We were able to show that two lysine-independent mutants of Streptococcus faecalis ATCC 8043 contained the enzymes for the usual bacterial pathway for lysine biosynthesis. Because of this synthetic capacity, one mutant, the Lys+OHLyss strain, could not grow in the presence of hydroxylysine without a lysine supplement. Both lysine and hydroxylysine inhibited the first enzyme of the pathway, aspartokinase. Unlike the Escherichia coli enzyme, S.faecalis dihydrodipicolinic acid synthetase was not inhibited by either lysine or hydroxylysine. Both amino acids caused the repression of dihydrodipicolinic acid synthetase and diaminopimelic acid decarboxylase. Failure of Lys+OHLyss strain to grow in hydroxylysine-supplemented medium was caused by the mimicking of lysine control by hydroxylysine. Because hydroxylysine could not completely substitute for lysine and lysine could not be synthesized, the organism did not grow. We tested three lysine analogues and found that they prevented lysine-depletion lysis in the Lsy-OHLyss strain, as did hydroxylysine. Each analogue seemed to support cell wall mucopeptide synthesis, although ornithine did not. Preliminary data indicated that these analogues like hydroxylysine, have growth-inhibitory action on the Lys+OHLyss strain, but not the Lys+OHLysr strain. The nature of the specificity of the lysine-adding enzyme for cell wall mucopeptide synthesis is discussed. MATERIALS AND METHODS Compounds. L-Aspartic ,B-semialdehyde was prepared with the method used by Black and Wright (2). Diaminopimelic acid (DAP; a mixture of the LL and meso isomers) and a-aminoadipic acid-6-14C were obtained from Calbiochem (Los Angeles, Calif.). Aspartate-4-'4C, pyruvate-2-'4C,uniformly labeled "4C-L-lysine, and 5-hydroxylysine-6-'4C (mixture of four isomers) were obtained from the New England Nuclear Corp. (Boston, Mass.). a, e-DAP (labeled with tritium) was obtained from the Nuclear-Chicago Corp. (Des Plaines, Ill.). The lysine analogues 4, 5-trans-dehydrolysine, 2, 6-diamino-4-hexynoic acid, and L-thiosine [S-(,3-aminoethyl)-L-cysteine] were the generous gift of William Shive, Department of Chemistry, University of Texas. Cultures anid media. S. faecalis 8043 and two variants of this strain whose isolation was previously described (17) were used. The strains were designated 856