Heat-shock regulation was detected for three out of the five members of the groESL multigene family in Bradyrhizobium japonicum. The results uncovered the simultaneous presence of two distinct heat-shock control systems which so far have not been reported to co-exist in a single prokaryotic organism. The first system concerns groESL1 whose transcription is controlled in a sigma32-dependent manner similar to that known from work done with Escherichia coli. Heat-shock control of groESL4 is mediated by the second system, which is characterized by an inverted-repeat DNA structure originally described as a heat-shock regulatory element (CIRCE) in Bacillus subtilis. This element represses expression of groESL4 under non-stress conditions, as inferred from the increased expression of a groESL4'-'lacZ fusion suffering a 4 bp deletion within the CIRCE element. The two control systems clearly differ with respect to the temperature dependence and the kinetics of the heat-shock response, and they also respond differently to the stress signal elicited by incorporation of the amino acid analogue p-F-phenylalanine into cellular protein. Knock-out mutations in groEL4 resulted in an increased expression of groESL4, suggesting that repression via CIRCE depends, itself, upon the cellular level of GroEL4 protein.

The Bradyrhizobium japonicum NifA protein, the central regulator for nitrogen fixation gene expression, is encoded in the fixRnifA operon. This operon is activated during free-living anaerobic growth and in the symbiotic root nodule bacteroid state. In addition, it is expressed in aerobic conditions, albeit at a low level. Here, we report that this pattern of expression is due to the presence of two overlapping promoters: fixRp 1 , which is of the ؊24/؊12 class recognized by the RNA polymerase 54 , and fixRp 2 , which shares homology with the ؊35 and ؊10 regions found in other putative B. japonicum housekeeping promoters. Primer extension analyses showed that fixRp 1 directed the synthesis of a transcript, P1, that starts 12 nucleotides downstream of the ؊12 region. In addition to 54 , P1 was dependent on NifA and low oxygen tension. Transcripts originating from fixRp 2 started at two sites: one coincided with P1, while the most abundant, P2, initiated just two nucleotides further downstream of P1. Expression from fixRp 2 was dependent on the upstream ؊68 promoter region, a region known to bind a putative activator protein, but it was independent of 54 and NifA. This promoter was expressed in aerobic and anaerobic conditions but was not expressed in 30-day-old bacteroids. Mutations in the conserved ؊12 region for the 54 promoter did not show any transcript, because these mutations also disrupted the overlapping ؊10 region of the fixRp 2 promoter. Conversely, mutations at the ؊24 region only affected the 54 -dependent P1 transcript, having no effect on the expression of P2. In the absence of 54 , anaerobic expression from the fixRp 2 promoter was enhanced threefold, suggesting that in the wild-type strain, the two RNA polymerase holoenzymes must compete for binding to the same promoter region.In the soybean root nodule endosymbiont Bradyrhizobium japonicum, as in most of the nitrogen-fixing bacteria, expression of the genes involved in free-living and symbiotic nitrogen fixation (nif and fix) is controlled by the oxygen status of the cell. This control is exerted by regulating both the activity and the synthesis of the central regulatory protein NifA (13,24,29). NifA belongs to the bacterial enhancer-binding protein family of transcriptional regulators that activate promoters recognized by the RNA polymerase holoenzyme containing the alternative sigma factor 54 (E 54 ) (15,21). The 54 promoters are unique among the prokaryotic promoters in having two closely spaced conserved sequences, centered at Ϫ12 and Ϫ24 bp from the transcription start, instead of the more common Ϫ10 and Ϫ35 boxes found in most of the prokaryotic promoters (23,27

This review presents a comparison between the complex genetic regulatory networks that control nitrogen fixation in three representative rhizobial species, Rhizobium meliloti, Bradyrhizobium japonicum, and Azorhizobium caulinodans. Transcription of nitrogen fixation genes (nif and fix genes) in these bacteria is induced primarily by low-oxygen conditions. Low-oxygen sensing and transmission of this signal to the level of nif and fix gene expression involve at least five regulatory proteins, FixL, FixJ, FixK, NifA, and RpoN (sigma 54). The characteristic features of these proteins and their functions within species-specific regulatory pathways are described. Oxygen interferes with the activities of two transcriptional activators, FixJ and NifA. FixJ activity is modulated via phosphorylation-dephosphorylation by the cognate sensor hemoprotein FixL. In addition to the oxygen responsiveness of the NifA protein, synthesis of NifA is oxygen regulated at the level of transcription. This type of control includes FixLJ in R. meliloti and FixLJ-FixK in A. caulinodans or is brought about by autoregulation in B. japonicum. NifA, in concert with sigma 54 RNA polymerase, activates transcription from -24/-12-type promoters associated with nif and fix genes and additional genes that are not directly involved in nitrogen fixation. The FixK proteins constitute a subgroup of the Crp-Fnr family of bacterial regulators. Although the involvement of FixLJ and FixK in nifA regulation is remarkably different in the three rhizobial species discussed here, they constitute a regulatory cascade that uniformly controls the expression of genes (fixNOQP) encoding a distinct cytochrome oxidase complex probably required for bacterial respiration under low-oxygen conditions. In B. japonicum, the FixLJ-FixK cascade also controls genes for nitrate respiration and for one of two sigma 54 proteins.

At least five highly conserved, but disparately regulated groESL operons are present in Bradyrhizobium japonicum. Expression of groESL 3 is coregulated with symbiotic nitrogen fixation genes, implying a role of GroESL chaperonins in the nitrogen fixation process. Null mutants of individual groEL genes, however, were not impaired in symbiotic nitrogen fixation activity. By contrast, the groEL 3 -plus-groEL 4 double mutant strain D4, which is mutated in those groEL genes that contribute most to the GroEL pool under symbiotic conditions, exhibited less than 5% Fix activity as compared to the wildtype. Expression of lacZ fusions made to several representative nif and fix genes was not, or only marginally, reduced in mutant D4, indicating that the requirement of chaperonins for nitrogen fixation does not occur at the level of RegSR-NifA-σ 54 -or FixLJ-FixK 2 -dependent gene regulation. Instead, immunoblot analyses revealed that the level of NifH and NifDK nitrogenase proteins was drastically decreased in extracts prepared from D4 bacteroids and from free-living cells grown anaerobically. Transcriptional fusions of the anaerobically induced groESL 3 promoter (P3) to all five B. japonicum groESL operons and also to groESL from Escherichia coli were integrated into the chromosome of mutant D4. Strains harboring P3 fused to groESL 1 , groESL 2 , groESL 5 , or E. coli groESL partially complemented the symbiotic defect of mutant D4, whereas the wild-type phenotype was completely restored in strains complemented with P3 fused to groESL 3 (control) or groESL 4 . Likewise, the growth defect of an E. coli groEL mutant could be corrected at least partially by individual B. japonicum groESL operons. In conclusion, both series of complementation analyses were not indicative of a strict substrate specificity of any of the B. japonicum groESL gene products, which is in good agreement with their high degree of sequence conservation.

No abstract

No abstract

No abstract

No abstract