The production of 2,4-DAPG was negatively regulated by DsbA1
In an approach to identify novel regulators of 2,4-diacetyphloroglucinol (2,4-DAPG) production in P. fluorescens, the gacA mutant strain PM203 was subjected to a random Tn5 insertion mutagenesis. Among the 5,000 mutants tested, four mutants exhibited the antifungal activity against plant pathogen Rhizoctonia solani compared with the gacA mutant (Table S2). Sequence analysis showed that in one of the mutants, X-2, the transposon was inserted into the dsbA1 gene. The dsbA gene encodes a major periplasmic disulfide-bond-forming protein. An in silico analysis revealed two genes in P. fluorescens 2P24 genome (accession number CP025542) encoding DsbA family proteins (DsbA1 [C0J56_00210] and DsbA2 [C0J56_08555]), which have 28% and 13% amino acid sequence identity with DsbA from E. coli, respectively. In addition, two genes encoding proteins homologous to DsbB (DsbB1 [C0J56_24475] and DsbB2 [C0J56_29125]), which is required for reoxidizing DsbA’s cysteines to regenerate its activity, are found in the 2P24 genome. DsbB1 and DsbB2 of P. fluorescens 2P24 share 29% and 26% identity with E. coli DsbB, respectively.
DsbA family proteins are involved in the oxidative folding of various proteins [19]. To determine whether DsbA1 regulates 2,4-DAPG production, we checked the effect of Dsb proteins on the expression of phlA in strain 2P24. Translation fusion assays showed that mutation in dsbA1, dsbA2, dsbB1, or dsbB2 could not influence the phlA′-′lacZ expression (Figure 1A). Whereas HPLC analysis indicated that more 2,4-DAPG was produced in the dsbA1 and the dsbB1 dsbB2 double mutant than in the wild type (Figure 1B). By contrast the dsbA2 and the single dsbB mutants produced similar amounts to strain 2P24 (Figure 1). Introduction of the plasmid-borne dsbA1 gene in the dsbA1 mutant restored 2,4-DAPG produced to the level of wild-type strain. Similarly, the introduction of the plasmid-borne dsbB1 gene or dsbB2 gene into the dsbB1 dsbB2 double mutant restored the production of 2,4-DAPG (Figure 1B). These results indicated that DsbA1, DsbB1, and DsbB2, but not DsbA2, act as negative regulatory elements in the synthesis of 2,4-DAPG.
DsbA1 regulates the production of 2,4-DAPG in a Gac/Rsm-independent manner
Our results showed that the production of 2,4-DAPG was significantly increased in the mutant X-2. To verify this phenotype, we further constructed the dsbA1 gacA mutant and tested its effect on 2,4-DAPG production. Compared to the gacA mutant, 2,4-DAPG production was significantly increased in the dsbA1 gacA double mutant. This could be complemented by introducing a copy of wild-type dsbA1 on the plasmid pBBR-dsbA1 (Figure 2A).
The GacS/GacA system exerts its function via the small regulatory RNA (sRNA) RsmX, RsmY, and RsmZ to sequester the CsrA/RsmA family proteins RsmA and RsmE [1]. To determine whether DsbA1 negatively regulated 2,4-DAPG production via sRNAs or RsmA and RsmE proteins, we compared the expression of these regulatory elements in wild-type and the dsbA1 mutant. Similar to the wild-type, mutation of dsbA1 could not change rsmX, rsmY, and rsmZ genes expression (Figure 2B). Western blot assay further showed that similar levels of the RsmA and RsmE proteins were observed between the dsbA1 mutant and the wild-type strain 2P24 (Figure 2C & 2D). Taken together, these results suggested that DsbA1 affects the production of 2,4-DAPG in a Gac/Rsm-independent manner in P. fluorescens.
The C235, C275, and C578 cysteine residues of Gcd are essential for the interaction of DsbA1 in vivo
The function of DsbA1 is to form disulfide bonds between consecutive cysteine residues in its target proteins, we thus hypothesized that DsbA1 might catalyze the formation of disulfide bonds on a regulator of 2,4-DAPG production, which is localized on cell membrane or in the periplasmic space. Several proteins containing cysteine residues, including the pathway-specific transcriptional repressor PhlF [20], outer membrane protein OprF [21], and glucose dehydrogenase Gcd [22] were selected for a bacterial two-hybrid system with DsbA1. A strong interaction was only detected between DsbA1 and Gcd (Figure 3 & S1), a glucose dehydrogenase that is required for the conversion of glucose to gluconic acid [23]. Analysis using PredictProtein (http://www.predictprotein.org) suggested that Gcd is a transmembrane protein with six cysteine residues C235, C275, C306, C330, C578, and C678 in the periplasmic space. Individual mutagenesis of these periplasmic cysteine residues into serine revealed the critical roles of C235, C275, and C578 in the interaction between Gcd and DsbA1 (Figure 3). In addition, we noticed that the fusions containing only Gcd were unable to reconstitute significant b-galactosidase activities when coexpressed in E. coli, suggesting that Gcd exerts its biological function as a monomer (Figure 3B).
DsbA1 represses 2,4-DAPG production in a Gcd-dependent manner
The direct interaction between DsbA1 and Gcd raised the possibility that DsbA1 might regulate the production of 2,4-DAPG via Gcd. We thus examined the effect of Gcd on the production of 2,4-DAPG. b-Galactosidase reporter assays showed that the translation phlA′-′lacZ fusion did not differ significantly in the gcd mutant from that in the wild-type (Figure 4A), but 2,4-DAPG production was 3-fold lower than that in the wild-type 2P24. The plasmid-borne gcd gene restored 2,4-DAPG production in the gcd mutant, indicating the positive regulation of Gcd on 2,4-DAPG production (Figure 4B). Furthermore, we observed that repression of 2,4-DAPG production in the dsbA1 mutant was abolished by in-frame deletion of gcd, indicating that DsbA1-mediated repression of 2,4-DAPG is Gcd-dependent (Figure 4B).
Given that DsbA1 interacts with Gcd and that DsbA1 negatively, but Gcd positively influences the concentration of 2,4-DAPG, we hypothesized that mutation in dsbA1 would improve the activity of Gcd. To test this hypothesis, we checked the concentration of 2,4-DAPG in the Gcd cysteine mutations. Interestingly, the C235S, C275S, and C578S mutations increased the concentration of 2,4-DAPG. Whereas the C306S, C330S, and C678S mutations could not change the concentration of 2,4-DAPG in the cells of P. fluorescens (Figure 4B). Gcd catalyzes the conversion of glucose to gluconic acid, which is efficient to solubilize mineral phosphate on NBRIP agar plates. The halo size produced by the wild-type 2P24 on NBRIP plate was about 11 mm in diameter, whereas those formed by the C235S, C275S, and C578S mutations were about 15 mm, indicating that mutations of C235, C275, and C578 improved the function of Gcd (Figure 5).
The effect of dsbA1, dsbB1, and dsbB2 genes on the swimming motility and twitching motility
Previous data showed that DsbA is essential for E. coli cell motility [24]. To verify the role of DsbA and DsbB proteins in cell motility, we examined the motility of strain 2P24 and its derivatives. The results showed that the dsbA1 mutant was defective in both swimming and twitching motilities, however, the dsbA2 mutant had a normal phenotype (Figure 6). Although the single dsbB mutants exhibited significant defects in swimming and twitching motilities, disruption of both the dsbB1 and dsbB2 genes resulted in severe defects in cell motilities (Figure 6). These results indicated DsbA1, DsbB1, and DsbB2 are essential for P. fluorescens 2P24 cell motility.