Microbial chemotaxis has recently been proposed as a widespread phenomenon among motile bacteria towards several distinct xenobiotic compounds and it may therefore be advantageous to use such bacteria in bioremediation . It is suggested that chemotaxis can enhance biodegradation by effectively improving 'pollutant bioavailability' and/or by promoting the formation of microbial consortia with diverse microorganisms harboring complementary degradation capabilities [7, 8, 31, 32]. Several studies have now reported the isolation and characterization of bacteria responding chemotactically to a wide variety of hazardous environmental pollutants, including toluene, trinitrotoluene, atrazine and a variety of nitroaromatic compounds [7–9, 33]. However, information pertaining to bacterial chemotaxis towards some of the recently introduced, highly recalcitrant, chlorinated xenobiotic compounds (e.g. chloro-nitroaromatic compounds, polychlorinated biphenyls, chlorinated anilines etc.) is extremely scarce .
Results presented in this report clearly demonstrate that Burkholderia sp. strain SJ98 is chemotactic towards five CNACs. Furthermore, there is a strong association between the chemotaxis and metabolic transformation of the compounds; a chemotactic response was only observed towards those CNACs that the strain could either completely degrade or co-metabolically transform in the presence of alternative carbon sources. Based on observed intermediates, the following catabolic pathways are proposed for CNACs degradation in strain SJ98: (1) both 4C2NB and 5C2NB are degraded via ONB and 3HAA; (2) 2C4NB is transformed to 3,4DHBA via PNB; and (3) 2C3NP is transformed to 3NC via MNP. The degradation pathway for 2C4NP is via PNP, 4NC and BT, as has already been reported . Interestingly, some of the intermediates identified from the five chemoattractant CNACs degradation/transformation were previously characterized chemoattractants for strain SJ98. These are (1) PNP and 4NC in the 2C4NP pathway; (2) ONB in the 4C2NB and 5C2NB pathways;  PNB in the 2C4NB pathway; and (4) MNP in the 2C3NP pathway. These pathways and chemotactic intermediates have been summarized in Additional file: Figure S3. Chemotaxis of strain SJ98 towards 2C4NP, 4C2NB and 5C2NB and also towards some of their metabolic intermediates strongly suggests metabolism depended chemotaxis to this strains towards these CNACs.
Previous studies have suggested two mechanisms for bacterial chemotaxis towards xenobiotic compounds . The first involves transmembrane signaling by a bacterial chemoreceptor wherein binding of the ligand to the extracellular domain of the chemoreceptor generates a transducible signal and results in chemotaxis. This mechanism is independent of metabolism of the chemoattractant and can therefore also be induced by non-metabolizable structural analogues of the chemoattractant. The second possible mechanism involves energy flux, wherein changes in cellular energy levels resulting from metabolism of chemoattractant molecules induce the chemotactic response. It is necessary for the chemoattractant to be metabolized for this mechanism to be operative . Empirical work on various systems to date provides support for both mechanisms. In support of the first mechanism, Liu and Parales recently reported that Pseudomonas sp. strain ADP was chemotactic towards both atrazine, which it could metabolise, and its s-triazine analogue ametryn, which it could not . They also showed that atrazine degradation and chemotaxis are genetically distinct phenotypes in strain ADP. By contrast, support for the second mechanism comes from studies of the chemotaxis by Pseudomonas putida G7 towards naphthalene [6, 36], P. putida F1 towards toluene , and Ralstonia eutropha JMP134 towards 2,4-dichlorophenoxyacetate , which have all reported the phenomenon to be dependent on and genetically linked to the metabolism of the chemoattractant. It remains to be determined whether the proximal triggers for the chemotactic response are the CNACs themselves or their, e.g. NAC, metabolites.
Our results suggest that a more complex mechanism may operate in respect of the chemotaxis of strain SJ98 towards CNACs. The fact that strain SJ98 does not show chemotaxis towards the non-metabolizable structural analogue 4C2NP suggests metabolism-dependent effects. However, the ability of strain SJ98 to be attracted towards co-metabolically transformed NACs  and CNACs is a notable departure from previous examples of metabolism-dependent mechanisms and raises questions as to the extent of energy flux needed for metabolism-dependent chemotaxis.
Also significant is our finding that cells of strain SJ98 induced to metabolise CNACs can exhibit selective chemotaxis towards CNACs which is not inhibited by co-occurrence of simpler compounds like aspartate or succinate as alternative chemoattractants. This finding confirms that CNAC chemotaxis by this strain is at least to some degree a separate phenomenon from some of the precedents. This could also be an important advantage in the potential application of this strain in the in situ bioremediation of CNAC-contaminated sites. Specific regulation of chemotaxis towards the target compound in contaminated environments often comprising a complex mix of multiple potential chemoattractants could significantly improve the efficiency of in situ bioremediation. The chemotaxis of strain SJ98 towards CNACs therefore could be a fruitful model system for studying both basic and applied aspects of target-specific bacterial chemotaxis.