcina mazei (NP_632493 [GenBank] ). Three block deletions in the S. shibatae {beta}-turn are boxed, and a block substitution is highlighted in blue. B, eukaryotic poly(A) polymerases. GenBankTM accession numbers are in parentheses as follows: bovine (P25500 [GenBank] ); human (P51003 [GenBank] ); human testis-specific (Q9NRJ5); mouse (Mus musculus) (NP_766143 [GenBank] ); mouse, (M. musculus), testis-specific (NP_064327 [GenBank] ); Rattus norvegicus (AAH79046 [GenBank] ; chicken, Gallus gallus (AAC16061 [GenBank] ; zebrafish, Danio rerio (NP_956915 [GenBank] ); goldfish, Carassius auratus (BAB39139 [GenBank] ; Xenopus laevis (AAH61931 [GenBank] ; Caenorhabditis briggsae (CAE64813 [GenBank] ; Caenorhabditis elegans (NP_496067 [GenBank] ); C. elegans, Gld-2 (AAM94369S); Drosophila melanogaster (AAF36978 [GenBank] ; budding yeast, S. cerevisiae (P29468 [GenBank] ); fission yeast, Schizosaccharomyces pombe (CAA22808 [GenBank] ; Aspergillus nidulans (EAA65903 [GenBank] ; rice, Oryza sativa (CAE04327 [GenBank] ; Arabidopsis thaliana (CAB80002 [GenBank] . C, the {beta}-turn is required for CCA addition. A gel electrophoresis assay was used to measure incorporation of [{alpha}-32P]CTP or [{alpha}-32P]ATP (indicated as CTP* and ATP*) into a tRNA substrate lacking A, CA, or CCA (tRNA-NCC, tRNA-NC, or tRNA-N where N is the discriminator base). Assays are as described under "Experimental Procedures," with products resolved by 12% denaturing PAGE, and activity quantified by phosphorimaging (Amersham Biosciences). The average of three independent experiments is shown, with error bars indicating standard deviation.

 
CCA-adding Enzyme Expression and Assay Conditions—Hexahistidine-tagged Sulfolobus CCA-adding enzymes and tRNA substrates lacking CCA, CA, and A were prepared as described (33). All tRNA substrates were derived by runoff transcription from a B. subtilis aspartate tRNA construct (34) or a modification thereof (35). Uniformly labeled tRNAs were transcribed for 30 min at 37 °C in a reaction containing 40 mM Tris-HCl (pH 8.0 at 20 °C), 6 mM MgCl2, 10 mM dithiothreitol, 2 mM spermidine, 20 units/µl T7 RNA polymerase (Roche Applied Science), 500 µg/ml DNA template, 500 µM ATP, CTP, and GTP, 50 µM UTP, and 50 µCi of [{alpha}-32P]UTP (3,000 Ci/mmol, Amersham Biosciences).

tRNA transcripts were purified by denaturing 12% PAGE, visualized by UV shadowing, and carefully excised to remove any aberrant products. Standard 10-µl CCA addition assays contained 100 mM glycine/NaOH (pH 9.0), 10 mM MgCl2, 1 mM dithiothreitol, 5 µM CTP, 50 µM ATP, 0.25 µM tRNA substrates, 100 nM [{alpha}-32P]CTP or [{alpha}-32P]ATP (3,000 Ci/mmol, Amersham Biosciences), and 10 nM purified protein at 70 °C for 5 min. The reactions were terminated by addition of 5 µl of 95% formamide containing 20 mM sodium EDTA (pH 8.0), xylene cyanol (0.2%), and bromphenol blue (0.2%). Products were resolved by 12% denaturing PAGE and quantified using a Storm PhosphorImager (Amersham Biosciences).

Mutants were screened for UTP or GTP incorporation under standard assay conditions using 100 nM [{alpha}-32P]UTP or [{alpha}-32P]GTP in the presence of all four unlabeled ribonucleoside triphosphates (1 mM), 2 µM tRNA substrates, and 20 nM of enzyme at 70 °C for 15 min. Km values for GTP were determined at 2 µM tRNA-NCC by a linear reciprocal plot of initial velocities against nucleotide concentrations from 50 to 2000 µM GTP.

Kinetic parameters were assayed by addition of [{alpha}-32P]CTP to tRNA-NC under standard assay conditions. Km values for CTP were determined at 2 µM tRNA by a linear reciprocal plot of initial velocities against nucleotide concentrations from 6.25 to 50 µM CTP.

Gel mobility shift assays were carried out as described (13) using purified CCA-adding enzyme and 5 x 104 cpm (1 pmol) of uniformly labeled tRNA lacking CA (tRNA-NC). The 10-µl binding reactions contained 160 mM KCl, 30 mM Tris-HCl (pH 7.5), 0.1 mM dithiothreitol, 0.1 mM EDTA, 5 mM MgCl2, and 10% glycerol. After incubation at 25 °C for 15 min in the absence of CTP or ATP, the reactions were chilled on ice and loaded on a 10% nondenaturing gel (29:1 acrylamide:bisacrylamide, 20 cm long x 0.4 mm thick) in 1x TBE for electrophoresis at 220 V and at room temperature. Assays were quantified using a Storm Phosphor-Imager (Amersham Biosciences). The ratio of free to bound tRNA was measured as a function of enzyme concentration, and Kd was calculated from the slope of a double-reciprocal plot (36).

Incorporation of CTP analogs was assayed in a standard 10-µl reaction containing 2 pmol of uniformly labeled tRNA-NC and 10 nM of purified recombinant CCA-adding enzyme. The reactions were incubated at 70 °C for 15 min, and the tRNA products were resolved by high resolution denaturing 12% PAGE (20 x 48-cm gel at 800 V for 16 h) and autoradiographed or quantitated by PhosphorImager (Amersham Biosciences).


   RESULTS AND DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS AND DISCUSSION
REFERENCES
 
Modeling the Structure of the S. shibatae CCA-adding Enzyme—To understand how CCA-adding enzymes add CCA without using a nucleic acid template (2, 5) and without translocation or rotation of the growing tRNA on the enzyme (13), we sought to identify key residues involved in nucleotide selection, primer positioning, and tRNA binding in the class I archaeal S. shibatae CCA-adding enzyme. As structures had not yet been determined for any CCA-adding enzyme, we asked five different protein fold recognition programs ("threaders") to search the protein structural data base for potential homologs of the Sulfolobus enzyme (5). As described under "Experimental Procedures," all five programs modeled the Sulfolobus structure on the bovine (24) and yeast poly(A) polymerase crystal structures (23). This was not surprising. Archaeal CCA-adding enzymes and eukaryotic poly(A) polymerases are both class I members of the nucleotidyltransferase superfamily (see Refs. 3–5 and 37), and the primary sequences can be readily aligned over an N-terminal region of about 30 kDa (see supplemental Fig. S1), although the C-terminal regions are highly divergent and cannot be aligned (24). In contrast, eubacterial and eukaryotic CCA-adding enzymes, as well as eubacterial poly(A) polymerases, are class II nucleotidyltransferases that can be aligned with class I only within the nucleotidyltransferase domain (5).

The Sulfolobus structure predicted by the 3D-PSSM threader is shown in Fig. 1A; for convenience, selected results from our mutational analysis are superimposed on the model and discussed below. The modeled Sulfolobus structure has the same nucleotidyltransferase "palm" domain as the poly(A) polymerases; five {beta}-strands backed by two {alpha}-helices form a platform supporting two carboxylates (5, 37) that are required for the nearly universal two metal ion mechanism of phosphoester bond transfer (6, 7). In addition to the catalytic palm domain, the homology model extends to nearby helices of the "fingers" domain that participate in nucleotide recognition in other polymerases (38, 39), and may also do so in eukaryotic poly(A) polymerases (see Refs. 16, 23, and 24). The predicted Sulfolobus structure (Fig. 1A) agrees well with the actual structure of the highly homologous A. fulgidus CCA-adding enzyme (Fig. 1B), which was determined subsequently (26, 27).



View larger version (56K):
[in this window]
[in a new window]
  		FIG. 1.
Actual and predicted structures of class I archaeal CCA-adding enzymes. A, predicted structure of the S. shibatae CCA-adding enzyme, with selected mutations and phenotypes superimposed as in the key below. B, actual structure of the highly homologous A. fulgidus CCA-adding enzyme (PDB 1TFW [PDB] (21)). The Sulfolobus protein sequence was submitted to five threaders for fold recognition, but only the model generated by 3D-PSSM using the yeast poly(A) polymerase coordinates (PDB 1FA0 [PDB] (23)) is shown here, as all five threaders were in substantial agreement. Secondary structure elements supporting the active site carboxylates are indicated in orange and the conserved {beta}-turn in green; orange {beta}-strands are numbered in white, and nearby {alpha}-helices are indicated in blue or violet. The violet helices correspond to helix H (Archaeoglobus CCA-adding enzyme (21)), helix G (bovine poly(A) polymerase (24)), and helix F (yeast poly(A) polymerase (23)); the blue helices begin with helix F(Archaeoglobus), helix H (bovine poly(A) polymerase), or helix G (yeast poly(A) polymerase) and continue into helix G (Archaeoglobus CCA-adding enzyme, helix I (bovine poly(A) polymerase), or helix H (yeast poly(A) polymerase). For scale, selected protein side chains are shown as stick models with nitrogen as blue, carbon as yellow, and oxygen as red; side chain orientation in A is arbitrary, but in B is seen in a cocrystal of the enzyme with a tRNA-NCC75 mimic and incoming ATP. The incoming ATP in the cocrystal structure is shown as a stick model in cyan (B). For clarity, tRNA substrate and bound divalent metal ions are omitted.

 
An Evolutionarily Versatile {beta}-Turn in Class I Nucleotidyltransferases—We suggested previously that CCA addition might reflect collaborative templating, meaning that the template for nucleotide addition would be neither protein alone nor the growing 3' end of tRNA but an evolving ribonucleoprotein structure that is remodeled upon addition of each successive nucleotide (13). After comparing the predicted structure of the Sulfolobus enzyme (Fig. 1A) with known structures for other class I enzymes, we were particularly intrigued by an evolutionarily versatile {beta}-turn preceding strand 4 of the nucleotidyltransferase motif that serves different purposes in different enzymes (Fig. 2). In the template-dependent rat DNA polymerase {beta} (40, 41), this {beta}-turn holds the DNA template strand in position between the fingers domain and the catalytic palm domain. In the template-independent mouse terminal deoxynucleotidyltransferase, an expanded version of this {beta}-turn occupies the position of the DNA template strand in DNA polymerase {beta}, ensuring untemplated addition to the end of a duplex DNA substrate (42). In the template-independent yeast (23) and bovine poly(A) polymerases (24), the same {beta}-turn may bind the incoming ATP and/or the 3' terminus of the RNA primer. Thus it seemed possible that the equivalent {beta}-turn seen in the predicted structure of the Sulfolobus CCA-adding enzyme (Fig. 1A) might hold the primer terminus, replace the template strand, or serve as a component of a ribonucleoprotein template for CCA addition.



View larger version (69K):
[in this window]
[in a new window]
  		FIG. 2.
An evolutionarily versatile {beta}-turn in the nucleotidyltransferase motif of four members of the class I nucleotidyltransferase superfamily. A, class I archaeal A. fulgidus CCA-adding enzyme (PDB 1TFW [PDB] ). B, Saccharomyces cerevisiae poly(A) polymerase (PDB 1FAO [PDB] ). C, rat DNA polymerase {beta} (PDB 2BPG [PDB] ). D, mouse terminal deoxynucleotidyltransferase (PDB 1KEJ [PDB] ). All structures are oriented as in Fig. 1. The head or nucleotidyltransferase domain is shown in magenta, the neck (and fingers when present) in blue, and the evolutionarily versatile {beta}-turn connecting strands 3 and 4 in green. The incoming nucleotide is shown as a space-filling model with phosphorus as violet and the other colors as in Fig. 1. The three catalytic carboxylates are displayed as stick representations, using the same color scheme as in the space-filling models. For clarity, RNA and DNA substrates, some connecting loops, and bound divalent metal ions are omitted. For consistency, the actual structure of the Archaeoglobus CCA-adding enzyme (21) is shown, although the experimental work reported here was based on a predicted structure (Fig. 1A) and was substantially completed before the crystal structure was determined. These and all other images were drawn in PyMOL (51).

 
The {beta}-turn sequence is highly conserved among archaeal CCA-adding enzymes (Fig. 3A) but exhibits no primary sequence homology with class I eukaryotic poly(A) polymerases other than a proline at the junction between the {beta}-turn and strand 4, which may incline the {beta}-turn relative to the nucleotidyltransferase platform (Fig. 3B). To begin to explore the functional importance of the {beta}-turn, we made three block deletions and one block substitution in which we replaced the Sulfolobus {beta}-turn residues (89–96) with the corresponding residues (150–157) from bovine poly(A) polymerase in hopes of converting the CCA-adding enzyme into an oligo(A)- or poly(A)-adding enzyme (Fig. 3B). We expressed and assayed the mutant CCA-adding enzymes for addition of labeled CTP or ATP to the unlabeled substrates tRNA-N, tRNA-NC, and tRNA-NCC (Fig. 3C). We also screened the mutant enzymes for misincorporation of UTP or GTP and, in light of the close sequence relationship between archaeal CCA-adding enzymes and eukaryotic poly(A) polymerases (5, 37, 43), for poly(C), poly(A), or poly(A,C) polymerase activity (data not shown). The block deletion mutants of 5 or 6 residues (D5, {Delta}90–94, and D6, {Delta}90–95) severely inhibited CCA addition, but deletion of only 4 residues (D4, {Delta}92–95) retained 20% C75-adding activity and 5–10% C74- and A76-adding activity (Fig. 3C). Substitution of the Sulfolobus {beta}-turn by the corresponding bovine poly(A) polymerase residues in mutant S8 did not affect C75 addition but severely inhibited addition of C74 and A76. These data confirmed the importance of the highly conserved {beta}-turn for CCA-adding activity, and the curious specificity of the D4 and S8 mutations suggested that the {beta}-turn might indeed play a role in positioning the primer strand or templating CCA addition.

Mutations Near the Active Site Can Differentially Affect Addition of C74, C75, or A76—We next constructed a series of missense mutants to explore the function of the Sulfolobus {beta}-turn in greater detail (Fig. 4). Although none of these mutants had oligo(A) or poly(A) polymerase activity, we were surprised to find mutations that specifically affected C74, C75, or A76 addition, or combinations thereof (Fig. 4, Table II, and data not shown). Mutant H93V, in which Sulfolobus H93 is changed to match V154 in bovine poly(A) polymerase, adds C74 and C75 but fails to add A76. Y95V, in which Sulfolobus Y95 is changed to match V156 of bovine poly(A) polymerase, adds C75 and A76 but not C74. On the other hand, two other mutations that change the Sulfolobus {beta}-turn to match bovine poly(A) polymerase had nonspecific effects; Y90E was nearly wild type for CCA addition despite replacement of a conserved bulky hydrophobic residue by a carboxylate, whereas E92F was nearly inactive for CCA addition, although tRNA binding was apparently normal as discussed below (Fig. S3). Finally, the double mutants H93V/Y95V and Y90E/E92F were completely inactive for CCA addition; H93V/Y95V could not add C75, as might have been expected if addition of C74, C75, and A76 were independent, nor did Y90E rescue E92F (Fig. 4).



View larger version (36K):
[in this window]
[in a new window]
  		FIG. 4.
Mutations in the vicinity of the Sulfolobus CCA-adding enzyme {beta}-turn differentially affect addition of C74, C75, or A76. Mutants are designated using the single letter code and residue number; for example, H93V indicates mutation of histidine 93 to valine. WT, wild type. Assays performed in triplicate as in Fig. 3C; variation did not exceed 9%.

 


View this table:
[in this window]
[in a new window]
  		TABLE II
Mutations surrounding the predicted active site of the S. shibatae CCA-adding enzyme can affect C, A, or CA addition

 
The ability of H93V to add CTP but not ATP, and of Y95V to add C75 and A76 but not C74, demonstrated that addition of C74, C75, and A76 can be separated by mutation and that C74 and C75 addition do not simply reuse the same determinants. The failure of the H93V/Y95V double mutant to add C75 further indicated that mutations affecting ATP and CTP addition interact with each other, suggesting that CTP and ATP addition share common determinants. The variety and specificity of these phenotypes encouraged us to generate a larger collection of mutations surrounding the active site of the Sulfolobus enzyme and including helices E, F, and G as well as the nucleotidyltransferase domain (Fig. S1). All mutants were assayed with tRNA substrates lacking A, CA, or CCA, and the data are presented in Table II. The phenotypes of selected mutants are superimposed on the predicted Sulfolobus structure (Fig. 1B).

Archaeoglobus Structures Explain Many of the Sulfolobus Mutant Phenotypes—The Sulfolobus and Archaeoglobus CCA-adding enzymes are highly homologous class I enzymes (Fig. S1). Archaeoglobus structures can therefore be used to interpret the phenotypes of Sulfolobus mutants, and the Sulfolobus data can in turn provide an independent test of conclusions drawn from the Archaeoglobus cocrystal structures (21, 26, 27).

As background for the structural analysis, we first describe the four major conclusions from the class I crystal: 1) the class I nucleotide-binding site is very different from class II; 2) the enzyme uses a ribonucleoprotein template; 3) the catalytic head domain moves away from the neck domain as CCA addition proceeds; and 4) the {beta}-turn must adopt four consecutive conformations in order to present the discriminator base N73, C74, and C75 to the active site and to terminate the reaction after CCA addition is complete, without requiring translocation or rotation of the tRNA acceptor stem.

In the class II enzyme from B. stearothermophilus (16), CTP and ATP addition are directed by a pure protein template composed of R157 and D154, which effectively base pairs almost identically with the Watson/Crick edges of CTP and ATP (Fig. 5, right). Presumably, only a small conformational change is required to accommodate the larger purine after the pyrimidines C74 and C75 have been added. The nucleotide-binding site of another class II enzyme, from Aquifex aeolicus, almost certainly behaves similarly (22).



View larger version (24K):
[in this window]
[in a new window]
  		FIG. 5.
Archaeoglobus structures can explain Sulfolobus phenotypes. Comparison of base-specific interactions with incoming CTP in class I and class II CCA-adding enzymes. Left, class I enzyme from Archaeoglobus. The base interacts with a ribonucleoprotein template formed collaboratively by protein (H129 and R221) and the 3' end of tRNA (additionally stabilized by R125). Coordinates are for the Archaeoglobus complexes (21), but for clarity, residue positions are labeled according to the homologous Sulfolobus enzyme (see Fig. S1). Right, class II enzyme from B. stearothermophilus. The base interacts with a pure protein template (D154 and R157). The polypeptide side chains and incoming nucleotide are shown as stick models with colors as in Fig. 2; other nucleotides are shown in wheat, with phosphate oxygens red, except for C74 which is shown as a thin gray line. Hydrogen bonds are shown as dashed lines in black for those seen in the crystal structure (21) and in red for those implied by the mutational analysis. The exocyclic N4 group of the incoming CTP is circled in red; this group interacts with the backbone phosphates of A73 and C74 but is missing in zebularine (Fig. 8). The main chain of the evolutionarily flexible {beta}-turn is indicated in green as in Figs. 1 and 2.

 
In sharp contrast, the class I enzyme from Archaeoglobus uses a ribonucleoprotein template for CTP and ATP addition. Unlike class II enzymes, where the pure protein template enables the apoenzyme to bind CTP or ATP in the absence of tRNA substrate (16), the Archaeoglobus apoenzyme cannot bind nucleotides specifically (26, 27) because the 3' end of the tRNA substrate is an integral part of the nucleotide-binding site (Fig. 5, left) (21) as predicted by the collaborative templating model (13).

The location of the {beta}-turn close to the active site (Figs. 1, 2, 3) and the phenotypes of the Sulfolobus {beta}-turn mutants (Fig. 4 and Table II) suggested that this structure plays a major role in templating CCA addition; however, crystallographic snapshots of C75 and A76 addition, as well as a complex of mature tRNA with the Archaeoglobus enzyme (21), clinched the case (Fig. 6). As seen in the crystal structures, one task of the {beta}-turn is to present the 3' end of the growing tRNA to the incoming nucleotide bound at the active site; the other task is to position the 3' end of the tRNA so it can function as the RNA component of the ribonucleoprotein template for nucleotide addition. These two tasks are necessarily linked because 1) the 3' end of the tRNA is sandwiched between the {beta}-turn and the protein components of the ribonucleoprotein template, and 2) the enzyme must read the length and sequence of the 3' end to know whether CTP or ATP addition is required or CCA addition is complete.2



View larger version (37K):
[in this window]
[in a new window]
  		FIG. 6.
Snapshots of CCA addition by the Archaeoglobus enzyme. Coordinates are from PDB 1TFW [PDB] (21). A, complex with tRNA-NC74 and incoming CTP in orange. B, complex with tRNA-NCC75 and incoming ATP in red. C, complex with mature tRNA. Orientation is approximately as in Fig. 5. The asterisk in D53* indicates that the equivalent residue in Archaeglobus is E59. Nucleotides A73, C74, C75, and A76 are shown in pink, blue, orange, and red, respectively; the acceptor stem is shown in wheat; and protein side chains are shown with carbon in green ({beta}-turn) or yellow (elsewhere), nitrogen in blue, oxygen in red, and phosphorus in violet. Metal ions A and B are shown as cyan spheres; hydrogen bonds as dashed lines.

 

The {beta}-turn accomplishes these linked tasks by adopting four consecutive conformations that bulge (or scrunch) the growing 3' end, thus positioning each new 3'-hydroxyl group identically for attack on the next incoming nucleotide (Fig. 6). The second (C75 addition), third (A76 addition), and fourth conformations (mature CCA end) are seen in the cocrystal structures (21); a cocrystal corresponding to the first conformation (C74 addition) has remained elusive. To enable the tRNA acceptor stem to remain immobile on the enzyme as C75 and A76 are added (13), the enzyme makes room for the additional nucleotides by repositioning the catalytic head domain further from the neck domain (21) (Fig. 7). Although repositioning affects many residues in the vicinity of the active site, these motions are largest for the {beta}-turn itself (Fig. 7A) and especially for the side chains emanating from the turn, which undergo an intricate, coordinated set of rotations and displacements (Fig. 7B). Note, however, that repositioning of the head domain relative to the neck does not cause reshaping of the {beta}-turn but rather provides a context in which it can occur.



View larger version (32K):
[in this window]
[in a new window]
  		FIG. 7.
Conformational changes in the flexible {beta}-turn of the head domain. Coordinates are from PDB 1TFW [PDB] (21). A, superposition of three Archaeoglobus CCA enzyme complexes: with tRNA-NC74 mimic and incoming CTP (magenta, PDB 1TFY [PDB] ); with tRNA-NCC75 mimic and incoming ATP (blue, PDB 1TFW [PDB] ); and with mature tRNA (yellow, PDB 1SZ1 [PDB] ). Complexes were aligned by superposing the tRNA-binding domains: left, selected protein side chains only; center, superposition of left and right panels; right, tRNA only. Incoming CTP and ATP are shown in orange and red, respectively; tRNA-NC74 and tRNA-NCC75 are shown in violet and cyan, respectively; mature tRNA was omitted for clarity. B, movement of {beta}-turn residues during stepwise CCA addition. Images and orientation are as in A. Sulfolobus nomenclature with equivalent A. fulgidus residues are shown in parentheses.

 
{beta}-Turn Mutations Can Affect Templating Functions—Use of a ribonucleoprotein template by class I enzymes complicates the mutant phenotypes and the corresponding structural explanations. For example, we had previously used an array of CTP and ATP analogs to explore the nature of the nucleotide-binding site in class I and class II CCA-adding enzymes, and we had been able to predict several of the key interactions seen by crystallography (16, 21, 35). Although the CTP analog zebularine is a good substrate for the wild type Sulfolobus enzyme (35), we were surprised to find that {beta}-turn mutants H93V and Y95V (Archaeoglobus H97 and Y99) would not incorporate zebularine into tRNA-NC (Fig. 8). The simplest explanation for the H93V and Y95V phenotypes is that bifurcated hydrogen bonds formed between the exocyclic N4 of incoming C75 and the backbone phosphates of A73 and C74 in the ribonucleoprotein template (Fig. 5) help to position the 3' end of tRNA for attack on the incoming nucleotide. Zebularine, lacking exocyclic N4 (Fig. 8), would be unable to interact with the backbone phosphates and would therefore be more dependent on the ability of residues in the {beta}-turn to properly position the 3' end of tRNA. Thus, because the 3' end of tRNA is sandwiched between the {beta}-turn on one side and protein components of the ribonucleoprotein template on the other, mutations in bulging (or "scrunching") functions can affect templating functions.



View larger version (42K):
[in this window]
[in a new window]
  		FIG. 8.
{beta}-Turn mutations can selectively inhibit zebularine incorporation into tRNA-NC. Addition of CTP analogs zebularine, pseudoisocytidine, and 6-azacytidine was assayed as in Fig. 3C but using 200 µM analog triphosphate, 2 µM uniformly 32P-labeled tRNA-NC substrate, and 10 nM mutant enzyme.

 
H129 and R221 are Protein Components of a Ribonucleoprotein Template—The Archaeoglobus crystal structures (21) indicate that H129 (Archaeoglobus H133) forms similar hydrogen bonds with the 2'-hydroxyls of incoming CTP and ATP (Figs. 5 and 6), suggesting that H129 might serve only to discriminate NTPs from the dNTPs that are not substrates for CCA-adding enzymes (1, 2, 5). Yet H129A inhibits addition of C74 and C75 but not A76 (Table II). The simplest interpretation is that H129 forms an additional hydrogen bond, not seen in the crystal structure, with O2 of the incoming CTP, but cannot do so with incoming ATP (Fig. 5, red dashed line, and Fig. 6). This additional interaction could also help to explain why the apparent Km value of the Sulfolobus enzyme for ATP is manyfold higher than for CTP (5, 14). Most interestingly, the guanidinium group of R221 (Archaeoglobus R224) forms bridging hydrogen bonds between tRNA backbone phosphates and N3 of CTP or N1 of ATP (21), thereby helping to position the 3' end of tRNA in the active site. These R221 interactions blur the boundaries between tRNA as substrate and tRNA as part of the ribonucleoprotein template for nucleotide addition, as anticipated by the collaborative templating model (13). Accordingly, the R221A and R221E mutations in this ribonucleoprotein template severely inhibit C74, C75, and A76 addition (Table II).

A Role for the Third Carboxylate in CCA Addition by Class I Enzymes—A single active site must be responsible for addition of all three nucleotides in the class I enzymes, because mutation of either of the two highly conserved carboxylates in the nucleotidyltransferase motif of the Sulfolobus enzyme (D53 and D55, see Fig. S1) abolishes both CTP and ATP addition (14). This is the expected result for phosphoryltransfer reactions catalyzed by a two metal ion mechanism (6, 7) and is consistent with binding of two metal ions by the catalytic carboxylates E59 and D61 in the complex of the Archaeoglobus enzyme with tRNA-NC74 and incoming CTP (Fig. 9, left); metal "A" is absent from the complex with tRNA-NCC75 and incoming ATP (Fig. 9, right) because tartrate buffer was added to prevent reaction by chelating Mg2+ (21).



View larger version (41K):
[in this window]
[in a new window]
  		FIG. 9.
The third carboxylate D106 plays a significant role in A76 addition only. Coordinates are from PDB 1TFW [PDB] (21). Geometries surrounding the metal ions (cyan) are shown for complexes of tRNA-NC74 with incoming CTP and for tRNA-NCC75 with incoming ATP.

 
Although three carboxylates are absolutely required for the activity of two other class I nucleotidyltransferases, DNA polymerase {beta} (40, 41) and bovine poly(A) polymerase (24), we were unable to identify an essential third carboxylate in the Sulfolobus enzyme; rather, we found that D106A selectively impaired addition of A76 but not C74 or C75 (14).

The Archaeoglobus cocrystal structures may now provide an explanation for this surprising result. D106 (Archaeoglobus D110) apparently plays no role in C75 addition (Fig. 9, left), but it forms a hydrogen bond with the 3'-ribose hydroxyl of C75 during A76 addition (Fig. 9, right). Although the absence of metal A in the A76-adding complex (Fig. 9, right) is a complicating factor, it seems likely that the missing metal A would interact with D106, as well as with D55, and the ribose hydroxyl as in the previous step (Fig. 9, left). The implication is that refolding of both C74 and C75 would bring the C75 ribose 3'-hydroxyl within reach of D106 (2.98 Å), whereas bulging (or scrunching) of C74 alone would not (4.02 Å). Thus selective inhibition of A76 addition by D106A may imply that the primary function of D106 in CCA-adding enzymes of class I is to position C75 and not to participate in catalysis. In this unusual polymerase where a small number of protein side chains operating in tight quarters must substitute for a nucleic acid template, it may not be possible to bring a third carboxylate into play until the terminal A addition.

None of the additional mutations in and around D106 had as strong an effect as D106A (Table II). The conservative substitution, D106E, was less selectively impaired than D106A, but D106Q and D106N were nearly normal, consistent with the ability of amides to hydrogen-bond to a hydroxyl or chelate divalent metals (44) (metallo.scripps.edu; see Ref. 4). Nearby alanine substitutions in V103, V105, and V108 also had no apparent effect (Table II).

Mutants Exhibiting Progressive Inhibition, A76 > C75 > C74 or C74 > C75 > A76 —R125 (Archaeoglobus R129) corresponds to R186 in yeast poly(A) polymerase, where it may interact with the N6 amino group of the primer base or select for adenine over guanine (23). Sulfolobus R125A retained 30% C74-adding activity, 10% C75-adding activity, and 5% A76-adding activity (Fig. 4 and Table II). This unexpected phenotype initially suggested that each successive nucleotide is more difficult to add than the last, as might be expected if R125A affected a pocket that is progressively filled by the growing 3' end as postulated by both the collaborative templating (13) and double scrunching models (16). Moreover, R125A was not the only mutation that affected CCA addition progressively (Fig. 4). R125Y (Archaeoglobus R129), Y90A (Archaeoglobus Y94), V96A (Archaeoglobus V100), and I97A (Archaeoglobus H101) also inhibit A76 > C75 > C74, albeit more mildly than R125A, whereas V54A (Archaeoglobus V60) and P94A (Archaeoglobus P98) have the opposite effect of inhibiting C74 > C75 > A76 (Table II).

As anticipated by modeling (Fig. 1A) and confirmed by the crystal structure (Figs. 1B and 10), all mutations causing progressive inhibition mapped in the vicinity of the active site, and all appear to fall into the following two categories: those within the {beta}-turn (Y90A, P94A, V96A, and I97A) and those close to the incoming nucleotide (V54A and R125A). V96 and I97 are located in the stalk of the {beta}-turn, P94 in the turn itself, and could affect the ability of the {beta}-turn to assume consecutive conformations as CCA is added (Fig. 10). Although V96 and I97 exhibit no significant conformational changes upon C75 and A76 addition (data not shown), Y90 undergoes a significant displacement and rotation during C75 and A76 addition as seen for other {beta}-turn residues (Fig. 7B). With Y90 located at the interface between the head and neck domains, these conformational changes induced by the growing 3' end of tRNA could wedge the head and neck domains further apart as CCA addition proceeds (21) (see Fig. 7).



View larger version (60K):
[in this window]
[in a new window]
  		FIG. 10.
Mutations causing progressive inhibition of CCA addition (A76 > C75 > C74 or C74 > C75 > A76) cluster near the active site. Coordinates of the tRNA-NC74 mimic (magenta) with incoming CTP (blue) are from PDB 1TFY [PDB] (21); orientation is approximately as in Figs. 5 and 6. D53 and D55 are shown chelating metal A (also see Fig. 9); V54 emanates from the opposite side of {beta}-strand 2. H129 is shown for reference, and R221 is shown for comparison to R125.

 
V54 is located between the two aspartates, D53 and D55, that position the catalytic metals on the nucleotidyltransferase platform (14), but it is not obvious why the V54A mutation would cause progressive rather than uniform inhibition. Conceivably, repositioning of the head domain relative to the neck domain as CCA addition proceeds does not merely displace (21) but also progressively distorts the head domain (see Fig. 7A).

R125 is also located between the head and neck domains, but closer to the hinge region, where it stabilizes the ribonucleoprotein template by fastening down the critical R221 determinant and helping to position the 3' end of tRNA with a hydrogen bond to the C72 phosphate. Most interestingly, R125 undergoes significant displacement and rotation as CCA addition proceeds (Fig. S4), much like residues in the {beta}-turn (Fig. 7B). Thus progressive inhibition by R125A may reflect failure to reshape the ribonucleoprotein template or failure to wedge the head and neck domains further apart, as CCA addition proceeds (21) (see Fig. 7).

Missense Mutants Rarely Exhibit Compromised Specificity— We screened 48 of 57 mutant CCA-adding enzymes reported here for the ability to use UTP or GTP as substrates. Most surprisingly, none of mutants incorporated UTP in the presence of all four unlabeled ribonucleoside triphosphates (1 mM); only P94A exhibited compromised specificity, incorporating both ATP and GTP into tRNA-NCC, albeit with Km values for GTP and ATP of >3 and ~1.5 mM, respectively (Fig. S2, and data not shown); and even mutants R221A and R221E (Archaeoglobus R224) in a residue normally involved in base-specific recognition of N3 of CTP and N1 of ATP (Figs. 5 and 6) did not incorporate GTP.

The failure of so many mutations surrounding the active site to cause nucleotide misincorporation initially suggested that the CTP- and ATP-binding sites must overlap, as in the pure protein template of the distantly related eubacterial class II CCA-adding enzyme from B. stearothermophilus (16). The Archaeoglobus cocrystal structures subsequently confirmed that the CTP- and ATP-binding sites also overlap in class I enzymes, although class I uses a more complicated ribonucleoprotein template (21, 43). One consequence of overlapping nucleotide-binding sites is that mutations are more likely to affect binding of both CTP and ATP, thus generating an inactive enzyme rather than one with altered nucleotide specificity. Only H129A affects binding of CTP specifically, presumably reflecting loss of a hydrogen bond to the O2 of incoming CTP (Fig. 5). Additional mutants in the vicinity of the incoming nucleotide triphosphate are described in the Supplemental Material.

Choreographed Consecutive Conformations of the {beta}-Turn— As mentioned above, the {beta}-turn must assume at least four consecutive conformations during CCA addition (C74, C75, and A76 addition, followed by a final conformation that prevents further addition to completed CCA). The latter three conformations are seen in the cocrystal structures (21) (Fig. 6), but the first remains a matter of conjecture. However, data arguing against translocation or rotation of the acceptor stem following C74 addition2 suggest that in the first conformation the {beta}-turn wedges between A73 and the C72:G1 base pair, pushing A73 up toward the incoming CTP.3 Most intriguingly, the Y95V mutation (Archaeoglobus Y99) inhibits C74 addition only, suggesting that Y95 may be necessary in the first conformation to present A73 to incoming CTP but to be replaceable thereafter by V (Fig. 3C and Table II). Thus, Y95 may function as a wedge, as seen for tyrosine and phenylalanine side chains at other sharp turns of the nucleic acid backbone within RNA and DNA polymerases (38, 45, 46). Y90 is unlikely perform this function, as Y90A is nearly normal for C74 addition (Table II).

In fact, the {beta}-turn is already known to function twice as a wedge (21). First, during C75 addition, the {beta}-turn drives A91, E92, and H93 between A73 and C74, pinning A73 against the C72:G1 base pair and lifting C74 up to the incoming CTP. Second, during A76 addition, a blunter {beta}-turn creates a floor consisting of Y95, H93, A91, and E92 that continues to pin A73 against the C72:G1 base pair but uses C74 as a wedge to push C75 up toward the incoming ATP. The central role of H93 in forming a flat floor for A76 addition may explain why H93V adds C74 and C75 but not A76 (Fig. 4 and Table II). Valine may be able to stack between A91 and P94 during C75 addition but may be too flexible to form a floor during A76 addition (Fig. 6). The dramatic 90° rotation of H93 before A76 addition lends support to this interpretation (Fig. 7B). Similarly, although the conservative substitution E92Q was almost fully active (Table II), consistent with the ability of an amide to hydrogen-bond with the Watson/Crick edge of C74 during A76 addition (Fig. 6B), the nonconservative substitution E92F is nearly inactive for C75 addition (Fig. 4), although phenylalanine might have been expected to replace E92 by stacking on A73 (Fig. 6A). This suggests that the conformation and packing of the {beta}-turn (Fig. 7B) are as critical for C75 as A76 addition and that reorientation of individual side chains within the {beta}-turn, not simply an integral movement of the {beta}-turn as a whole (Fig. 7), is required for the intricate molecular choreography. This makes it all the more surprising that A91Y (Archaeoglobus A95), which substitutes a bulky residue for a universally conserved alanine within the {beta}-turn (Fig. 3A), was nearly wild type for CCA addition and that Y90A (Archaeoglobus Y94), which substitutes a small residue for the nearly universally conserved tyrosine (Fig. 3A), was not inactive but inhibited C74 > C75 > A76 addition progressively (Table II).

After A76 is added, the completed CCA stacks helically on the acceptor stem (Fig. 6C), pushing the head domain yet further from the neck (Fig. 7) and extending the 3' end of the mature tRNA beyond the reach of the active site. The {beta}-turn facilitates this helical stack by assuming a fourth conformation in which P94, H93, and Y90 stack to form a flat floor with two ridges, E92 and Y95, that interact with the C74 (3.5 Å away) and A76 bases (2.8 Å away), respectively (Fig. 6C). One further complication is that the {beta}-turn is unlikely to go directly from one conformation to the next, because it must almost certainly unfold to release pyrophosphate and admit the incoming nucleotide, and then refold before nucleotide addition can take place (see Supplemental Movie 2 in Ref. 21), just as in the "open" (nucleotide entry) and "closed" (nucleotide addition) states of a conventional RNA or DNA polymerase (6, 47). Thus the growing 3' end of tRNA may be disordered in the absence of incoming nucleotide; binding of the correct nucleotide would lock the 3' end of the tRNA and the {beta}-turn into the correct conformation for addition of that nucleotide. All four of the {beta}-turn conformations would then reflect an induced fit. The correct conformation could only be formed by the ribonucleoprotein complex and not by enzyme or tRNA alone.

tRNA Binding and the Catalytic Site Are Independent—Mutants that fail to add CTP or ATP could be defective in tRNA binding, nucleotide binding, catalysis, or any (as yet unknown) intermediate step. To assay the dissociation constant (Kd) for tRNA, we performed gel mobility shift assays with purified mutant enzymes and labeled tRNA-NC (Fig. S3 and Table III). To assay the apparent Km for CTP binding, we measured addition of [{alpha}-32P]CTP to tRNA-DC in the presence of nearly saturating 1 mM ATP (5) (data not shown). K149A affects the apparent Km for CTP rather than the Kd for tRNA, as expected for a residue whose the Archaeoglobus (K152) or bovine and yeast poly(A) polymerase counterparts (K228 and K215, respectively) appear to bind the {gamma}-phosphate of ATP or the outgoing pyrophosphate (23, 24). In contrast, K153A (Table II) is only modestly defective in Kd for tRNA and Km for CTP (Fig. S3 and Table III), suggesting that this severe mutation mainly affects kcat. Most importantly, none of the mutations we tested surrounding the active site had more than a 3-fold effect on Kd for tRNA binding (Table III), although all had significant defects in CCA addition (Table II). We conclude that tRNA binding, which involves primarily interactions with the acceptor stem (13, 17, 21), does not extend to or depend on binding of the growing 3'-terminal CCA sequence to the enzyme. Anchoring of the tRNA acceptor stem to the enzyme serves the twin purposes of leaving the 3' end free to refold within the catalytic center and favoring (but not requiring) processive CCA addition.


View this table:
[in this window]
[in a new window]
  		TABLE III
Affinity of mutant CCA-adding enzymes for tRNA-NC and CTP

 
Perspective—Our mutational analysis of the Sulfolobus CCA-adding enzyme confirms and extends the conclusion, based on cocrystal structures of the highly homologous class I enzyme from Archaeoglobus (21), that an evolutionarily flexible {beta}-turn characteristic of nucleotidyltransferase domains (Fig. 2) plays a key role in CCA addition by assuming four consecutive conformations: three to specify addition of C74, C75, and A76, and a fourth to prevent further addition when the CCA terminus is complete (Fig. 6). The equivalent {beta}-turn is short and far less highly conserved in the class II enzymes (48), where it may need to work in concert with a long proline-rich loop between strand 5 and helix D that is disordered in both the class II apoenzyme (16, 49) and the cocrystal corresponding to ATP addition (22). Thus, although class I and class II enzymes diverge completely in sequence and structure outside the nucleotidyltransferase domain (16, 26, 27, 49), both classes may exploit the ability of loops at the edges of the conserved nucleotidyltransferase domain to adopt multiple conformations as CCA is added. Moreover, although class II enzymes use a pure protein template (16, 22) instead of the ribonucleoprotein template characteristic of class I (21), in both classes the 3' end of tRNA is nestled between the loop(s) on one side and the protein components of the template on the other side. On the other hand, although the pure protein and ribonucleoprotein templates used by class II and class I are very different, it is not surprising that the two classes respond almost identically to a panel of CTP and ATP analogs (35); each class recognizes the only two Watson/Crick determinants (cytidine N4 and N3, adenine N1 and N6) that are shared by C and A and distinguish them from U and G. Whether these similarities testify to the evolutionary adaptability of the nucleotidyltransferase domain (3, 5, 37), to a common ancestor for class I and class II enzymes, or to evolutionary interconversion of CCA-adding enzymes and poly(A) polymerases (5, 8, 9, 50) remains as much a mystery as ever.


   FOOTNOTES
 
* This work was supported by National Institutes of Health Grant GM59804 (to A. M. W.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. Back

{boxs} The on-line version of this article (available at http://www.jbc.org) contains additional text and Figs. S1-S5. Back

{ddagger} To whom correspondence should be addressed. Tel.: 206-543-1768; Fax: 206-685-9231; E-mail: amweiner{at}u.washington.edu.

1 The abbreviation used is: PDB, Protein Data Bank. Back

2 H-D. Cho and A. M. Weiner, unpublished data. Back

3 H-D. Cho, Y. Chen, G. Varani, and A. M. Weiner, manuscript in preparation. Back


   ACKNOWLEDGMENTS
 
We thank Yong Xiong and Tom Steitz for ongoing discussions throughout the course of this work and for communicating results in advance of publication.



   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS AND DISCUSSION
 REFERENCES
 

  1. Sprinzl, M., and Cramer, F. (1979) Prog. Nucleic Acids Res. Mol. Biol. 22, 1-69[Medline] [Order article via Infotrieve]
  2. Deutscher, M. P. (1982) Enzymes 15, 183-215
  3. Holm, L., and Sander, C. (1995) Trends Biochem. Sci. 20, 345-347[CrossRef][Medline] [Order article via Infotrieve]
  4. Aravind, L., and Koonin, E. V. (1999) Nucleic Acids Res. 27, 1609-1618[Abstract/Free Full Text]
  5. Yue, D., Maizels, N., and Weiner, A. M. (1996) RNA (N. Y.) 2, 895-908
  6. Brautigam, C. A., and Steitz, T. A. (1998) Curr. Opin. Struct. Biol. 8, 54-63[CrossRef][Medline] [Order article via Infotrieve]
  7. Steitz, T. A., and Steitz, J. A. (1993) Proc. Natl. Acad. Sci. U. S. A. 90, 6498-6502[Abstract/Free Full Text]
  8. Tomita, K., and Weiner, A. M. (2001) Science 294, 1334-1336[Abstract/Free Full Text]
  9. Tomita, K., and Weiner, A. M. (2002) J. Biol. Chem. 277, 48192-48198[Abstract/Free Full Text]
  10. Zhu, L., and Deutscher, M. P. (1987) EMBO J. 6, 2473-2477[Medline] [Order article via Infotrieve]
  11. Liu, M., and Horowitz, J. (1994) Proc. Natl. Acad. Sci. U. S. A. 91, 10389-10393[Abstract/Free Full Text]
  12. Aebi, M., Kirchner, G., Chen, J. Y., Vijayraghavan, U., Jacobson, A., Martin, N. C., and Abelson, J. (1990) J. Biol. Chem. 265, 16216-16220[Abstract/Free Full Text]
  13. Shi, P. Y., Maizels, N., and Weiner, A. M. (1998) EMBO J. 17, 3197-3206[CrossRef][Medline] [Order article via Infotrieve]
  14. Yue, D., Weiner, A. M., and Maizels, N. (1998) J. Biol. Chem. 273, 29693-29700[Abstract/Free Full Text]
  15. Li, F., Wang, J., and Steitz, T. A. (2000) J. Mol. Biol. 304, 483-492[CrossRef][Medline] [Order article via Infotrieve]
  16. Li, F., Xiong, Y., Wang, J., Cho, H. D., Tomita, K., Weiner, A. M., and Steitz, T. A. (2002) Cell 111, 815-824[CrossRef][Medline] [Order article via Infotrieve]
  17. Cho, H. D., and Weiner, A. M. (2004) J. Biol. Chem. 279, 40130-40136[Abstract/Free Full Text]
  18. Hou, Y. M. (2000) RNA (N. Y.) 6, 1031-1043
  19. Tomari, Y., Suzuki, T., Watanabe, K., and Ueda, T. (2000) Genes Cells 5, 689-698[Abstract]
  20. Weiner, A. M. (2004) Curr. Biol. 14, R883-R885[CrossRef][Medline] [Order article via Infotrieve]
  21. Xiong, Y., and Steitz, T. A. (2004) Nature 430, 640-645[CrossRef][Medline] [Order article via Infotrieve]
  22. Tomita, K., Fukai, S., Ishitani, R., Ueda, T., Takeuchi, N., Vassylyev, D. G., and Nureki, O. (2004) Nature 430, 700-704[CrossRef][Medline] [Order article via Infotrieve]
  23. Bard, J., Zhelkovsky, A. M., Helmling, S., Earnest, T. N., Moore, C. L., and Bohm, A. (2000) Science 289, 1346-1349[Abstract/Free Full Text]
  24. Martin, G., Keller, W., and Doublie, S. (2000) EMBO J. 19, 4193-4203[CrossRef][Medline] [Order article via Infotrieve]
  25. Panchenko, A. R., and Bryant, S. H. (2002) Protein Sci. 11, 361-370[CrossRef][Medline] [Order article via Infotrieve]
  26. Xiong, Y., Li, F., Wang, J., Weiner, A. M., and Steitz, T. A. (2003) Mol. Cell 12, 1165-1172[CrossRef][Medline] [Order article via Infotrieve]
  27. Okabe, M., Tomita, K., Ishitani, R., Ishii, R., Takeuchi, N., Arisaka, F., Nureki, O., and Yokoyama, S. (2003) EMBO J. 22, 5918-5927[CrossRef][Medline] [Order article via Infotrieve]
  28. Kelley, L. A., MacCallum, R. M., and Sternberg, M. J. (2000) J. Mol. Biol. 299, 499-520[Medline] [Order article via Infotrieve]
  29. Fischer, D. (2000) Pacific Symposium on Biocomputing, January, Hawaii, pp. 119-130, World Scientific (www.cs.bgu.ac.il/~bioinbgu/)
  30. Karchin, R., and Hughey, R. (1998) Bioinformatics 14, 772-782[Abstract/Free Full Text]
  31. Shi, J., Blundell, T. L., and Mizuguchi, K. (2001) J. Mol. Biol. 310, 243-257[CrossRef][Medline] [Order article via Infotrieve]
  32. Jones, D. T. (1999) J. Mol. Biol. 287, 797-815[CrossRef][Medline] [Order article via Infotrieve]
  33. Cho, H. D., Tomita, K., Suzuki, T., and Weiner, A. M. (2002) J. Biol. Chem. 277, 3447-3455[Abstract/Free Full Text]
  34. Oh, B. K., and Pace, N. R. (1994) Nucleic Acids Res. 22, 4087-4094[Abstract/Free Full Text]
  35. Cho, H. D., Oyelere, A. K., Strobel, S. A., and Weiner, A. M. (2003) RNA (N. Y.) 9, 970-981
  36. Carey, J., Cameron, V., de Haseth, P. L., and Uhlenbeck, O. C. (1983) Biochemistry 22, 2601-2610[CrossRef][Medline] [Order article via Infotrieve]
  37. Martin, G., and Keller, W. (1996) EMBO J. 15, 2593-2603[Medline] [Order article via Infotrieve]
  38. Doublie, S., and Ellenberger, T. (1998) Curr. Opin. Struct. Biol. 8, 704-712[CrossRef][Medline] [Order article via Infotrieve]
  39. Franklin, M. C., Wang, J., and Steitz, T. A. (2001) Cell 105, 657-667[CrossRef][Medline] [Order article via Infotrieve]
  40. Sawaya, M. R., Pelletier, H., Kumar, A., Wilson, S. H., and Kraut, J. (1994) Science 264, 1930-1935[Abstract/Free Full Text]
  41. Davies, J. F., II, Almassy, R. J., Hostomska, Z., Ferre, R. A., and Hostomsky, Z. (1994) Cell 76, 1123-1133[CrossRef][Medline] [Order article via Infotrieve]
  42. Delarue, M., Boule, J. B., Lescar, J., Expert-Bezancon, N., Jourdan, N., Sukumar, N., Rougeon, F., and Papanicolaou, C. (2002) EMBO J. 21, 427-439[CrossRef][Medline] [Order article via Infotrieve]
  43. Higgins, D. G., Thompson, J. D., and Gibson, T. J. (1996) Methods Enzymol. 266, 383-402[Medline] [Order article via Infotrieve]
  44. Castagnetto, J. M., Hennessy, S. W., Roberts, V. A., Getzoff, E. D., Tainer, J. A., and Pique, M. E. (2002) Nucleic Acids Res. 30, 379-382[Abstract/Free Full Text]
  45. Eom, S. H., Wang, J., and Steitz, T. A. (1996) Nature 382, 278-281[CrossRef][Medline] [Order article via Infotrieve]
  46. Kiefer, J. R., Mao, C., Braman, J. C., and Beese, L. S. (1998) Nature 391, 304-307[CrossRef][Medline] [Order article via Infotrieve]
  47. Doublie, S. (2004) Structure (Lond.) 12, 1749-1750[Medline] [Order article via Infotrieve]
  48. Martin, G., and Keller, W. (2004) RNA (N. Y.) 10, 899-906
  49. Augustin, M. A., Reichert, A. S., Betat, H., Huber, R., Morl, M., and Steegborn, C. (2003) J. Mol. Biol. 328, 985-994[CrossRef][Medline] [Order article via Infotrieve]
  50. Raynal, L. C., Krisch, H. M., and Carpousis, A. J. (1998) J. Bacteriol. 180, 6276-6282[Abstract/Free Full Text]
  51. DeLano, W. L. (2002) The PyMol Molecular Graphics System, DeLano Scientific, San Carlos, CA
  52. Jeanmougin, F., Thompson, J. D., Gouy, M., Higgins, D. G., and Gibson, T. J. (1998) Trends Biochem. Sci. 23, 403-405[CrossRef][Medline] [Order article via Infotrieve]

Add to CiteULike CiteULike   Add to Complore Complore   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us   Add to Digg Digg   Add to Reddit Reddit   Add to Technorati Technorati    What's this?


This article has been cited by other articles:


Home page
 RNAHome page
H. D. Cho, V. D. Sood, D. Baker, and A. M. Weiner
On the role of a conserved, potentially helix-breaking residue in the tRNA-binding {alpha}-helix of archaeal CCA-adding enzymes
RNA, July 1, 2008; 14(7): 1284 - 1289.
[Abstract] [Full Text] [PDF]

Home page
 Proc. Natl. Acad. Sci. USAHome page
H. D. Cho, C. L. M. J. Verlinde, and A. M. Weiner
Reengineering CCA-adding enzymes to function as (U,G)- or dCdCdA-adding enzymes or poly(C,A) and poly(U,G) polymerases
PNAS, January 2, 2007; 104(1): 54 - 59.
[Abstract] [Full Text] [PDF]

Home page
 J. Biol. Chem.Home page
H. D. Cho, Y. Chen, G. Varani, and A. M. Weiner
A Model for C74 Addition by CCA-adding Enzymes: C74 ADDITION, LIKE C75 AND A76 ADDITION, DOES NOT INVOLVE tRNA TRANSLOCATION
J. Biol. Chem., April 7, 2006; 281(14): 9801 - 9811.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplemental Data
Right arrow All Versions of this Article:
280/10/9555    most recent
M412594200v1
Right arrow Submit a Letter to Editor
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowRequest Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Cho, H. D.
Right arrow Articles by Weiner, A. M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Cho, H. D.
Right arrow Articles by Weiner, A. M.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
 All ASBMB Journals   Molecular and Cellular Proteomics 
 Journal of Lipid Research   ASBMB Today 
Copyright © 2005 by the American Society for Biochemistry and Molecular Biology.
Advertisement
spacer
Advertisement
Advertisement