Expression of the putA gene encoding proline dehydrogenase from Rhodobacter capsulatus is independent of NtrC regulation but requires an Lrp-like activator protein.

Four Rhodobacter capsulatus mutants unable to grow with proline as the sole nitrogen source were isolated by random Tn5 mutagenesis. The Tn5 insertions were mapped within two adjacent chromosomal EcoRI fragments. DNA sequence analysis of this region revealed three open reading frames designated selD, putR, and putA. The putA gene codes for a protein of 1,127 amino acid residues which is homologous to PutA of Salmonella typhimurium and Escherichia coli. The central part of R. capsulatus PutA showed homology to proline dehydrogenase of Saccharomyces cerevisiae (Put1) and Drosophila melanogaster (SlgA). The C-terminal part of PutA exhibited homology to Put2 (pyrroline-5-carboxylate dehydrogenase) of S. cerevisiae and to aldehyde dehydrogenases from different organisms. Therefore, it seems likely that in R. capsulatus, as in enteric bacteria, both enzymatic steps for proline degradation are catalyzed by a single polypeptide (PutA). The deduced amino acid sequence of PutR (154 amino acid residues) showed homology to the small regulatory proteins Lrp, BkdR, and AsnC. The putR gene, which is divergently transcribed from putA, is essential for proline utilization and codes for an activator of putA expression. The expression of putA was induced by proline and was not affected by ammonia or other amino acids. In addition, putA expression was autoregulated by PutA itself. Mutations in glnB, nifR1 (ntrC), and NifR4 (ntrA encoding sigma 54) had no influence on put gene expression. The open reading frame located downstream of R. capsulatus putR exhibited strong homology to the E. coli selD gene, which is involved in selenium metabolism. R. capsulatus selD mutants exhibited a Put+ phenotype, demonstrating that selD is required neither for viability nor for proline utilization.

Optimizing photoheterotrophic H2 production by Rhodobacter capsulatus upon interposon mutagenesis in the hupL gene.

In Rhodobacter capsulatus, the hupL gene encoding the large subunit of the uptake-hydrogenase (Hup) enzyme complex was mutated by insertion of an interposon. The mutant neither synthesized an active hydrogenase nor grew photoautotrophically. Under conditions of nitrogen (N) limitation, photoheterotrophic cultures of the wild type and the mutant evolved H2 by activity of the nitrogenase enzyme complex. When grown with glutamate as an N source and either D,L-malate or L-lactate as carbon sources, the efficiency of H2 production by the HupL mutant was higher than 90%, whereas wild-type cultures exhibited efficiencies of 54% (with D,L-malate) and 64% (with L-lactate), respectively. With NH4+ as the N source, efficiencies of H2 production were 70% (mutant) and 52% (wild type).

Isolation of Escherichia coli mutants defective in uptake of molybdate.

For the study of molybdenum uptake by Escherichia coli, we generated Tn5lac transposition mutants, which were screened for the pleiotropic loss of molybdoenzyme activities. Three mutants A1, A4, and M22 were finally selected for further analysis. Even in the presence of 100 microM molybdate in the growth medium, no active nitrate reductase, formate dehydrogenase, and trimethylamine-N-oxide reductase were detected in these mutants, indicating that the intracellular supply of molybdenum was not sufficient. This was also supported by the observation that introduction of plasmid pWK225 carrying the complete nif regulon of Klebsiella pneumoniae did not lead to a functional expression of nitrogenase. Finally, molybdenum determination by induced coupled plasma mass spectroscopy confirmed a significant reduction of cell-bound molybdenum in the mutants compared with that in wild-type E. coli, even at high molybdate concentrations in the medium. A genomic library established with the plasmid mini-F-derived cop(ts) vector pJE258 allowed the isolation of cosmid pBK229 complementing the molybdate uptake deficiency of the chlD mutant and the Tn5lac-induced mutants. Certain subfragments of pBK229 which do not contain the chlD gene are still able to complement the Tn5lac mutants. Mapping experiments showed that the Tn5lac insertions did not occur within the chromosomal region present in pBK229 but did occur very close to that region. We assume that the Tn5lac insertions have a polar effect, thus preventing the expression of transport genes, or that a positively acting regulatory element was inactivated.