Genetic and chemical characterization of ibuprofen degradation by Sphingomonas Ibu-2

Abstract

Sphingomonas Ibu-2 has the unusual ability to cleave the acid side chain from the pharmaceutical ibuprofen and related arylacetic acid derivatives to yield corresponding catechols under aerobic conditions via a previously uncharacterized mechanism. Screening a chromosomal library of Ibu-2 DNA in Escherichia coli EPI300 allowed us to identify one fosmid clone (pFOS3G7) that conferred the ability to metabolize ibuprofen to isobutylcatechol. Characterization of pFOS3G7 loss-of-function transposon mutants permitted identification of five ORFs, ipfABDEF, whose predicted amino acid sequences bore similarity to the large and small units of an aromatic dioxygenase (ipfAB), a sterol carrier protein X (SCPx) thiolase (ipfD), a domain of unknown function 35 (DUF35) protein (ipfE) and an aromatic CoA ligase (ipfF). Two additional ORFs, ipfH and ipfI, which encode putative ferredoxin reductase and ferredoxin components of an aromatic dioxygenase system, respectively, were also identified on pFOS3G7. Complementation of a markerless loss-of-function ipfD deletion mutant restored catechol production as did complementation of the ipfF Tn mutant. Expression of subcloned ipfABDEF alone in E. coli did not impart full metabolic activity unless coexpressed with ipfHI. CoA ligation followed by ring oxidation is common to phenylacetic acid pathways. However, the need for a putative SCPx thiolase (IpfD) and DUF35 protein (IpfE) in aerobic arylacetic acid degradation is unprecedented. This work provides preliminary insights into the mechanism behind this novel arylacetic acid-deacylating, catechol-generating activity.

Fig. 1.

ORFs found on two separate contigs in pFOS3G7. Small black arrows represent the location and orientation of loss-of-function transposon insertions. The ORFs identified in dark grey are described in this study. Numbered light-grey arrows represent putative ORFs identified via additional sequencing. Top similarity hits according to blastx analysis of these ORFs, including accession numbers and identity/similarity percentages, are as follows: (1) catechol-2,3-dioxygenase, P47228, 90/94 %; (2) 4-oxalocrotonate tautomerase, Q9RHM8, 84/95 %; (3) periplasmic binding protein of ABC transport system, P21175, 25/46 %; (4) catechol 2,3-dioxygenase, P11122, 69/74 %; (5) 4-hydroxy-2-oxovalerate aldolase, O85977, 90/95 %; (6) 4-oxalocrotonate decarboxylase, Q9KWS3, 53/73 %; (7) 4-oxalocrotonate isomerase, Q9RHM8, 34/59 %; (8) plant-like ferredoxin, P23103, 49/65 %; (9) dehydrogenase, P23102, 54/69 %. Large black arrows represent ORFs with high similarity to conserved transposase genes. The locations of the BamHI and NsiI restriction sites used in the generation of pIPFA–F on contig 1 at positions 1951–1956 and 7565–7560, respectively, are indicated.

Fig. 2.

(a) Percentage of 0.24 mM ibuprofen remaining after 2 days of incubation in E. coli EPI300(pFOS3G7), loss-of-function mutants and the two successful complementation constructs (ipfF and ipfD) as determined by HPLC analysis; n = 3, standard deviations were too small to be visualized effectively (typically <1 % of the means). Control: E. coli EPI300 with no vector. (b) Catecholic polymer accumulation in E. coli EPI300(pFOS3G7), loss-of-function mutants and the two successful complement constructs grown in LB with 0.24 mM ibuprofen and 1.5 mM ferric chloride.

Fig. 3.

(a) Phenylacetate (light grey) and catechol (dark grey) concentration in E. coli EPI300 cultures harbouring pIPFA-F and/or pIPFHI following 18 h of incubation with 1 mM phenylacetate. Negative controls consisted of vectorless E. coli EPI300. pIPFA-F pIPFHI cultures (strain IPFA-FHI) had less residual substrate and more catecholic product after 18 h than other cultures (n = 3, P<0.005). (b) Ibuprofen (light grey) and isobutylcatechol (dark grey) concentration after 18 h incubation with 1 mM ibuprofen (n = 3, P<0.025). (c) E. coli EPI300 harbouring the indicated plasmids following 18 h of incubation with 1 mM ibuprofen (IPF) or phenylacetate (PAA), and containing 1.5 mM ferric chloride for catecholic metabolite visualization.

Fig. 4.

Parent compounds (I) and corresponding catechols (IV) were produced by Sphingomonas Ibu-2 and detected by GC/MS and/or HPLC (Murdoch & Hay, 2005). Expression of ipfABDEFHI in E. coli was demonstrated in this study to be sufficient for the deacylation of phenylacetic acid and ibuprofen (R¹ = methyl, R² = isobutyl, R³ = H). The identity of metabolite II was determined for phenylacetate and ibuprofen via ipfF CoA assays, while the identity of metabolite III is hypothesized based on the putative identities of ipfABHI.

Fig. 5.

Reactions catalysed by sterol carrier protein X family members (SCPx) in animals and bacteria. (a) Thiolytic decondensation of propionyl-CoA from 24-oxo-3α,7α,12α-trihydroxy-5β-cholestanoyl-CoA (Takeuchi et al., 2004). (b) Thiolytic decondensation of propionyl-CoA from 3-ketopristanoyl-CoA (Westin et al., 2007). (c) Thiolytic decondensation of succinyl-CoA from benzoyl-CoA by Aromatoleum aromaticum EbN1 (Kube et al., 2004; Kühner et al., 2005) and Thauera aromatica (Leuthner & Heider, 2000).