Michel Sylvestre

In this work, we examined the profile of metabolites produced from the doubly para -substituted biphenyl analogs 4,4′-dihydroxybiphenyl, 4-hydroxy-4′-chlorobiphenyl, 3-hydroxy-4,4′-dichlorobiphenyl, and 3,3′-dihydroxy-4,4′-chlorobiphenyl by biphenyl-induced Pandoraea pnomenusa B356 and by its biphenyl dioxygenase (BPDO). 4-Hydroxy-4′-chlorobiphenyl was hydroxylated principally through a 2,3-dioxygenation of the hydroxylated ring to generate 2,3-dihydro-2,3,4-trihydroxy-4′-chlorobiphenyl and 3,4-dihydroxy-4′-chlorobiphenyl after the removal of water. The former was further oxidized by the biphenyl dioxygenase to produce ultimately 3,4,5-trihydroxy-4′-chlorobiphenyl, a dead-end metabolite. 3-Hydroxy-4,4′-dichlorobiphenyl was oxygenated on both rings. Hydroxylation of the nonhydroxylated ring generated 2,3,3′-trihydroxy-4′-chlorobiphenyl with concomitant dechlorination, and 2,3,3′-trihydroxy-4′-chlorobiphenyl was ultimately metabolized to 2-hydroxy-4-chlorobenzoate, but hydroxylation of the hydroxylated ring generated dead-end metabolites. 3,3′-Dihydroxy-4,4′-dichlorobiphenyl was principally metabolized through a 2,3-dioxygenation to generate 2,3-dihydro-2,3,3′-trihydroxy-4,4′-dichlorobiphenyl, which was ultimately converted to 3-hydroxy-4-chlorobenzoate. Similar metabolites were produced when the biphenyl dioxygenase of Burkholderia xenovorans LB400 was used to catalyze the reactions, except that for the three substrates used, the BPDO of LB400 was less efficient than that of B356, and unlike that of B356, it was unable to further oxidize the initial reaction products. Together the data show that BPDO oxidation of doubly para -substituted hydroxychlorobiphenyls may generate nonnegligible amounts of dead-end metabolites. Therefore, biphenyl dioxygenase could produce metabolites other than those expected, corresponding to dihydrodihydroxy metabolites from initial doubly para -substituted substrates. This finding shows that a clear picture of the fate of polychlorinated biphenyls in contaminated sites will require more insights into the bacterial metabolism of hydroxychlorobiphenyls and the chemistry of the dihydrodihydroxylated metabolites derived from them.

There is evidence that many plant secondary metabolites may act as signal molecules to trigger the bacterial ability to metabolize polychlorinated biphenyls (PCBs) during the rhizoremediation process. However, the bases for the PCB rhizoremediation process are still largely unknown. The rhizobacterium Rhodococcus erythropolis U23A is unable to use flavanone as a growth substrate. However, on the basis of an assay that monitors the amount of 4-chlorobenzoate produced from 4-chlorobiphenyl by cells grown co-metabolically on flavanone plus sodium acetate, this flavonoid was previously found to be a potential inducer of the U23A biphenyl catabolic pathway. In this work, and using the same assay, we identified ten other flavonoids that did not support growth, but that acted as inducers of the U23A biphenyl pathway, and we confirmed flavonoid induction of the biphenyl catabolic pathway using quantitative real-time polymerase chain reaction (RT-qPCR) on the bphA gene. We also examined the effect of the growth co-substrate on flavonoid induction. Sodium acetate was replaced by glucose, mannose, sucrose, or mannitol, which are sugars found in plant root exudates. The data showed that the level of induction of strain U23A biphenyl-degrading enzymes was significantly influenced by the nature and concentration of the flavonoid in the growth medium, as well as by the substrate used for growth. Sucrose allowed for an optimal induction response for most flavonoids. Some flavonoids, such as flavone and isoflavone, were better inducers of the biphenyl catabolic enzymes than biphenyl itself. We also found that all flavonoids tested in this work were metabolized by strain U23A during co-metabolic growth, but that the metabolite profiles, as well as the level of efficiency of degradation, differed for each flavonoid. To obtain insight into how flavonoids interact with strain U23A to promote polychlorinated biphenyl (PCB) degradation, we determined the concentration of flavanone at which optimal PCB-degrading performance of strain U23A was achieved. We showed that it corresponded to the concentration required to fully induce the biphenyl catabolic pathway of the strain. Together, our data demonstrate that optimal PCB degradation during the rhizoremediation process will require the adjustment of several parameters, including the presence of the appropriate flavonoids at the proper concentrations and the presence of proper growth substrates that positively influence the ability of flavonoids to induce the pathway.

Xylanases are used in many industrial processes including pulp bleaching, baking, detergent, and the hydrolysis of plant cell wall in biofuels production. In this work we have evolved a single domain GH10 xylanase, Xyn10A_ASPNG, from Aspergillus niger to improve its thermostability. We introduced a rational approach involving as the first step a computational analysis to guide the design of a mutagenesis library in targeted regions which identified thermal important residues that were subsequently randomly mutagenized through rounds of iterative saturation mutagenesis (ISM). Focusing on five residues, four rounds of ISM had generated a quintuple mutant 4S1 (R25W/V29A/I31L/L43F/T58I) which exhibited thermal inactivation half‐life (t_1/2) at 60°C that was prolonged by 30 folds in comparison with wild‐type enzyme. Whereas the wild‐type enzyme retained 0.2% of its initial activity after a heat treatment of 10 min at 60°C and was completely inactivated after 2 min at 65°C, 4S1 mutant retained 30% of its initial activity after 15 min heating at 65°C. Furthermore, the mutant melting temperature (T_m) increased by 17.4°C compared to the wild type. Each of the five mutations in 4S1 was found to contribute to thermoresistance, but the dramatic improvement of enzyme thermoresistance of 4S1 was attributed to the synergistic effects of the five mutations. Comparison of biochemical data and model structure between 4S1 and the wild‐type enzyme suggested that the N‐terminal coil of the enzyme is important in stabilizing GH10 xylanase structure. Based on model structure analyses, we propose that enforced hydrophobic interactions within N‐terminal elements and between N‐ and C‐terminal ends are responsible for the improved thermostability of Xyn10A_ASPNG. Biotechnol. Bioeng. 2015;112: 1081–1091.

In this work, we have compared the ability of Pandoraea pnomenusa B356 and of Burkholderia xenovorans LB400 to metabolize diphenylmethane and benzophenone, two biphenyl analogs in which the phenyl rings are bonded to a single carbon. Both chemicals are of environmental concern. P. pnomenusa B356 grew well on diphenylmethane. On the basis of growth kinetics analyses, diphenylmethane and biphenyl were shown to induce the same catabolic pathway. The profile of metabolites produced during growth of strain B356 on diphenylmethane was the same as the one produced by isolated enzymes of the biphenyl catabolic pathway acting individually or in coupled reactions. The biphenyl dioxygenase oxidizes diphenylmethane to 3-benzylcyclohexa-3,5-diene-1,2-diol very efficiently, and ultimately this metabolite is transformed to phenylacetic acid, which is further metabolized by a lower pathway. Strain B356 was also able to cometabolize benzophenone through its biphenyl pathway, although in this case, this substrate was unable to induce the biphenyl catabolic pathway and the degradation was incomplete, with accumulation of 2-hydroxy-6,7-dioxo-7-phenylheptanoic acid. Unlike strain B356, B. xenovorans LB400 did not grow on diphenylmethane. Its biphenyl pathway enzymes metabolized diphenylmethane, but they poorly metabolize benzophenone. The fact that the biphenyl catabolic pathway of strain B356 metabolized diphenylmethane and benzophenone more efficiently than that of strain LB400 brings us to postulate that in strain B356, this pathway evolved divergently to serve other functions not related to biphenyl degradation.

Summary: The fate of polychlorinated biphenyls ( PCBs ) in soil is driven by a combination of interacting biological processes. Several investigations have brought evidence that the rhizosphere provides a remarkable ecological niche to enhance the PCB degradation process by rhizobacteria. The bacterial oxidative enzymes involved in PCB degradation have been investigated extensively and novel engineered enzymes exhibiting enhanced catalytic activities toward more persistent PCB s have been described. Furthermore, recent studies suggest that approaches involving processes based on plant‐microbe associations are very promising to remediate PCB ‐contaminated sites. In this review emphasis will be placed on the current state of knowledge regarding the strategies that are proposed to engineer the enzymes of the PCB ‐degrading bacterial oxidative pathway and to design PCB ‐degrading plant‐microbe systems to remediate PCB ‐contaminated soil.

Many investigations have provided evidence that plant secondary metabolites, especially flavonoids, may serve as signal molecules to trigger the abilities of bacteria to degrade chlorobiphenyls in soil. However, the bases for this interaction are largely unknown. In this work, we found that BphAE _B356 , the biphenyl/chlorobiphenyl dioxygenase from Pandoraea pnomenusa B356, is significantly better fitted to metabolize flavone, isoflavone, and flavanone than BphAE _LB400 from Burkholderia xenovorans LB400. Unlike those of BphAE _LB400 , the kinetic parameters of BphAE _B356 toward these flavonoids were in the same range as for biphenyl. In addition, remarkably, the biphenyl catabolic pathway of strain B356 was strongly induced by isoflavone, whereas none of the three flavonoids induced the catabolic pathway of strain LB400. Docking experiments that replaced biphenyl in the biphenyl-bound form of the enzymes with flavone, isoflavone, or flavanone showed that the superior ability of BphAE _B356 over BphAE _LB400 is principally attributable to the replacement of Phe336 of BphAE _LB400 by Ile334 and of Thr335 of BphAE _LB400 by Gly333 of BphAE _B356 . However, biochemical and structural comparison of BphAE _B356 with BphAE _p4 , a mutant of BphAE _LB400 which was obtained in a previous work by the double substitution Phe336Met Thr335Ala of BphAE _LB400 , provided evidence that other residues or structural features of BphAE _B356 whose precise identification the docking experiment did not allow are also responsible for the superior catalytic abilities of BphAE _B356 . Together, these data provide supporting evidence that the biphenyl catabolic pathways have evolved divergently among proteobacteria, where some of them may serve ecological functions related to the metabolism of plant secondary metabolites in soil.

Rhodococcus erythropolis U23A is a polychlorinated biphenyl (PCB)-degrading bacterium isolated from the rhizosphere of plants grown on a PCB-contaminated soil. Strain U23A bphA exhibited 99% identity with bphA1 of Rhodococcus globerulus P6. We grew Arabidopsis thaliana in a hydroponic axenic system, collected, and concentrated the plant secondary metabolite-containing root exudates. Strain U23A exhibited a chemotactic response toward these root exudates. In a root colonizing assay, the number of cells of strain U23A associated to the plant roots (5.7 × 10⁵ CFU g⁻¹) was greater than the number remaining in the surrounding sand (4.5 × 10⁴ CFU g⁻¹). Furthermore, the exudates could support the growth of strain U23A. In a resting cell suspension assay, cells grown in a minimal medium containing Arabidopsis root exudates as sole growth substrate were able to metabolize 2,3,4′- and 2,3′,4-trichlorobiphenyl. However, no significant degradation of any of congeners was observed for control cells grown on Luria–Bertani medium. Although strain U23A was unable to grow on any of the flavonoids identified in root exudates, biphenyl-induced cells metabolized flavanone, one of the major root exudate components. In addition, when used as co-substrate with sodium acetate, flavanone was as efficient as biphenyl to induce the biphenyl catabolic pathway of strain U23A. Together, these data provide supporting evidence that some rhodococci can live in soil in close association with plant roots and that root exudates can support their growth and trigger their PCB-degrading ability. This suggests that, like the flagellated Gram-negative bacteria, non-flagellated rhodococci may also play a key role in the degradation of persistent pollutants.