The first ethylene-oxidising bacterium, Mycobacterium sp. strain E20, was isolated in 1978 (de Bont and Harder 1978). Subsequently, the characterisation of the enzymes involved in aerobic ethylene metabolism was aimed at utilising such strains for microbial bioremediation of chlorinated ethenes, such as vinyl chloride, as many of these are environmental pollutants (Mattes et al. 2010). However, a separate stream of research has identified another potential use for ethylene-oxidising bacteria – in regulating and monitoring fruit ripening.
As the key hormone responsible for the ripening of most fruits, ethylene levels need to be monitored at every step of the fruit packaging process, and there is a
promising market for a cheap, portable and convenient ethylene sensor. Hence, we turned to Mycobacterium NBB4, the best characterised ethylene-oxidising bacterium, with the aim of designing a cell-based biosensor for ethylene.
In Mycobacterium NBB4, the enzymes required for ethylene metabolism are contained within an operon organised in two bidirectional gene clusters. One gene cluster encodes the catabolic enzymes required to convert ethylene into acetyl-CoA, the final product of the degradation pathway which is subsequently directed to the tricarboxylic acid cycle. The other gene cluster encodes the enzymes responsible for synthesising coenzyme M, a cofactor required for the function of the second enzyme in the metabolic pathway.
While the enzymes required for ethylene metabolism have been investigated in detail (Coleman et al. 2010), the regulatory determinants of ethylene metabolism have yet to be characterised. Studies examining the expression of ethylene-oxidising genes in Mycobacterium NBB4 when grown on different substrates have demonstrated that these genes are only expressed when the strain is exposed to ethylene (Figure 2; Coleman et al. 2011). This suggests that these genes are under the control of an inducible, rather than constitutive, promoter.
A promising candidate for regulation of this operon are two genes present in the gene clusters, designated EtnR1 and EtnR2. We suspect that together, EtnR1 and EtnR2 represent a two-component regulatory system responsible for activating gene expression from the putative promoter EtnP upon exposure to ethylene. Such a two-component system is common in bacteria, involving a histidine protein kinase that phosphorlates another regulatory protein when it senses the environmental stimulus (Stock et al. 2000)
There is strong evidence supporting such a hypothesis:
EtnR1 and EtnR2 are the only genes of unknown function in the well-characterised ethylene metabolism operon.
The regulation of the genes responsible for ethylene degradation is also unknown. This may suggest that EtnR1 and EtnR2 are involved in regulating this process.
EtnR1 and EtnR2 are highly conserved across ethylene-oxidising Mycobacterium.
All other ethylene-oxidising mycobacterial strains such as Mycobacterium rhodesiae NBB3, rhodesiae JS60 and tusciae JS617 contain homologues of EtnR1 and EtnR2 in similar positions in alkene monooxygenase clusters, that have highly conserved residues:
EtnR1 - Mouse over the images to Zoom
Figure 1. Amino acid sequence similarity between EtnR1 (left) and EtnR2 (right) homologues from Mycobacterium NBB4, NBB3, JS60 and JS617. * indicates fully conserved, : indicates strongly similar amino acids, . indicates weakly similar amino acids.
These EtnR1 and EtnR2 homologues all share >70% amino acid identity. Homologues in bacteria that do not oxidise ethylene, on the other hand, only share ~50% or less amino acid identity. Furthermore, Mycobacterium strains that do not oxidise ethylene lack any homologues of EtnR1 and EtnR2.
The closest homologue of EtnR1 is CdaR which forms part of a two component regulatory system common in bacterial gene regulation.
CdaR has been identified as a potential regulator of bacterial diadenylate cyclase, which synthesises the second messenger cyclic diadenylate monophosphate (Mehne et al. 2013, Rismondo et al. 2016). While CdaR has not been well studied, it has been suggested to be a negative regulator of diadenylate cyclase.
EtnR1 has DNA-binding and kinase domains and EtnR2 has a MEDS domain.
From protein domain analysis and searching of protein databases, EtnR1 was found to have a helix-turn-helix domain near the C terminus which is a DNA-binding motif commonly associated with negative repressor transcription factors in prokaryotic systems. EtnR1 was also found to have a kinase domain near the N terminus, which is commonly linked with activating/deactivating phosphorylation marks. EtnR2 has a MEDS domain, which is commonly associated with a substrate sensing-role, being found in methane-sensing for methylotrophs and methanogens.