Systematic screening of 6.104 independent Tn5 insertion mutants of Escherichia coli yielded one new hydrogenase locus, hydF, mapping near 64.8 min, i.e., close to the hydL locus (K. Stoker, L. F. Oltmann, and A. H. Stouthamer, J. Bacteriol. 170:1220-1226, 1988. It regulated specifically the activity of the hydrogenase isoenzymes, formate dehydrogenase and lyase activities being unaffected. In hydF mutants, hydrogenase 1 and 2 activities were reduced to 1 % of the parental level, whereas the electrophoretically labile part was present at about 20% of the parental level. H2 uptake was also reduced to about 20%, which suggested a relationship between these two activities. Experiments with 63Ni indicated that hydrogenase isoenzymes 1 and 2 might be present in these strains but in an inactive form. The hydF product might therefore be a posttranslational activator. At least three other mutant classes were isolated. Additional data were obtained on coisolated, nickel-restorable hydC mutants (L. F. Wu and M.-A. Mandrand-Berthelot, Biochimie 68:167-179, 1986). These strains were found to suffer a general impairment of nickel uptake. Restoration of hydrogenase activities was specific for NiCl2 and inhibited by chloramphenicol, which indicated an effect either on the transcription of hydrogenase(-associated) genes or by cotranslational incorporation in nickel-containing enzymes (e.g., in hydrogenases). The hydC mutation could not be complemented in trans, evidence that the hydC product is not a nickel transport protein but rather a cis-acting regulatory gene. Parent HB101, hydF mutants, and the other mutants were further analyzed by monitoring the induction of hydrogenase and hydrogenase-associated activities upon transition of cells from aerobic to anaerobic growth. These experiments also revealed a correlation between the early-induced H2 uptake route and labile hydrogenase activity. The formate hydrogenlyase induction patterns followed quite well the slower induction patterns of hydrogenases 1 and 2. Expression of most of these components is inducible, depending on the redox state of the environment, the presence of substrates (e.g., formate), and the source of energy (27). The underlying genetic regulation mechanisms, however, are still poorly understood. Although more than 10 genetic loci have been described thus far, mapping near 58 min (11, 15, 22, 24, 30), 65 min (15, 28), and 77 min (31), none of these searches was exhaustive, and it is still unknown whether these studies revealed all loci involved in H, metabolism. We therefore undertook a systematic screening of a large number of Tn5 insertion mutants; this search yielded one new locus, which was further characterized.Another problem is how the products expressed by these loci carry out their functions. This too is largely unknown, although some suggestions have been made. For example, hydC (31) and the loci affected in mutants FD-12 (11) and AK23 (6) might be involved in nickel metabolism, the hyd-17 * Corresponding author. 831 locus might make up the structu...