Aminoacyl-tRNA synthetases are multidomain enzymes that often possess two activities to ensure translational accuracy. A synthetic active site catalyzes tRNA aminoacylation, while an editing active site hydrolyzes mischarged tRNAs. Prolyl-tRNA synthetases (ProRS) have been shown to misacylate Cys onto tRNA Pro , but lack a Cys-specific editing function. The synthetase-like Haemophilus influenzae YbaK protein was recently shown to hydrolyze misacylated Cys-tRNA Pro in trans. However, the mechanism of specific substrate selection by this single domain hydrolase is unknown. Here, we demonstrate that YbaK alone appears to lack specific tRNA recognition capabilities. Moreover, YbaK cannot compete for aminoacyl-tRNAs in the presence of elongation factor Tu, suggesting that YbaK acts before release of the aminoacyl-tRNA from the synthetase. In support of this idea, cross-linking studies reveal the formation of binary (ProRS⅐YbaK) and ternary (ProRS⅐YbaK⅐tRNA) complexes. The binding constants for the interaction between ProRS and YbaK are 550 nM and 45 nM in the absence and presence of tRNA Pro , respectively. These results suggest that the specificity of trans-editing by YbaK is ensured through formation of a novel ProRS⅐YbaK⅐tRNA complex.Aminoacyl-tRNA synthetases maintain the high fidelity of protein synthesis by activating specific amino acids and transferring them to cognate tRNAs in a process known as aminoacylation. However, some synthetases are known to misactivate noncognate amino acids, transferring them onto their cognate tRNAs. Therefore, to ensure accurate translation of the genetic code, these enzymes, which include members of both classes of synthetases, have evolved proofreading or editing functions (1-17). Pretransfer editing involves hydrolysis of the misactivated aminoacyl-adenylate, whereas post-transfer editing refers to specific hydrolysis of misacylated tRNA.To clear noncognate amino acids, a double-sieve mechanism was proposed, wherein larger amino acids are rejected by the aminoacylation active site (i.e. coarse sieve) while smaller amino acids are cleared in a second editing active site (i.e. fine sieve) (18,19). This hypothesis was confirmed by more recent biochemical and x-ray crystallographic studies (20). Although the exact site of pretransfer editing is still an open question in most systems (21), class I editing enzymes, including isoleucyl-tRNA synthetase, valyl-tRNA synthetase, and leucyl-tRNA synthetase, use the highly conserved connective polypeptide 1 domain to edit misacylated tRNAs (1, 2, 4 -8, 12). In contrast, the post-transfer editing domains used by class II synthetases are much more diverse. The internal editing domain of alanyl-tRNA synthetase (AlaRS) 2 and the N-terminal editing domain of threonyl-tRNA synthetase (with the exception of the archaebacterial enzymes) share sequence similarity (9, 22), while the editing domains of bacterial prolyl-tRNA synthetase (ProRS) (13) and phenylalanyl-tRNA synthetase (17) appear to be unique, sharing no sequence homology to any of the ot...