All known DNA and RNA polymerases catalyze the formation of phosphodiester bonds in a 5′ to 3′ guidance, suggesting this place is a fundamental feature of maintaining and dispersing familial information. The transfer rna His guanylyltransferase ( Thg1 ) is a member of a unique enzyme syndicate whose members catalyze an unprecedented reaction in biology : 3′-5′ addition of nucleotides to nucleic acid substrates. The 2.3-Å crystal structure of homo THG1 ( hTHG1 ) reported here shows that, despite the lack of sequence similarity, hTHG1 shares unexpected structural homology with basic 5′-3′ deoxyribonucleic acid polymerases and adenylyl/guanylyl cyclases, two enzyme families known to use a two-metal-ion mechanism for catalysis. The ability of the same geomorphologic computer architecture to catalyze both 5′-3′ and 3′-5′ reactions raises authoritative questions concerning excerpt of the 5′-3′ mechanism during the development of nucleotide polymerases. All nucleotide polymerases, including deoxyribonucleic acid and RNA polymerases, reverse transcriptase, and telomerase, catalyze nucleotide summation in the 5′ to 3′ direction. The chemical reaction involves the nucleophilic attack of a polynucleotide terminal 3′-OH onto the α-phosphate of an incoming nucleotide, followed by release of the pyrophosphate moiety. Although the 5′ to 3′ direction has been adopted by all polymerases and transferases described to go steady, there is one celebrated exception : the enzyme tRNAHis guanylyltransferase ( Thg1 ). Thg1 catalyzes the highly unusual 3′-5′ summation of a single guanine to the 5′-end of tRNAHis ( 1, 2 ). This reaction is an obligatory step in the growth of this transfer rna because the extra 5′ base, G-1, constitutes a primary identity element for the aminoacyl-tRNA synthetase ( HisRS ) that attaches the amino acid histidine to the 3′-end of the transfer rna ( 3 – 9 ). Thg1 is therefore necessity for maintaining the fidelity of protein deduction. consistent with the critical nature of the G-1 residue, THG1 is an necessity gene in yeast and RNAi-mediated hush of the Thg1 homolog in homo cells results in austere cell-cycle progression and growth defects ( 2, 10, 11 ). Thg1 is widely conserved throughout eukarya, and Thg1 homologs are show in many archaea and bacteria. In eukarya, G-1 addition occurs opposite a universally conserved A73 and therefore is the leave of a nontemplated 3′-5′ addition reaction. In accession, yeast Thg1 catalyzes a second reaction in vitro, extending transfer rna substrates in the 3′-5′ steering in a template-directed manner driven by Watson–Crick pairing ( 12 ). Thg1 enzymes in archaea besides catalyze template-dependent 3′-5′ addition, but do not catalyze nontemplated G-1 addition ( 13 ), suggesting that the templated 3′-5′ accession reaction likely represents an ancestral activity of the earliest Thg1 family members. The 3′-5′ accession of G-1 to tRNAHis occur via three chemical reactions, all catalyzed by Thg1 ( 2, 14 ) ( Fig. 1 ). First, the 5′-monophosphorylated transfer rna that results from RNase P cleavage of pre-tRNAHis is activated using ATP, creating a 5′-adenylylated-tRNAHis intermediate. This adenylylation step mirrors the energizing step in aminoacyl-tRNA synthetases in which the amino acid receives an AMP moiety anterior to being charged on the cognate transfer rna ( 15, 16 ). In the moment dance step, the 3′-hydroxyl of GTP attacks the activate average, yielding triphosphorylated ( palatopharyngoplasty ) G-1-tRNAHis. last, the 5′ pyrophosphate is removed, yielding mature, monophosphorylated ( phosphorus ) G-1-containing tRNAHis. Although these chemical steps are evocative of activities catalyzed by well-studied RNA or DNA ligases or mRNA capping guanylyltransferase ( 17 ), no obvious sequence similarity exists between Thg1 and these or any other known enzyme families to suggest a possible molecular mechanism for Thg1 catalysis.

Thg1 is the only know example of an enzyme that catalyzes templated nucleotide addition in the 3′-5′ direction, opposition to that of all known DNA and RNA polymerases. Thus the molecular mechanism of this enzyme is of great interest and structural portrayal of Thg1 is essential for understanding this probable singular enzymology. We report here the structure of a eukaryal Thg1 family penis determined at 2.3-Å resolution. The social organization of human THG1 ( hTHG1 ) reveals a shared active locate computer architecture with basic 5′-3′ deoxyribonucleic acid polymerases and adenylyl/guanylyl cyclases, two families of enzymes that use a similar two-metal-ion mechanism for catalysis. analysis of a crystal structure of hTHG1 constipate to nucleotide and magnesium suggests that a preadenylylation building complex may have been captured .

Results

Overall Structure of Human THG1.

The 2.3-Å crystal social organization of hTHG1 represents a previously undetermined structure of an enzyme catalyzing 3′-5′ nucleotide additions ( Table S1 ). The hTHG1 concept used for structural studies was composed of 269 amino acids with a calculate molecular system of weights of 32 kDa. The purify protein eluted from gel excommunication chromatography with an apparent molecular system of weights of ∼165 kDa ( Fig. S1 ), reproducible with formation of a higher regulate multimer in solution and with the tetrameric form of the enzyme observed in the crystals ( Fig. 2 ). The tetrameric phase of hTHG1 appears to be highly conserved, because yeast, archaeal, and bacterial Thg1 enzymes similarly eluted from mousse exclusion with a molecular weight reproducible with a tetramer. hTHG1 was crystallized in two different crystal systems, tetragonal and rhombohedral, and in both cases the homotetramer appears as a dimer of dimers ( Fig. 2 ), in which the first dimer constitutes the crystal asymmetrical whole and the second is generated by symmetry .Fig. 2. Fig. 2. Ribbon diagram of the homo THG1 homotetramer. The tetramer consists of a dimer of dimers. Monomers are colored as follows : grey, monomer A ; chicken, monomer B ; aristocratic, monomer A′ ; and fleeceable, monomer B′. The farseeing branch composed of β-strands β6 and β7 is seen in its entirety only in the rhombohedral crystals. Disordered residues are indicated with spheres. The hTHG1 monomer is made of a β-sheet write of six antiparallel strands flanked by three or four α-helices on each side. In addition, two antiparallel β-strands ( β6 and β7 ) form a long branch that is seen only in the rhombohedral crystals. Interactions between the two monomers in the asymmetrical unit are largely mediated by residues from coil αD and β4. several hydrogen bail interactions are made between main-chain atoms of β4 and side-chain atoms of αD, an model of which is seen between G129-N and T98-OH. Alteration of T98 to alanine disrupted multimer formation, as judged by mousse exclusion chromatography ( Fig. S1 ), and yielded a discrepancy deficient in G-1 action ( table 1 ). In addition to hydrogen bonding interactions, two salt bridges ( K95A–D128B ; E13A–R130B ) stabilize the interface between the two monomers. The importance of these residues is underscored by mutational studies of Saccharomyces cerevisiae Thg1 ( ScThg1 ), with which hTHG1 shares 52 % sequence identity : change of any of these four strictly conserved residues in yeast ( ScK96, ScD131, ScE13, or ScR133 ) to alanine powerfully diminishes G-1 accession activity ( 18 ). table 1. G-1 addition activity of yeast and human Thg1 variants The N-terminal segment comprising the first 20 residues from one monomer in the A/B dimer wraps around the early monomer in such a means that residues in coil αA contact the nucleotide adhere web site of the other monomer ( see below ). The interface between the two monomers buries 4,200 Å2 ( 19 ). The intertwine N-terminal segments from the first dimer provide a chopine to interact with their symmetry-related counterparts from the second gear dimer. The dimer/dimer ( AB/A′B′ ) interface is less across-the-board than the monomer/monomer ( A/B ) interface, burying only 1,800 Å2 ( 19 ). The tetrameric form was calculated to be the most stable oligomeric form ( ΔG = -68 kcal/mol vs. -22 kcal/mol for the dimer ) ( 20 ) .

Unexpected Homology to DNA Polymerases.

Because hTHG1 shares no significant succession similarity with any know protein, it was unclear whether hTHG1 would have any geomorphologic homologs. surprisingly, significant homology was found with guanylyl and adenylyl cyclases [ βαββαβ theme of C1a domain ; Z score = 8.0 ; Protein Data Bank ( PDB ) ID codes 2W01 ( 21 ) and 3E8A ( 22 ) ] ( 23 ), along with the palm domain of several traditional polymerases including T7 DNA polymerase, a family A polymerase [ Z score = 5.3, PDB ID code 1T7P ( 24 ) ], and DNA polymerase II, a member of the B family [ Z score = 5.9, PDB ID code 1Q8I, ( 25 ) ]. The fold of the hTHG1 βαββαβ motif ( residues 22–135 ) most closely matches that of the cyclases. however, the superposition of hTHG1 with the polymerases suggests that the Thg1 mechanism is closest to that of the A family polymerases ( Fig. 3 ), based on the position of three highly conserve carboxylate residues ( see below ). [ All DNA polymerases harbor these three carboxylates in the polymerase active site ( 26, 27 ) and members of the A and B families differ in the proportional position of the three residues ( Fig. 3 A ) ]. After the hTHG1 structure was solved, Aravind and coworkers published a computational analysis proposing a model which is consistent with parts of our biochemical and structural data, including the potential interest of three carboxylates and bivalent alloy ions in catalysis ( see below ) ( 28 ) .Fig. 3. Fig. 3. The catalytic effect of hTHG1 most close resembles DNA polymerases of the A syndicate. ( A ) Comparison of topology diagrams for hTHG1, adenylate cyclase ( PDB ID code 3E8A ) ( 22 ), and handle domains of T7 DNA polymerase ( family A, PDB ID code 1T8E ) ( 42 ), and RB69 gp43 ( family B, PDB ID code 2OYQ ) ( 43 ). The count of the β-strands is for the palm knowledge domain entirely and locations of the catalytic carboxylate residues are indicated by black diamonds. ( B ) superposition of hTHG1 ( residues 22–135, blue sky ) with the palm sphere of T7 DNA polymerase ( residues 466–486, 608–698, grey ) overlays catalytic carboxylates. The incoming nucleotide and metallic ions from the T7 DNA polymerase complex are shown ( 1T7P ) ( 24 ).

Read more: David Prowse

The core β-sheet of the decoration knowledge domain of T7 DNA polymerase superimposes with the corresponding theme in hTHG1 with an rmsd of 1.8 Å. The superposition pinpointed three residues that correspond to the three catalytic carboxylates of all known DNA polymerases ( Fig. 3 B ) ( 29 ). In contrast, adenylyl and guanylyl cyclases contain lone two catalytic aspartates and a cysteine, alanine or glycine in stead of the third carboxylate ( 21, 30 ). The highly conserved hTHG1 carboxylates D29, D76, and E77 correspond to D475, D654, and E655 in T7 DNA polymerase. Alteration of any of these three residues to alanine decreased G-1 summation bodily process by hTHG1 ( table 1 ) in a manner reproducible with that expected for catalytically crucial residues. Whereas D29A and D76A hTHG1 variants exhibited > 100-fold decreased specific bodily process compared with wild-type hTHG1, E77 exhibited a more modest ( 20-fold ) decrease in activeness similar to the relatively minor effect of the E655A alteration in T7 DNA polymerase ( 24 ). interestingly, the alanine alteration at the analogous position in ScThg1 ( ScE78 ) sternly impacts ScThg1 natural process ( table 1 ), suggesting species-specific differences in the roles of the highly conserve glutamate. The superposition of the three strictly conserved Thg1 carboxylates with those of T7 and their crucial function in catalysis powerfully suggest that Thg1 by chance uses the two-metal-ion mechanism of basic 5′-3′ polymerases ( 31 – 33 ) .

Two Metal Ions in the hTHG1 Active Site.

During the three-step nucleotide addition reaction, Thg1 first binds ATP for adenylylation then GTP for the nucleotidyl remove reaction ( Fig. 1 ). We obtained 2.95-Å diffraction data from a building complex of hTHG1 with 2′-deoxy-GTP ( dGTP ) and Mg2+ ( the enzyme can use dGTP angstrom well as GTP, referee. 12 ). The resulting electron concentration maps showed clearly density for one dGTP and one triphosphate moiety ( see below ) in each monomer of the asymmetrical unit ( Fig. 4 ). The dGTP is located in what is predicted to be the active voice locate of the enzyme based on its structural homology to DNA polymerases : A superposition of the dGTP-bound hTHG1 structure with a T7 DNA polymerase complex with DNA and ddGTP overlays the two nucleotides ( Fig. S2 ). In the hTHG1 cocrystal social organization, the guanine free-base stacks against two conserved hydrophobic residues, A37 and F42. The Watson–Crick face of the basal is within hydrogen bonding distance of two protein backbone atoms : O6 and N1 contact the D47 amide and A43 carbonyl, respectively. The 3′OH of the deoxyribose makes a hydrogen bond with H34 while the font of the sugar is within van five hundred Waals touch of the atoms forming the peptide bond between F33 and H34. The β- and γ-phosphates interact with a main-chain amide via a nonbridging oxygen ( β-phosphate with H34 and γ-phosphate with N32 ), and all three phosphates coordinate at least one of the two alloy ions. The rmsd between the dGTP-bound and unliganded structures is 0.45 Å based on an alignment length of 234 amino acids, suggesting no major geomorphologic difference between the two structures. There are, however, celebrated changes in side-chain orientation course that occur around the active locate .Fig. 4. Fig. 4. structure of the dGTP-bound mannequin of hTHG1. ( A ) overall view of the dimer in the asymmetrical unit showing monomer A ( grey ) and monomer B ( yellow ) and the oblige nucleotides and metals. ( B ) Close-up of the hTHG1 nucleotide binding locate. The structure reveals a boundary dGTP with two Mg2+ ions, A and B ( empurpled spheres ). besides present in the crystal is an extra triphosphate moiety and Mg2+. A fake annealing exclude function contoured at 3σ is shown in blue. The catalytic carboxylates Asp29, Asp76, and Glu77 are shown, as are extra residues involved in nucleotide tie and hTHG1 routine. The dGTP-bound crystal structure revealed two bind Mg2+ ions associated with the nucleoside triphosphate ( Fig. 4 B ). The bearing of these ions was confirmed using an anomalous dispute Fourier map calculated with data acquired from manganese soaked crystals ( Fig. S3 ). The outdistance between the two metallic element ions is 4.3 Å. As with class A DNA polymerases, the two hTHG1 aspartates ( D29 and D76 ) coordinate the two bivalent metal ions, whereas E77 points away from the metals. Metal B contacts nonbridging oxygens of all three phosphates of dGTP and the two catalytic aspartates D29 and D76. An interaction with the main-chain oxygen of G30 completes the octahedral coordination. Metal A interacts with the two aspartates and a nonbridging oxygen of the α-phosphate .

Nucleotide-Bound hTHG1 Structure.

The three-step chemical reaction catalyzed by Thg1 requires diverse interactions with multiple nucleotide substrates, including ATP during adenylylation, GTP during nucleotidyl transportation, and the 5′-monophosphorylated and 5′-triphosphorylated ends of the transfer rna for adenylylation and pyrophosphate removal, respectively ( Fig. 1 ). therefore, the boundary dGTP visualized in the hTHG1 social organization could reflect the position of any of these nucleotide species. To examine these possibilities, we altered H34 and S75, two hTHG1 residues that contact the boundary dGTP ( Fig. 4 B ), to alanine and used single-turnover kinetic assays to individually measure rates of adenylylation ( step 1 ) or nucleotidyl transfer ( step 2 ) during G-1 addition. H34 is highly conserved ( replaced only by S, T, or K in eukarya ) and is positioned to accept a hydrogen bail from the 3′-hydroxyl of the bandaged dGTP. S75 is universally conserved in Thg1 enzymes from all domains of animation, and the serine hydroxyl is ∼4 Å away from the N7 of the guanine base. removal of a side chain from a residue that activates the 3′-hydroxyl nucleophile for catalysis is expected to substantially decrease the rate of nucleotidyl remove. however, the maximal pace constants for adenylylation ( kaden ) and nucleotidyl transfer ( kntrans ) are lone modestly decreased by alteration of H34 to alanine ( board 2 and Fig. S4 ), suggesting that H34 does not serve this character. similarly, the S75A change did not significantly affect kaden or kntrans, indicating that neither of these residues participates directly in the chemical steps for adenylylation or nucleotidyl transmit. however, a 10-fold increase in the KD, app, ATP for the adenylylation pace of the reaction is observed for hTHG1 S75A, with a relatively smaller effect on the KD, app, GTP for nucleotidyl transfer. Although the placement of this particular ligand-bound structure along the reaction coordinate is uncertain at this early stage, these results suggest that the bound dGTP in the structure does not represent the side of the entrance GTP that is incorporated into the growing polynucleotide range. rather, the results suggest that the jump dGTP may reveal the position of the ATP that is used for the activation footprint ( adenylylation ). This interpretation is far supported by the predilection of the tie down dGTP nucleotide with its triphosphate moiety coordinating two metal ions and thus poised for chemistry to occur at the α-phosphate of the restrict NTP, such as occurs during adenylylation ( Fig. 1 ). furthermore, the 3′OH of the dGTP does not contact either metallic ion and is not positioned for nucleophilic approach. mesa 2. kinetic analysis of residues interacting with the nucleotide notably, interactions observed between hTHG1 and the nucleotide base are not restricted to a guanine. The interaction of the Watson–Crick face of the basis with main-chain atoms are such that the bind web site could accommodate either a guanine or an adenine by alternating anchor carbonyl/amide contacts. The interaction between guanine N7 and S75 is besides possible with either purine. importantly, although ATP is preferred by hTHG1, GTP is able to substitute for ATP in the transfer rna activation step.

Read more: Wikipedia

A Second Active Site Triphosphate.

As described above, the electron density maps revealed the presence of a second bind atom in each monomer, a triphosphate. The presence of the triphosphate group implies that a second base nucleotide was bound ; the base and sugar moieties are not seen in the electron concentration function presumably because of the miss of specific interactions. This situation is evocative of that identify for the class I 3′-CCA-adding enzymes ( transfer rna nucleotidyl transferases ), where the basal moiety of the nucleotide is not tightly bound in the absence of the transfer rna substrate ( 34 – 36 ). The triphosphate moiety interacts with respective strictly conserved residues : the α-phosphate contacts R27 and R130, whereas the γ-phosphate contacts R130, a well as R92 and K95 of the adjacent monomer. A metallic ion, presumably Mg2+, is besides bound by the α-, β-, and γ-phosphates. The concentration was assigned to a metal ion rather than a water atom by analogy with other polymerase structures deposited in the Protein Data Bank in which the triphosphate tail constantly ligates a alloy ion. Altering any of the aforesaid positively charged residues to alanine in ScThg1 causes a marked decrease in G-1 addition bodily process, suggesting that this locate is critical for catalysis ( 18 ). Based on the data presented above, the second tie nucleotide is not probable to reflect the place of the ATP used for adenylylation. therefore, the note triphosphate may pinpoint a tie web site for the incoming nucleotide that participates in 3′-5′ addition or the transfer rna molecule ( Fig. 1 ) ; further biochemical and structural portrayal with bound transfer rna and nucleotide substrates will be needed to evaluate these possibilities .

tRNA Recognition by hTHG1.

The important question of how hTHG1 binds to substrate transfer rna is not addressed by the stream hTHG1 structure, because no know tRNA-binding structural motifs were identified. A character in tRNA-binding may help to explain the observation that several highly conserved residues ( H152, K187, E189, N198, and K208 ) are critically significant to catalysis ( 18 ) ( table 1 ), yet are located in a small coiling subdomain ( helices αF and αG ) 20–30 Å from the nucleotide bind web site ( Fig. 4 A ). In addition, previous biochemical portrayal of ScThg1 implicated a individual aspartate residue ( ScD68, analogous to hTHG1 D67 ) in tRNAHis anticodon recognition ( 18 ). however, attempts to build a model of tRNA-bound hTHG1 suggest multiple possibilities for the orientation and mode of transfer rna substrate constipate, and extra biochemical and geomorphologic data are needed to critically evaluate these electric potential alternatives. Thus the exact bind mode of transfer rna to Thg1 remains uncertain, and will have to await the crystal structure of the enzyme with its polynucleotide substrate .