The Pediococcus pentosaceus glutamate racemase gene product complemented the D-glutamate auxotrophy of Escherichia coli WM335. Amino acid sequence analysis of the two proteins revealed 28% identity, primarily in six clusters scattered throughout the sequence. Further analyses indicated secondary structure similarities between the two proteins. These data support a recent report that the dga (murl) gene product is a glutamate racemase.The peptidoglycan component of the bacterial cell wall is synthesized by a number of unique enzymes (10). Several of these enzymes either synthesize D-amino acids or add them to the monomeric subunit precursor in the cytoplasm. One of these enzymatic activities, D-glutamic acid biosynthesis, has not been characterized in most species of bacteria, including Escherichia coli. Early studies with a few species revealed two different biosynthetic routes of D-glutamic acid biosynthesis. In Bacillus spp., D-glutamate is formed from D-alanine and a-ketoglutarate by a D-amino acid transaminase (11,18,20,21), while Lactobacillus spp. (7,8,19) and Pediococcus spp. use a racemase which converts L-glutamate to D-glutamate (1, 2, 13, 14). All prior attempts to detect either of these activities in E. coli had been unsuccessful (5,13,14). In addition, the gene(s) responsible for D-glutamate biosynthesis in this organism also had not been identified until recently (5, 6). The dga (murl) gene was identified by its ability to complement the D-glutamate auxotrophic phenotype of the only known Dglutamic-acid-requiring mutant, E. coli B/r WM335 (9, 12), and could be stably maintained on high-copy-number pUC vectors (5, 6). Doublet et al. recently reported data that suggest that dga (murI) encodes a racemase enzyme in E. coli (4). This finding is supported by our own experimental data and protein sequence analysis.Glutamate racemase complements an E. coli D-glutamate auxotroph. Nakajima et al. described the cloning of the glutamate racemase gene of Pediococcus pentosaceus and its expression in E. coli (13). Plasmid pICR222, which consists of a 1.65-kb HindIII-EcoRI fragment containing the racemase gene cloned into pUC18 (Fig. 1) Fig. 1]). Subsequently, 110 bp was removed from the ORF on pRDD022 via a DraII deletion (recombinant plasmid designated pRDD023 [ Fig. 1]). Both pRDD022 and pRDD023 were transformed into WM335 and checked for their ability to complement the D-glutamic acid-requiring phenotype. Plasmid pRDD022 was able to complement the phenotype, while plasmid pRDD023 could not (data not shown). These data provide strong evidence that the 795-bp ORF encodes an activity which allows WM335 to grow normally in the absence of exogenous Dglutamic acid. Since pUC19 is present in E. coli cells in high copy numbers (>100 copies), the ORF was subcloned into a low-copy-number vector, pWSK29 (approximately five copies per cell [22]), to determine whether a relatively low gene dosage would still complement WM335. The recombinant plasmid, designated pRDD024 (Fig. 1)