In the present work, four bacterial morphotypes from the same ginger rhizosphere bacterial community were isolated and identified as a consequence of their ability to grow on an enrichment medium  containing 3-oxo-C6-HSL as the sole carbon and nitrogen source. BLAST search analyses of the 16S rDNA sequences identified the strains as belonging to the genera Acinetobacter, Burkholderia, Klebsiella and Microbacterium. In semi-quantitative whole cell assays, we evaluated the AHL-inactivating spectrum of the three Gram-negative isolates. The broadest range of activity was noted for Klebsiella strain Se14 which inactivated each of the 24 structurally diverse AHLs evaluated including the D-isomer of 3-oxo-C6-HSL. Similarly Acinetobacter strain GG2 exhibited a broad spectrum of activity but was less effective against short chain AHLs. In contrast, Burkholderia GG4 was inactive against the unsubstituted AHLs but was active against the 3-oxo-AHLs. Although AHL-degrading activity has not previously been characterized in the genus Burkholderia, a soil isolate from this genus capable of growing on AHLs as the sole nitrogen but not carbon source was reported by Yang et al . This differs from Burkholderia strain GG4 which did not grow on 3-oxo-C6-HSL as a source of both carbon and nitrogen and probably came through the enrichment process as a consequence of AHL turnover by the other bacteria in the ginger rhizosphere community. Nevertheless, when GG4 was incubated with 3-oxo-C6-HSL in PBS buffer, GG4 reduced this AHL to the corresponding 3-hydroxy compound. Similar results were obtained for 3-oxo-C4-HSL and 3-oxo-C8-HSL as well as the D-isomer of 3-oxo-C6-HSL indicating that the activity was not AHL chain length dependent or stereospecific. This simple reduction of a 3-oxo-AHL to the corresponding 3-hydroxy compound is likely to impact on QQ. For example, in Er. carotovora where carbapenem antibiotic biosynthesis and exoenzyme production are regulated by 3-oxo-C6-HSL, the corresponding 3-hydroxy compound has only 1% of the activity of the 3-oxo-AHL . For P. aeruginosa, the 3-hydroxy-C12-HSL was approximately 8-fold less active than the 3-oxo compound . These data suggest that simple modification of the 3-oxo moiety is likely to substantially reduce the activity of 3-oxo-AHLs and to contribute to the QQ activity within a bacterial community. A similar oxido-reductase activity has been observed for a strain of Rhodococcus erythropolis isolated from the tobacco rhizosphere . In contrast to Burkholderia strain GG4, this Gram positive bacterium (R. erythropolis) was unable to reduce 3-oxo-C6-HSL and required an AHL acyl chain of at least eight carbons . However in common with GG4, the activity was only observed on incubation of 3-oxo-AHLs with whole, live bacterial cells as cell lysates were inactive .
For Klebsiella and Acinetobacter, AHL-inactivating activity has previously been noted by Park et al  and Kang et al , respectively. For the former, an AHL-degrading enzyme (AhlK) related to AhlD from Arthrobacter has been cloned and sequenced and by homology suggested to be a lactonase . Here we have shown that the same gene is conserved in the Klebsiella ginger rhizosphere isolate Se14 and have demonstrated that the recombinant enzyme expressed in E. coli is indeed a lactonase with very broad AHL-inactivating activity including both short and long chain AHLs (with saturated or unsaturated acyl side chains of 4 to 14 carbons). These include N -(3-hydroxy-7-cis-tetradecanoyl)homoserine lactone (3-hydroxy-C14:1-HSL), an AHL which was originally termed the Rhizobium small bacteriocin  because it inhibits the growth of Rhizobium leguminosarum strains which carry a 'sensitivity locus' on Sym plasmids such as pRLJ1 . 3-hydroxy-C14:1-HSL is also produced by soil bacteria such as Pseudomonas fluorescens .
Acinetobacter GG2 also degraded a wide range of short and long chain AHLs via a lactonase activity although we were unable to identify the gene involved. Although the mechanism of AHL degradation has not previously been determined in this genus, an Acinetobacter strain isolated from cucumber rhizosphere has been reported to degrade both C6-HSL and N -octadecanoyl homoserine lactone (C18-HSL) as well as the AHLs produced by a biocontrol strain of Pseudomonas chlororaphis and a phytopathogenic strain of Burkholderia glumae . Interestingly, Acinetobacter GG2 not only degrades AHLs but also produces AHLs which we identified as 3-hydroxy-C12-HSL (major) and C12-HSL (minor). Previously Niu et al  showed that the human nosocomial pathogen, Acinetobacter baumannii, produces 3-hydroxy-C12-HSL and C12-HSL via the LuxI synthase, AbaI, the expression of which is AHL dependent. In A. baumannii, AHL-dependent QS appears to contribute to biofilm development since abaI mutants were less biofilm proficient than the parent strain .
Acinetobacter strains isolated from contact lenses have been reported not to produce any AHLs  whereas a study of 43 Acinetobacter strains isolated from both hospital patients and the environment and assayed using AHL biosensors showed that most of the strains examined produced AHLs although these were not chemically characterized . Furthermore a comparative genome analysis of three different Acinetobacter strains from three different environments revealed the presence of a luxIR -type locus in a multidrug resistant clinical A. baumannii isolate which was disrupted by an insertion element in a sensitive strain isolated from human body lice but completely absent from a soil isolate .
In Acinetobacter GG2, 3-hydroxy-C12-HSL accumulated in the growth medium reaching a maximal level after 12 h before rapidly being degraded. This indicates GG2 tightly controls its own AHL production and turnover and suggests that sustained expression (or repression) of the QS target genes is not required in stationary phase. The coupling of AHL synthesis and degradation in the same bacterium has previously been noted for Agrobacterium tumefaciens which produces and degrades 3-oxo-C8-HSL during early stationary phase via a lactonase encoded by attM which is activated by starvation signals and the stress alarmone (p)ppGpp [29, 30]. Similarly, a marine Shewanella strain which produces AHLs in late exponential phase degraded its long chain AHLs in stationary phase via both lactonase and acylase/amidase activities . In polymicrobial biofilms, this Shewanella isolate interfered with AHL production in other bacteria and as a consequence, their ability to enhance the settlement of algal zoospores was compromised . Here, we also found that the ginger rhizosphere Burkholderia isolate GG4 is not only capable of interfering with QS by reducing 3-oxo-AHLs to the corresponding 3-hydroxy compounds but also produces AHLs including 3-oxo-C6-HSL, C9-HSL and 3-hydroxy-C8-HSL. While most Burkholderia strains synthesize C6-HSL and C8-HSL [32, 33], 3-hydroxy-C8-HSL production has only been confirmed in the pathogen, Burkholderia mallei  and tentatively identified in the environmental non-pathogenic Burkholderia xenovorans . In B. mallei, C8-HSL and 3-hydroxy-C8-HSL are produced by two different AHL synthases (BmaI1 and BmaI3) . In Burkholderia GG4, it remains to be established whether 3-hydroxy-C8-HSL is produced directly via a LuxI-type synthase or is a consequence of the reduction of 3-oxo-C8-HSL.
Bacteria such as GG2, GG4 and Se14 which produce and/or modify/degrade QS signals are likely to have a major impact on the properties of polymicrobial bacterial communities. Here we have shown that the ginger rhizosphere isolates were each capable of reducing virulence factor production in both P. aeruginosa and Er. carotovora. However, GG4 was unable to down-regulate lecA (which codes for the cytotoxic galactophilic lectin A ) expression probably as a consequence of its inability to reduce C4-HSL  in contrast to elastase which is predominantly LasR/3-oxo-C12-HSL dependent .