Regulation of nod gene expression in Bradyrhizobium japonicum

Bradyrhizobium japonicum infects specific legume plants (e.g., soybean) and establishes a nitrogen fixing symbiosis.  The bacterial genes required for this process are the nod (for nodulation) genes. From our first discovery showing that plant produced isoflavones induced nod gene expression, models of nod gene regulation in B. japonicum have become more and more complex. This complexity is required to allow the bacterium to respond to changes environments (e.g., soil saprophyte to intracellular symbiont) and to finely regulate Nod signal synthesis (product of nod gene action).

Plant isoflavones induce nod gene expression through the action of NodD, a LysR-family transcriptional regulator. More recently, we discovered that a classical two-component regulatory, NodVW also responds to isoflavones and induces nod gene expression. The explanation for this redundancy is that NodV & NodW are required for nodulation of specific hosts for which NodD alone is insufficient.

Soybean nodule infected with a B. japonicum NolA mutant expressing a NodY-GUS (β-glucoronidase) fusion. Unlike the wild type, a NolA mutant shows constitutive in planta expression of the nod genes due to the disruption of the NolA/ NodD2 repression pathway.
©Copyright, 2002.

More recently, our research has focused on NolA and NodD2, which together make up a negative regulatory circuit for nod gene expression. The nolA gene is interesting since it encodes for the synthesis of three distinct proteins from two transcriptional start sites (P1 & P2). The largest of these proteins, NolA1, is required to induced expression of NolA2 and NolA3 from P2, as well as induce expression of NodD2. This latter protein then acts to repress nod gene expression. In a series of recent papers (see below), we have shown that the NolA/NodD2 system responds to tetrameric Nod signals, as a feedback inhibitory loop, and to a density dependent (quorum) regulator. This quorum regulator accumulates in culture medium at high population density. We have recently defined the structure of this molecule (now termed bradyoxetin) and shown that it is unrelated to the common homoserine lactone quorum signals.  Bradyoxetin appears to accumulate inside the symbiosome in which the bacteria reside in the infected plant cells of the nodule. Indeed, in contrast to the wild type, B. japonicum NolA mutants show constitutive in planta expression of the nod genes.

Bradyoxetin appears to be produced by all a-proteobacteria (phylogenetic group to which rhizobia belong). This group also contains important animal pathogens (e.g., Brucella, Bartonella) that also share an intracellular lifestyle with rhizobia. We now believe that bradyoxetin-like molecules are likely crucial for the intracellular growth and pathogenicity of these bacteria. Therefore, bradyoxetin is a promising drug target for treating infections by these bacteria. We are now exploring this interesting hypothesis