A tRNA-independent Mechanism for Transamidosome Assembly Promotes Aminoacyl-tRNA Transamidation*

Background: Some microorganisms use indirect tRNA aminoacylation to produce Asn-tRNAAsn; the necessary components are assembled into a tRNAAsn-dependent transamidosome complex. Results: A new protein, Hp0100, facilitates formation of an alternative, tRNA-independent transamidosome and increases the efficiency of Asp-tRNAAsn transamidation. Conclusion: Hp0100 is a component of a stable efficient Helicobacter pylori transamidosome. Significance: The Hp0100-containing transamidosome allows for optimal indirect biosynthesis of Asn-tRNAAsn. Many bacteria lack genes encoding asparaginyl- and/or glutaminyl-tRNA synthetase and consequently rely on an indirect path for the synthesis of both Asn-tRNAAsn and Gln-tRNAGln. In some bacteria such as Thermus thermophilus, efficient delivery of misacylated tRNA to the downstream amidotransferase (AdT) is ensured by formation of a stable, tRNA-dependent macromolecular complex called the Asn-transamidosome. This complex enables direct delivery of Asp-tRNAAsn from the non-discriminating aspartyl-tRNA synthetase to AdT, where it is converted into Asn-tRNAAsn. Previous characterization of the analogous Helicobacter pylori Asn-transamidosome revealed that it is dynamic and cannot be stably isolated, suggesting the possibility of an alternative mechanism to facilitate assembly of a stable complex. We have identified a novel protein partner called Hp0100 as a component of a stable, tRNA-independent H. pylori Asn-transamidosome; this complex contains a non-discriminating aspartyl-tRNA synthetase, AdT, and Hp0100 but does not require tRNAAsn for assembly. Hp0100 also enhances the capacity of AdT to convert Asp-tRNAAsn into Asn-tRNAAsn by ∼35-fold. Our results demonstrate that bacteria have adopted multiple divergent methods for transamidosome assembly and function.

Protein synthesis proceeds with high fidelity to ensure optimal survival. This process requires a full set of aminoacyl-tRNAs that are accurately aminoacylated so that the identity of the tRNA anticodon corresponds to the amino acid attached to its acceptor stem. Many microorganisms are missing one or more aminoacyl-tRNA synthetases, the enzymes that typically aminoacylate each tRNA. The obligate human pathogen Helicobacter pylori does not have either glutaminyl-or asparaginyl-tRNA synthetase (1,2). To compensate for the missing aminoacyl-tRNA synthetases, H. pylori utilizes an indirect pathway to produce Gln-tRNA Gln and Asn-tRNA Asn (3)(4)(5). The first step in the synthesis of Asn-tRNA Asn relies on a bacterial-type, nondiscriminating aspartyl-tRNA synthetase (ND-AspRS), 2 which aminoacylates tRNA Asp to generate Asp-tRNA Asp and misacylates tRNA Asn to produce Asp-tRNA Asn (6). Next, Asp-tRNA Asn is converted to Asn-tRNA Asn by a heterotrimeric, glutamine-dependent amidotransferase called AdT (or GatCAB). A similar pathway exists in H. pylori for Gln-tRNA Gln synthesis, where misacylated Glu-tRNA Gln is produced by a tRNA Gln -specific glutamyl-tRNA synthetase (GluRS2) (1,7). The same AdT converts Glu-tRNA Gln into Gln-tRNA Gln .
H. pylori and other organisms that utilize indirect tRNA aminoacylation must rely on mechanisms to prevent their misacylated tRNA intermediates from entering the ribosome, where they would cause errors in translation. Elongation factor Tu (EF-Tu) provides one such mechanism (8 -10). In Thermus thermophilus, a thermophilic bacterium, formation of a stable complex called the Asn-transamidosome also promotes accuracy. This macromolecular complex requires tRNA Asn for assembly with AdT and AspRS2 (an archaeal-type ND-AspRS) (11,12). The Asn-transamidosome ensures the stability of the aminoacyl ester bond of Asp-tRNA Asn , promotes its efficient conversion to Asn-tRNA Asn , and isolates it from EF-Tu and the ribosome until it is converted to Asn-tRNA Asn . The crystal structure of the T. thermophilus Asn-transamidosome shows that it contains two tRNAs in scaffold positions and two positioned to act as substrates for both ND-AspRS and AdT (12). Thus, this transamidosome is a tRNA-containing ribonucleoprotein complex. Similarly, the thermophilic archaeon Methanothermobacter thermoautotrophicus assembles a Gln-transamidosome from ND-GluRS and GatDE. (GatDE is a heterodimeric homolog of AdT that is specific for Gln-tRNA Gln biosynthesis (13).) Assembly of this Gln-transamidosome does not require tRNA Gln , but it is also less stable and more dynamic than the T. thermophilus Asn-transamidosome. Recently, pre-steady-state kinetic experiments and unsuccessful efforts to isolate the M. thermoautotrophicus Glntransamidosome argue that Gln-transamidosome formation and substrate channeling are not essential or even favorable, at least in archaea (14).
Unlike T. thermophilus and M. thermoautotrophicus, many bacteria, including H. pylori, use the heterotrimeric AdT (Gat-CAB) to compensate for the absence of both asparaginyl-and glutaminyl-tRNA synthetases; it is unknown if or how this dual specificity impacts Asn-and/or Gln-transamidosome formation and function. We recently examined the formation of Asnand Gln-transamidosomes from H. pylori components (e.g. ND-AspRS, tRNA Asn , and AdT or GluRS2, tRNA Gln , and AdT, respectively). Stable assembly of a macromolecular transamidosome complex was not observed in either case. However, quantitative steady-state kinetic analyses argued for transient dynamic assembly of both transamidosome particles (15,16).
These analyses left open the intriguing possibility that an additional mechanism to prevent translational errors remains undiscovered.
Here, we examined the hypothesis that H. pylori uses other proteins or enzymes either to facilitate transamidosome assembly or to stabilize and deliver each aminoacyl-tRNA from the misacylating aminoacyl-tRNA synthetase to AdT. A yeast twohybrid interaction profile of the H. pylori proteome (17) was used to identify a protein of unknown function called Hp0100. Hp0100 was selected for further studies based on its weak reported interactions with GatA (a subunit of AdT) and its stronger interactions with ND-AspRS. Here, we demonstrate that Hp0100 is a component of a new, stable, tRNA-independent transamidosome enzyme complex consisting of ND-AspRS, AdT, and Hp0100. The addition of Hp0100 also significantly accelerates the AdT-catalyzed rate of Asp-tRNA Asn transamidation. Consequently, Hp0100 simplifies the process of indirect tRNA aminoacylation by bringing all enzymatic players into proximity in a distinct Asn-transamidosome.

Overexpression and Purification of H. pylori ND-AspRS and
AdT-H. pylori ND-AspRS and AdT were overexpressed in Escherichia coli and purified as described previously (6,18). To avoid possible contamination by E. coli tRNA, cell lysates were treated with USB RNase A (6.25 g/ml; Affymetrix Inc.) prior to affinity purification.
In Vivo Production and Purification of H. pylori tRNA Asp and tRNA Asn -H. pylori tRNA Asp and tRNA Asn were produced in vivo in E. coli MV1184. They were purified, and their concen-trations were determined as described previously (6). Each tRNA was diluted to a stock solution of 150 M prior to use.
Cloning, Overexpression, and Purification of Hp0100-The hp0100 gene was amplified from H. pylori strain 26695 genomic DNA (American Type Culture Collection) and cloned into pQE-80L (Novagen) to encode the full-length protein with an N-terminal His 6 tag (pPTC034). E. coli DH5␣ calcium chloride competent cells were transformed with the pPTC034 plasmid. Cultures were grown, beginning with single colonies, at 37°C in LB medium. Protein expression was induced with isopropyl ␤-D-thiogalactopyranoside (1 mM) when the absorbance at 600 nm reached 0.6 -1.0. After 1 h, cells were harvested by centrifugation at 5000 rpm for 5 min. Hp0100 was first purified using nickel-nitrilotriacetic acid spin columns (Qiagen) following the manufacturer's instructions. Purification of Hp0100 resulted in co-purification of a contaminant that was the same approximate size as E. coli discriminating AspRS (D-AspRS). The identity of this contaminant as D-AspRS was demonstrated by aminoacylation assays with tRNA Asp (supplemental Fig. S1).
To avoid E. coli D-AspRS contamination, Hp0100 was purified by ion-exchange chromatography instead of nickel affinity. Harvested cells were homogenized in 3 ml of lysis buffer (20 mM Na 2 PO 4 , 5 mM ␤-mercaptoethanol, 6.25 g/ml RNase A, 15 l/ml saturated PMSF, and 1 mg/ml lysozyme at pH 7.4) and incubated on ice for 30 min. The lysate was then sonicated six times at 38% for 10 s on a Branson Digital Sonifier with 2-min recovery intervals on ice between each pulse. Cell debris was removed by centrifugation at 14,000 rpm for 1 h at 4°C. The filtered lysate (1 ml) was loaded onto a UNO S ion-exchange column (12-ml column volume; Bio-Rad). Hp0100 eluted at ϳ400 mM NaCl. After ion-exchange purification, Hp0100 did not contain detectable levels of contaminating E. coli D-AspRS (supplemental Fig. S1A) or tRNA (supplemental Fig. S2). The concentration of Hp0100 was determined by UV-visible spectroscopy (280 nm) using an extinction coefficient of ⑀ ϭ 30620 M Ϫ1 cm Ϫ1 as determined by the ExPASy Proteomics server (19). ND-AspRS Aminoacylation Assays-Aminoacylation assays were performed with tRNA Asp and tRNA Asn as described previously (6). For titration experiments, each tRNA concentration was 10 M, and ND-AspRS was used at 0.2 M. The concentration of Hp0100 was varied from 0.2 to 4 M as indicated.
Preparation of 32 P-Labeled Asp-tRNA Asn -The CCA-adding enzyme was used to incorporate [␣-32 P]AMP into the 3Ј-end of H. pylori tRNA Asp and tRNA Asn by replacing the unlabeled 3Ј-AMP with [␣-32 P]AMP (from [␣-32 P]ATP; American Radiolabeled Chemicals) as described previously (4,20,21) The plasmid expressing the CCA-adding enzyme was generously provided by Dr. Rebecca Alexander (Wake Forest University).
P1 Nuclease AdT Transamidation Assay-Transamidation activity was measured essentially as described previously (4,20,21). The assay buffer contained 5 M 32 P-labeled Asp-tRNA Asn (specific activity of ϳ4.5 Ci/mmol), 50 mM HEPES-OH (pH 7.5), 4 mM ATP, 8 mM MgCl 2 , 25 mM KCl, and 1 mM Gln with 10 nM AdT. Where indicated, Hp0100 was preincubated with AdT for 10 min at 4°C at final concentrations that ranged from 25 to 500 nM. Aliquots (5 l) were quenched at each time point, digested with P1 nuclease, and processed as described previ-ously (4,20,21). Reaction progress was monitored by developing each time point on a 20 ϫ 20 cm PEI-cellulose TLC plate (EMD Millipore) in a freshly prepared 100-ml solution of 10 mM ammonium chloride and 5% glacial acetic acid. For better resolution, the TLC plates were pretreated with water for 15 min and then dried prior to use. Plates were phosphorimaged after ϳ16 h of screen exposure. Asn-tRNA Asn production was normalized against Asp-tRNA Asn to eliminate errors arising from differences in sample spotting. For this reason, Asn-tRNA Asn production is reported as a percentage relative to Asp-tRNA Asn .
Electrophoretic Mobility Shift Assays-Macromolecular complex formation between AdT, ND-AspRS, and Hp0100 was examined by electrophoretic mobility shift assay. AdT (0.5 M), ND-AspRS (0.5 M), Hp0100 (2.5 M), and tRNA Asn (5 M) were preincubated for 30 min at 4°C in the indicated combinations. Complexes were examined on an 8% native polyacrylamide gel (330 ϫ 430 ϫ 1.5 mm 3 ) in Tris/glycine buffer (25 mM Tris and 200 mM glycine at pH 8.3). After electrophoresis at 20 V for 16 h at 4°C, bands were visualized by Western blotting using anti-His 6 antibody (AnaSpec) according to the manufacturer's instructions. Hp0100 and ND-AspRS both contain single N-terminal His 6 tags. However, His 6 -Hp0100 was not detectable with anti-His 6 antibody under native gel conditions (supplemental Fig. S3). His 6 -Hp0100 was readily visualized by Western blotting under denaturing conditions (data not shown). For AdT, both the GatC and GatB subunits contain His 6 tags.
Dynamic Light Scattering-Dynamic light scattering (DLS) analyses of ND-AspRS, AdT, Hp0100, and various binary and ternary complexes were performed at 25°C in a Zetasizer Nano S instrument (Malvern Instruments Ltd., Worcestershire, United Kingdom). Protein samples were diluted in buffer containing 50 mM NaH 2 PO 4 and 300 mM NaCl (refractive index ϭ 1.34) to prepare the required concentrations; 40-l samples were used for analyses. For particles that are smaller than the wavelength of incident light (), the intensity of light scattered (I) is related to the diameter of the particle (d) by the Rayleigh equation (I∞d 6 / 4 ). We used the Mie theory within the Nano S software to convert the intensity (I) versus size (d in nanometers) distribution data to volume versus size (d) distribution (supplemental Fig. S4). Mie theory provides a more realistic view of the importance of the observed peak/s. Surface Plasmon Resonance-Surface plasmon resonance (SPR) was used to characterize the binary interactions between Hp0100 and each of its possible binding partners (ND-AspRS, AdT, and tRNA Asn ). Hp0100 was immobilized on a CM5 sensor chip, and increasing concentrations of analyte (either protein or tRNA) were delivered to the chip. Dissociation constants (K d ) for each pair of interactions were determined either by fitting the SPR data directly to a Langmuir model using BIAevaluation version 4.0.1 or by plotting the equilibrium response units versus analyte concentration and fitting the resulting curve to the following equation: H-A ϭ [A 0 ]⅐R max /(K d ϩ [A 0 ]) (using KaleidaGraph version 4.0), where H-A represents the complex between Hp0100 (H) and the analyte (A) and is given in SPR equilibrium response units, [A 0 ] is the initial concentration of the analyte in solution, and R max represents the theoret-ical response units that would be observed upon 100% complex formation.

Published Yeast Two-hybrid Results Suggest Biologically Significant Interactions between AdT, ND-AspRS, and Hp0100, a
Protein of Unknown Function-In contrast to the stable, tRNAdependent Asn-transamidosome from T. thermophilus, an analogous H. pylori Asn-transamidosome could not be isolated (15). We investigated the possibility that a protein of unknown function is needed to promote assembly of a more stable complex. By mining published results of a protein-protein interaction map of the H. pylori proteome (17), we identified Hp0100 as a possible candidate because it showed yeast two-hybrid interactions with both AdT and ND-AspRS, the two enzymes in the T. thermophilus Asn-transamidosome. Hp0100 is a 368-amino acid protein that is conserved throughout the ⑀-proteobacteria; many bacterial genomes outside of this clade contain genes encoding the N-terminal half of Hp0100. The primary sequence of Hp0100 is not related to any protein of known function. However, sequence analyses suggest that its N terminus contains an ATP-binding motif that belongs to the adenine nucleotide ␣-hydrolase superfamily (AANH-like superfamily) (22).
E. coli D-AspRS Is Purified with His 6 -Hp0100-Building on the observation that Hp0100 interacts with both ND-AspRS and AdT by yeast two-hybrid screening, we hypothesized that this protein is a component of a novel bacterial-type transamidosome. We cloned the hp0100 gene and expressed Hp0100 in E. coli with an N-terminal His 6 tag. Nickel affinity chromatography led to the purification of Hp0100 with a single co-purifying contaminant that was approximately the same size as E. coli discriminating D-AspRS (supplemental Fig. S1A). Aminoacylation assays with tRNA Asp confirmed that the mixture of Hp0100 and this protein contaminant produced Asp-tRNA Asp , supporting our presumption that this contaminant is E. coli D-AspRS (supplemental Fig. S1B). These results offered the first in vivo and in vitro evidence for a relevant noncovalent interaction between Hp0100 and AspRS. (E. coli does not have a full-length ortholog of Hp0100.) Hp0100 was subsequently purified by ion-exchange chromatography to Ͼ95% purity with complete removal of the contaminating D-AspRS (supplemental Fig. S1A) (4,20,21).
Hp0100, AdT, and ND-AspRS Assemble into Binary and Ternary Complexes in the Absence of tRNA Asn -H. pylori ND-AspRS, tRNA Asn , and AdT do not assemble into a stable T. thermophilus-like Asn-transamidosome in gel-shift experiments (Fig. 1C, lane 5) (15). We recently showed that the H. pylori Asn-transamidosome is dynamic and unstable (15). We proposed a kinetic mechanism for translational fidelity wherein Asp-tRNA Asn would not dissociate from ND-AspRS prior to uptake by AdT (15). This mechanism was consistent with the data available at that time, but it was dissatisfyingly complex.
Alternatively, another unknown protein factor such as Hp0100 could induce stable transamidosome assembly, offering additional protection above that provided by a kinetic mechanism. To accurately assess the role of Hp0100 in the absence of tRNA, all enzymes used in experiments herein were pretreated with RNase A. RNA gel evaluation in comparison with tRNA standards confirmed that each enzyme contained Ͻ0.1% tRNA contamination (supplemental Fig. S2). Where indicated, tRNA Asn was added. To determine whether or not Hp0100 participates in stable Asn-transamidosome formation, we first sought to recapitulate the observed yeast two-hybrid interactions between Hp0100 and AdT and between Hp0100 and ND-AspRS. Native gel electrophoresis showed upward shifts when Hp0100 was combined with ND-AspRS (Fig. 1A) and with AdT (Fig. 1B), demonstrating that both of these indirect tRNA aminoacylation components separately interact with Hp0100, as was suggested by yeast two-hybrid screening. The extent of complex formation appears to be nearly complete in both cases. (Full blots are shown in supplemental Fig. S3.) We also used native gel electrophoresis to examine the contribution of Hp0100 to H. pylori transamidosome assembly. The addition of Hp0100 induced the formation of a heterotrimeric complex containing ND-AspRS, AdT, and Hp0100 (Fig.  1C, lane 4), and unlike the T. thermophilus transamidosome (11,12), this complex required Hp0100 instead of tRNA Asn (Fig.  1C, compare lanes 4 and 5). The native gels shown in Fig. 1 demonstrate that tRNA-independent complexes formed between Hp0100 and AdT, between Hp0100 and ND-AspRS, and between Hp0100, ND-AspRS, and AdT; these complexes were stable under native gel electrophoresis conditions.
We also observed these complexes in solution using DLS (Fig. 2). Hp0100 alone has a propensity to aggregate (observed by DLS (supplemental Fig. S4G) and size-exclusion chromatography (data not shown)); reproducible characterization of an Hp0100 particle was not possible. Attempts to observe interactions between ND-AspRS and AdT (in the absence of Hp0100) were unsuccessful. These negative results were expected based on the data from Fig. 1 that show that ND-AspRS and AdT did not assemble into a complex in the absence of Hp0100 and our previous efforts to isolate an H. pylori Asn-transamidosome (15). In all other cases, single predominant particles were observed (Ͼ90% of the total peak volume after volume size distribution data processing) and characterized. (All DLS spectra are provided in supplemental Fig. S4.) By DLS, ND-AspRS alone had a hydrodynamic diameter of 9.7 Ϯ 0.2 nm. This diameter increased to 11.5 Ϯ 0.3 nm upon the addition of equimolar Hp0100, indicating binary complex formation. AdT alone had an observed hydrodynamic diameter of 10.6 Ϯ 0.2 nm, which increased negligibly to 11.0 Ϯ 0.3 nm upon the addition of Hp0100. It is unclear why this shift in particle size was statistically insignificant. The particle diameters determined by DLS represent the diameters of a perfect hard sphere. It is possible that the Hp0100-AdT complex is more tubular and, consequently, less amenable to characterization by DLS or that assembly of this complex could result in compaction such that the observed diameter is similar to AdT alone. (The solution dimer between Hp0100 and AdT was further confirmed by SPR (Fig. 3B; see below).) Finally, the ternary complex of Hp0100, ND-AspRS, and AdT showed a diameter of 13.2 Ϯ 0.2 nm, a significant increase from the diameters of either dimeric complex. These results were insufficient to

FIGURE 2. Hp0100, AdT, and ND-AspRS assemble into an Asn-transamidosome in solution.
DLS was used to quantify changes in particle size with different binary and ternary combinations of Hp0100, AdT, and ND-AspRS as indicated. Results are reported only from spectra that showed a single predominant particle. Error bars represent standard (Std.) deviation from triplicate measurements. Molecular weights (Mol. Wt.) are given for each monomeric enzyme. The molecular weight for AdT is based on one copy of the heterotrimeric GatCAB enzyme. The molecular weights calculated for each complex are also based on 1:1 stoichiometries. Exact stoichiometries are given when evident. ND, not determinable from these DLS results. All spectra are provided in supplemental Fig. S4. determine complex stoichiometry, but they do confirm that Hp0100 promotes the formation of a stable transamidosomelike complex in solution.
Hp0100 Has Low Micromolar Affinities for Both ND-AspRS and AdT-SPR was used to quantitatively assess binary interactions between Hp0100 and ND-AspRS, between Hp0100 and AdT, and between Hp0100 and tRNA Asn (Fig. 3). Both AdT and ND-AspRS separately interacted with Hp0100 with low micromolar K d values (1.6 Ϯ 0.3 and 1.3 Ϯ 0.2 M, respectively) (Fig.  3). These dissociation constants are indistinguishable within error, consistent with assembly of a ternary complex. Weak interactions between Hp0100 and tRNA Asn were observed (K d Ͼ 10 M) (supplemental Fig. S5).
Hp0100 Does Not Affect ND-AspRS Activity-Next, we evaluated the impact of Hp0100 on ND-AspRS activity in the absence and presence of AdT. ND-AspRS aminoacylation assays were performed in the presence of increasing concentrations of Hp0100 (0 -4 M) without AdT; all substrates were at saturating concentrations, so the observed rates approximate k cat . Initial rates of aminoacylation were examined using both tRNA Asp and tRNA Asn to determine whether any impact could be observed and, if so, at what Hp0100 concentration this effect would reach saturation. Our results show that the addition of Hp0100 alone did not affect the ND-AspRS-catalyzed initial rates of aminoacylation of either tRNA even at concentrations of Hp0100 above the K d (supplemental Fig. S5).
To evaluate the impact of the ternary complex on ND-AspRS activity, Hp0100 and AdT were added to ND-AspRS at saturating concentrations (2 and 10 M, respectively), and ND-AspRS activity was measured (supplemental Fig. S6). Under these conditions, the addition of AdT and Hp0100 did not increase the rate of tRNA Asn aminoacylation. In fact, under these conditions, the addition of AdT alone or with Hp0100 caused slight drops in this rate, presumably because the excess AdT competed for tRNA Asn .
Hp0100 Increases the Rate of AdT-catalyzed Transamidation of Asp-tRNA Asn -We also quantitatively examined the impact of Hp0100 on the initial rate of transamidation of Asp-tRNA Asn by AdT (Fig. 4). Results from representative transamidation assays (Fig. 4, A and B) show that Hp0100 accelerated the rate of conversion of Asp-tRNA Asn to Asn-tRNA Asn . This rate enhancement reached saturation at ϳ400 nM Hp0100 (Fig. 4C), with an ϳ35-fold maximal increase in the rate of AdT catalysis. Because Hp0100 and AdT formed a stable complex at low micromolar concentrations (see Fig. 4C), one can hypothesize that these titration data reflect the uptake of Asp-tRNA Asn into the Hp0100-AdT complex. Thus, these data allowed us to approximate a K d of ϳ100 nM for Asp-tRNA Asn and the Hp0100-AdT complex. This value likely represents an upper limit because competition for Asp-tRNA Asn from free Hp0100 cannot be excluded.

DISCUSSION
The thermophilic bacterium T. thermophilus utilizes a tRNA Asn -dependent Asn-transamidosome to sequester Asp-tRNA Asn and deliver it directly to AdT, where it is converted to Asn-tRNA Asn . This transamidosome assembles from an archaeal-type ND-AspRS, tRNA Asn , and AdT; two of the four tRNAs bound to this complex are structural, making this complex a tRNA-containing ribonucleoprotein complex (12). Here, we have demonstrated that the H. pylori Asn-transamidosome is not a tRNA-containing ribonucleoprotein complex because the catalytically competent complex assembles in the absence of tRNA Asn , relegating this tRNA to the role of substrate only.
In looking at the H. pylori Asn-transamidosome, it is important to consider the differences between the archaeal-and bacterial-type ND-AspRS orthologs. The T. thermophilus Asn-transamidosome contains an archaeal-type ND-AspRS; H. pylori has a bacterial-type ND-AspRS with an extra insertion domain positioned where it could interfere with Asn-transamidosome assembly, at least based on the structure of the T. thermophilus transamidosome (12,15). Consistently, our previous characterization of the putative H. pylori Asn-transamidosome (prior to our discovery of Hp0100) revealed a dynamic complex that could not be stably isolated (15). It is intriguing to consider that this unstable association between ND-AspRS and AdT may be a result of the bacterial ND-AspRS insertion domain causing steric interference. Because H. pylori ND-AspRS, AdT, and tRNA Asn did not assemble into a stable Asn-transamidosome, we originally proposed a kinetic model for sequestration of Asp-tRNA Asn by ND-AspRS until delivery to AdT via transient association (15). This model fit the data at hand and likely contributes to translational accuracy to some extent. However, our discovery of Hp0100 enabled revision and simplification of this model (Fig.  5). By adding Hp0100 to a tRNA-independent Asn-transamidosome, it is clear that this complex exists in two forms. The first, which we call the "apo-transamidosome," contains ND-AspRS, Hp0100, and AdT. The second, "holo-transamidosome," arises upon tRNA Asn binding; AdT activity is enhanced within this complex. The formation of these complexes minimizes the need for a kinetic Asp-tRNA Asn retention mechanism. Apo-transamidosome formation does not require tRNA Asn , further distinguishing this complex from the T. thermophilus Asn-transamidosome.
Other tRNA aminoacylation systems offer precedence for the use of a bridging protein in complex assembly. For example, in yeast, Arc1p enables the assembly of a complex between Arc1p, GluRS, and methionyl-tRNA synthetase (23,24), and in Metazoa, at least nine aminoacyl-tRNA synthetases are assembled into a multi-aminoacyl-tRNA synthetase (MARS) complex with three other proteins that are essential for MARS assembly (25,26). To date, we have no evidence for formation of an Hp0100-driven MARS-like complex. GluRS2 is the most likely candidate for assembly into this kind of macromolecular complex because it produces Glu-tRNA Gln , which is subsequently converted to Gln-tRNA Gln by AdT. GluRS2 only weakly interacts with AdT (K d ϳ 40 M) (16), and it does not appear to interact with Hp0100. 3 Our discovery of the Hp0100-dependent Asn-transamidosome does not explain how H. pylori manages simultaneous faithful transamidation of both Asp-tRNA Asn and Glu-tRNA Gln . The solution to this challenge is likely to be one of two possibilities. Either AdT is expressed at sufficiently high levels to enable Asn-transamidosome formation with excess unbound AdT remaining available for Glu-tRNA Gln transamidation, or other as yet unidentified protein partners bring GluRS2 into this Asntransamidosome, making it a MARS-like complex that contains ND-AspRS, GluRS2, and AdT. Echoes of this putative MARSlike complex can be found within the published yeast two-hy-   shown. This cycle is subdivided into five steps as follows.
Step 2, tRNA Asn is recruited to this complex to form a holotransamidosome.
Step 3, ND-AspRS catalyzes the aspartylation of tRNA Asn to produce Asp-tRNA Asn .
Step 4, Asp-tRNA Asn is converted to Asn-tRNA Asn by AdT with ϳ35-fold acceleration in k cat .
Step 5, Asn-tRNA Asn is released from the complex, and the apo-transamidosome is regenerated. brid protein-protein interact map for H. pylori (15), but experimental evidence to distinguish between these two possibilities has remained elusive.
Because full-length Hp0100 is unique to the ⑀-proteobacteria, its role in transamidosome function and assembly is also likely to be limited to this bacterial clade. This scenario presents two intriguing possibilities for future consideration. First, one can imagine targeting assembly of the apo-transamidosome for the development of clade-specific antibiotics. Second, given that the ⑀-proteobacteria and T. thermophilus use different mechanisms for Asn-transamidosome assembly and function, it seems likely that other mechanisms for this critical process remain undiscovered. This mechanistic divergence suggests the exciting possibility that different branches of the tree of life elected different unique strategies for optimal indirect tRNA aminoacylation, sequestration of misacylated tRNAs, and translational accuracy.