icdB mutations map at 16 min, lead to the specific loss of citrate synthase, and are complemented by a prophage containing a gltA ؉ gene. Thus, they are allelic with gltA.The structural genes for all the known citric acid cycle enzymes of Escherichia coli except for a recently discovered second aconitase (AcnB) and a possible second citrate synthase have been mapped, cloned, and sequenced (3,4,13). Recently a new set of mutants altered in a gene affecting the proximal portion of the citric acid cycle were reported (12). The mutants required glutamate or 2-oxoglutarate, suggesting that the block prevented synthesis of 2-oxoglutarate. Because citrate or isocitrate could not substitute for 2-oxoglutarate or glutamate, the authors inferred that the mutant block was in isocitrate dehydrogenase and provisionally designated the corresponding gene icdB. Preliminary mapping studies suggested that icdB and a nearby transposon insertion, zbg-3::Tn10 (initially designated zgg-5483::Tn10 [12]), were at about 66 min on the chromosome (12), a region to which no citric acid cycle gene had been mapped. The structural genes for the first three enzymes in the cycle, gltA (citrate synthase), acnA (aconitase), and icdA (isocitrate dehydrogenase), are located at 16.3, 28.3, and 25.6 min, respectively (1, 14). Thus, the results suggested that icdB might be a regulatory gene.The inability of citrate or isocitrate to substitute for the glutamate requirement is not in itself indicative of a block in isocitrate dehydrogenase, because E. coli is not very permeable to citrate or isocitrate and these compounds do not satisfy the glutamate requirement of icdA mutants (6). In order to determine the nature of the icdB gene, I measured the activities of the relevant enzymes and mapped the mutations (11). Direct assay showed that the mutants were proficient for aconitase and isocitrate dehydrogenase activities but devoid of citrate synthase activity. As expected, transductants of a wild-type strain (RH822 [5]) selected for tetracycline resistance (from a donor containing both the transposon and an icdB mutation) and exhibiting a glutamate requirement (icdB) were deficient in citrate synthase activity. However, the mapping experiments showed no evidence for linkage to genes in the region of 66 min. I compared the restriction map of a cloned sequence close to icdB (in plasmid pSE-1001 [12]) with the restriction map of the entire E. coli chromosome (10). The two maps were congruent at 805 kb or 17.2 min, near the modABCD operon (7, 9, 15) but showed no similarity in the region of 66 min. Both icdB and the nearby transposon were found to cotransduce with markers near 17 min (nadA3052::Tn10-kan [16] and gal [1]). The data were consistent with the map order icdB-zbg-5483::Tn10-gal. If icdB is located to the left of gal, the icdB ϩ gene should be present in one or more of the lambda derivatives containing cloned genes from this segment of the chromosome (1, 10). In order to test this expectation, I lysogenized an icdB mutant strain separately with clon...