Characterization of the hcnABC gene cluster encoding hydrogen cyanide synthase and anaerobic regulation by ANR in the strictly aerobic biocontrol agent Pseudomonas fluorescens CHA0

Abstract

The secondary metabolite hydrogen cyanide (HCN) is produced by Pseudomonas fluorescens from glycine, essentially under microaerophilic conditions. The genetic basis of HCN synthesis in P. fluorescens CHA0 was investigated. The contiguous structural genes hcnABC encoding HCN synthase were expressed from the T7 promoter in Escherichia coli, resulting in HCN production in this bacterium. Analysis of the nucleotide sequence of the hcnABC genes showed that each HCN synthase subunit was similar to known enzymes involved in hydrogen transfer, i.e., to formate dehydrogenase (for HcnA) or amino acid oxidases (for HcnB and HcnC). These similarities and the presence of flavin adenine dinucleotide- or NAD(P)-binding motifs in HcnB and HcnC suggest that HCN synthase may act as a dehydrogenase in the reaction leading from glycine to HCN and CO2. The hcnA promoter was mapped by primer extension; the -40 sequence (TTGGC ... ATCAA) resembled the consensus FNR (fumarate and nitrate reductase regulator) binding sequence (TTGAT ... ATCAA). The gene encoding the FNR-like protein ANR (anaerobic regulator) was cloned from P. fluorescens CHA0 and sequenced. ANR of strain CHA0 was most similar to ANR of P. aeruginosa and CydR of Azotobacter vinelandii. An anr mutant of P. fluorescens (CHA21) produced little HCN and was unable to express an hcnA-lacZ translational fusion, whereas in wild-type strain CHA0, microaerophilic conditions strongly favored the expression of the hcnA-lacZ fusion. Mutant CHA21 as well as an hcn deletion mutant were impaired in their capacity to suppress black root rot of tobacco, a disease caused by Thielaviopsis basicola, under gnotobiotic conditions. This effect was most pronounced in water-saturated artificial soil, where the anr mutant had lost about 30% of disease suppression ability, compared with wild-type strain CHA0. These results show that the anaerobic regulator ANR is required for cyanide synthesis in the strictly aerobic strain CHA0 and suggest that ANR-mediated cyanogenesis contributes to the suppression of black root rot.

FIG. 1

Recombinant plasmids carrying the hcnABC region from P. fluorescens CHA0. Symbols: , genomic DNA from strain CHA0; , sites of Tn1725 insertions; ▸, orientation of kanamycin resistance and T7 promoters (P_Km and P_T7, respectively) on vector plasmids; T_φ, transcription terminator; x, time of Bal 31 digestion (minutes) and extent of deletions created by Bal 31. The HCN phenotype was assessed by qualitative and quantitative tests for HCN production as follows: −, <5 μM HCN; +, wild-type levels of HCN; overproduction of HCN.

FIG. 2

Nucleotide sequence of the hcn gene cluster of P. fluorescens CHA0 and deduced amino acid sequence of its protein products. The putative start codons and potential ribosome-binding site (SD) are underlined. The transcription start site is marked with +1. Double underlining below the sequence indicates a sequence (ANR-box) homologous to the E. coli FNR-binding site and the −10 region of the hcn promoter. Facing arrows indicate inverted repeats. Four cysteine residues in HcnA discussed as potential binding sites for an Fe—S cluster are indicated by diamonds. Eleven amino acid residues within HcnB and HcnC that define an ADP-binding motif, including three conserved glycine residues (boldface) (56), are marked by number signs, and the variable loop is indicated. Predicted transmembrane segments in HcnB and HcnC are boxed. Four amino acid residues in the C-terminal region of HcnC which are identical to the E. coli Pfl region containing the free glycyl radical in activated Pfl are circled. Introduction of artificial HindIII sites at the 5′ end of the hcnA gene and at the 3′ end of the truncated hcnC′ gene is described in the text.

FIG. 3

Expression of the proteins HcnA, HcnB, and HcnC′. The hcnABC′ genes were cloned in the T7 expression vector pEB16, giving plasmids pME3210 (hcnABC′) and pME3209 (hcnC′) (Fig. 1). Cultures of P. aeruginosa ADD1976 carrying either of these plasmids were induced with IPTG (2 mM) (+) or not induced (−). The proteins were labeled with [³⁵S]methionine, separated by electrophoresis in an SDS–15% polyacrylamide gel, and visualized by autoradiography.

FIG. 4

Cloning strategy for the anr gene of P. fluorescens CHA0 and mutant construction. Details are explained in the text. Plasmids are listed in Table 1. +, positive for complementation of the E. coli fnr mutant JRG1728; −, negative for complementation; anr_Pae, anr gene of P. aeruginosa.

FIG. 5

Alignment of the deduced amino acid sequence of ANR from P. fluorescens (Pfl) with those of P. aeruginosa (Pae) ANR and E. coli (Eco) FNR. The sequences were aligned by use of the computer programs PILEUP and GAP. Identities are indicated by vertical lines, and similarities are indicated by colons. The following features of the E. coli FNR protein are highlighted: ★, cysteine residues required for activity; +, residues suggested to interact with RNA polymerase (43); and =, amino acid residues probably involved in dimerization (30). The predicted helix-turn-helix motif is boxed and marked by a solid line above the sequence. Two amino acid residues in this DNA-binding region (Glu-209 and Ser-212) which are essential for binding to the FNR box (43) are shown in boldface.