Ortho-phenylphenol (OPP) is used in the post-harvest treatment of fruits to control fungal infestations during storage1. Its application results in the production of large wastewater volumes which require treatment on site2. The development of biological treatment systems based on the specific ability of microorganisms to degrade OPP could be a viable solution for the detoxification of these effluents. In this context, a Sphingomonas haloaromaticamans strain P3 was recently isolated from a soil collected from a wastewater disposal site3. The bacterium is using OPP as a carbon source and showed high potential for application in biodepuration and bioaugmentation strategies. However, the microbial metabolic pathway of the fungicide and the genetic systems driving its degradation by strain P3 remain unknown. Biodegradation is not always synonymous to detoxification; instead it occasionally leads to the formation of metabolic products that are more toxic or persistent than the parent compound4. Therefore, elucidation of the microbial metabolic pathway of OPP is a prerequisite for the downstream exploitation of strain P3 in any environmental application.

To date little is known about the microbial degradation of OPP. Kohler et al.5 were the first, and the only other study to date, that have studied the microbial degradation of OPP. They isolated a Pseudomonas azelaica strain HPB1 which was able to degrade OPP through the production of 2,3-dihydroxybiphenyl. Jaspers et al.6 identified a gene cluster, hbpCAD, encoding the upper metabolic pathway of OPP which involves the transformation of OPP to 2-hydroxypenta-2,4-dienoateand benzoic acid (BA). The downstream transformation of BA involved a meta-cleavage pathway, although its genetic organization and function was not revealed and the overall network of genes driving the full metabolic pathway of OPP is still not known. HbpA, a flavin-dependent monoxygenase responsible for the initial hydroxylation of OPP by P. azelaica HPB1, was isolated7 and characterized8.

Advances in high-throughput sequencing have shed light into the full genetic armoury of xenobiotic-degrading bacteria9,10, making protein annotation, metabolic pathway prediction and reconstruction feasible11,12. However, the mere presence of putative catabolic genes in a bacterial genome does not guarantee biodegradability in a given environment13. Proteomic analysis offers a unique dynamic view of the catabolic network of xenobiotic-degrading bacteria within the context of the overall cell response and adaptation to pollutant exposure14,15,16. Good knowledge of the physiological response of bacteria during exposure and degradation of organic pollutants is necessary for their future industrial exploitation.

The aim of the present study was to unravel the genetic mechanisms driving the metabolic pathway of OPP and the overall cellular response of S. haloaromaticamans strain P3 during degradation of OPP. To achieve this, a combination of genomics and proteomics coupled with transcription and chromatographic analysis was employed. The total bacterial genome was sequenced, assembled, annotated and used for mapping of the proteome of strain P3 growing on OPP, BA and succinate. Up-regulated enzymes with a putative role in the transformation of OPP were identified and their expression patterns during degradation of OPP and BA was determined by reverse transcription (RT)-q-PCR. The key enzyme involved in the first step of the metabolic pathway, a flavin-dependent monoxygenase, was isolated via heterologous expression, purified and its activity against OPP was verified in vitro. The combinatory use of advanced omic tools is expected to unravel the full genetic network driving microbial degradation of OPP for the first time as part of the wider cellular response of the studied bacterium to fungicide exposure. This knowledge will facilitate the exploitation of S. haloaromaticamans strain P3 in future bioremediation and biodepuration strategies.

Results and Discussion

Genomic analysis of S. haloaromaticamans

The draft genome of S. haloaromaticamans strain P3 had a size of 4.812.401 bp with a mean GC content of 62%, which falls within the average size and GC content values of known sphingomonad genomes (3.4–5.9 Mbp)17. Variation in the size and the GC content of the genomes of Sphingomonads has been attributed to the presence of plasmids and genomic islands18. A total of 4630 ORFs were predicted. The annotation of the sequenced genome generated 57 large contigs assembled into 13 scaffolds with the main ones being scaffolds 1 (4.5 Mbp), 2 (192 kbp) and 3 (92 kbp).

Genes with a putative role in the metabolism of aromatic compounds like OPP were localized in four well-organized operons. This is in contrast to other sphingomonads where the genes for the individual degrading pathways are localized in several gene clusters scattered in the genome19,20. Operons 1 and 2 encoded a BA ortho cleavage pathway (ben/cat operon). Operon 3 comprised genes with a putative role in the upper part of the metabolic pathway of OPP (opp operon) and genes from the lower biphenyl (bph) pathway. Finally operon 4 encoded an incomplete bph pathway (Table 1, Fig. 1). From these, only operon 1 was localized in the 4.5-Mb-scaffold, which based on its size, gene organization and the presence of several housekeeping genes could be considered as the bacterial chromosome. The other three catabolic operons were all located in the 92-kb-scaffold 3 and they were separated by transposases and integrases suggesting their acquisition through horizontal gene transfer. Prophages, transposons and insertion elements are common features of sphingomonads and have been deemed responsible for the genome evolution and the high catabolic versatility of this taxon17,21. The presence of plasmid stabilization proteins (stb) at the 5′ end of operon 3, a near complete Type IV secretion system (virD4 is missing) homologous to the VirB/virD4 secretion system of Agrobacterium tumefaciens22,23, a conjugal transfer protein (TraG)24 and a chromosome partitioning protein (Spo0J) downstream of operon 425 (Fig. 1) strongly suggest that the 55-kb fragment containing operons 2, 3 and 4 is most probably a modular transposon localized in a self-transmissible conjugative plasmid. Recently, Yan et al.10 showed via comparative genomics the key role of plasmid-localized transposons in the evolution of novel metabolic pathways for the degradation of phenylurea herbicides by Sphingomonads.

Table 1 A list of the catabolic genes contained in operons 1, 2, 3 and 4 found in the genome of S. haloaromaticamans strain P3. The organization of these genes in the respective operons are presented in Fig. 1.
Figure 1
figure 1

Genetic organization of operons 1, 2, 3 and 4 containing genes involved in the transformation of OPP by Sphingomonas haloaromaticamans strain P3. Operon 1 is located in scaffold 1 which represents the bacterial chromosome (4.5 Mb) and encodes a complete ortho - cleavage pathway for benzoic acid (BA). Operons 2, 3 and 4 are located in the 92-kb scaffold 3 and encode (2) a second complete ortho - cleavage pathway for BA, (3) enzymes with a putative role in the upper pathway of OPP, and a lower biphenyl (bph) pathway, and (4) an incomplete upper bph pathway. The genomic region downstream of the 3′ end of operon 4 encoding genes of the type IV secretion system and other plasmid conjugation proteins is also shown. Annotation of the different ORFs is shown in Table 1.

The ben/cat operons contained genes for the transformation of BA to catechol (benABCD) and then finally to acetyl-CoA and succinyl-CoA via the ortho - cleavage pathway (catABC and pcaDFIJ). A LysR-type transcriptional regulatory protein (benR) was identified at the 5′ end of each operon and was inversely oriented to other genes (Table 1, Fig. 1). The two ben/cat operons showed similar gene organization with the sole differences being: (i) the pcaDF gene organization; in operon 1 pcaF preceded pcaD and (ii) the position of benC, which was located between benB and benD in operon 1, while a putative benC (Ferredoxin–NAD(P)(+) reductase) was localized at the 3′ end of operon 2 (Fig. 1). The role of each of the ben/cat operons was further elucidated by proteomic and transcription analysis and is discussed in the relevant sections.

Operon 3 contained genes encoding a flavin-dependent monoxygenase (oppA1), a 2,3 dihydroxy, 1,2-dioxygenase (oppC) and two meta-ring fusion product hydrolases (oppD1 and oppD2) (Table 1, Fig. 1). These genes composed a gene cluster (oppD1ACD2), potentially homologous to the hbpCAD gene cluster driving the upper part of the metabolic pathway of OPP in P. azelaica HPB126. The translated products of hbpCAD genes showed 32–45% identities and 51–62% positives to their orthologs OppD1ACD2. Downstream of oppD1ACD2, genes encoding the complete lower biphenyl (bph) pathway were found, including a 2-keto-4-pentanoate hydratase (bphH1), an acetaldehyde dehydrogenase (bphJ) and a 4-hydroxy-2-oxovalerate aldolase (bphI). Denef et al.27 have shown that a similar gene cluster in Burkholderia xenovorans LB400 was responsible for the transformation of 2-hydroxypenta-2,4-dienoate (produced by the hydrolysis of the 2-hydroxy-6-oxo-6-phenyl–2,4-hexadienoic acid) to acetaldehyde and pyruvate. A transcriptional regulatory gene (oppR) showing high homology to a XylR_N superfamily of σ54-dependent transcriptional regulators was identified upstream of oppD1 (Table 1). This is in accordance with the regulation of the hpbCAD operon in P. azelaica which was operated by a σ54-dependent XylR/DmpR transcriptional regulator26.

Operon 4 encodes an incomplete upper bph pathway (for the transformation of biphenyl to BA and 2-hydroxypenta-2,4-dienoate). This consists of bphA1A2A3A4 (multi-component biphenyl dioxygenase), bphB (cis-biphenyl dihydrodiol dehydrogenase) and bphD (2-hydroxy-6-oxo-6-(2′-aminophenyl)hexa-2,4-dienoic acid hydrolase), only missing bphC (2,3-dihydroxybiphenyl dioxygenase). The latter is known to be responsible for the transformation of biphenyl-2,3-diol to 2-hydroxy-2,4-pentadienoate and benzoate28. The upper bph pathway is ubiquitous in soil bacteria29 and is usually found on transposable elements30. A second bphH2 and a gene encoding a second flavin-dependent monoxygenase (putative oppA2) were both located at the 5′ end of this operon. Two putative transcriptional regulatory genes were identified upstream of the bph genes (bphR1R2).

Phylogenetic analysis of the proteins encoded in the four catabolic operons provided insights into their origin and evolution. Operon 1 proteins (i.e. BenA, CatA, PcaD presented as indicative enzymes of the whole operon) clustered within the genus Sphingomonas and more specifically with proteins from a homologous operon of the HCH-degrading strain Sphingomonas MM131 (Supplementary Fig. S1). Their orthologs from operon 2 clustered with proteins found in bacteria from the wider sphingomonad complex (Sphingobium and Novosphingobium) (Supplementary Fig. S1). In contrast, proteins encoded in operons 3 (OppC, OppD1D2) and 4 (BphA3, BphB) were associated with taxonomically distant bacteria (Burkholderiaceae, Streptomyces sp.) (Supplementary Fig. S2). These results suggest that the ben/cat operon 2 was laterally acquired by a member of the sphingomonad complex via horizontal gene transfer, in line with its flanking by tranposases. Whereas operons 3 and 4 constitute a patchwork assembly, with operon 4 probably still undergoing evolution as indicated by the lack of the bphC gene.

Based on the genomic analysis of the S. haloaromaticamans strain P3, a putative metabolic pathway is proposed (Fig. 2a), where OPP is transformed to BA and 2-hydroxypenta-2,4-dienoate (operon 3). These are further transformed to Krebs cycle intermediates through the ben/cat ortho cleavage pathway (operons 1 or 2) and the lower bph pathway (operon 3), respectively. The pathway proposed is similar to the metabolic pathway of OPP by the P. azelaica strain HBP1 with the sole difference of the meta cleavage of BA operated in strain HBP15.

Figure 2
figure 2

(a) The proposed metabolic pathway of ortho-phenylphenol (OPP) depicted by the genomic analysis of the Sphingomonas haloaromaticamans strain P3; (b) the degradation of OPP by strain P3, and the formation of 2,3-dihydroxybiphenyl, benzoic acid (BA) and catechol in inoculated (■) and non-inoculated samples (□). Each value is the mean of three replicates ± the standard deviation.

Analytical determination of selected metabolic products of OPP

Key intermediate transformation products of the proposed metabolic pathway of OPP were detected by HPLC. The degradation of OPP by S. haloaromaticamans was rapid and concurred with the formation of BA, which peaked at 17 h and rapidly reduced thereafter (Fig. 2b). 2,3-dihydroxybiphenyl, the first metabolic product of the pathway, and catechol, the product of BA oxidation, were transiently detected at low levels. The transient formation of 2,3-dihydroxybiphenyl and catechol is probably a function of the capacity of S. haloaromaticamans to transform these intermediates at rates higher than that of their formation. Previous studies with P. azelaica HPB1 and other catechol-degrading bacteria26,32 have suggested that this metabolic strategy is a clever mechanism to overcome the toxicity of such metabolic intermediates to bacterial cells.

Proteomic analysis

Proteomic analysis explored the role of the catabolic operons on the microbial transformation of OPP and provided an overview of the cell adaptation responses of the P3 strain during the degradation of OPP and/or BA. The bacterium was grown under selective conditions with OPP, BA and succinate as sole carbon sources and its proteome was analyzed at the mid-log phase of growth. This coincided with the near complete dissipation of OPP and BA (13 h) (Supplementary Fig. S3). 2D gel-based proteomic analysis identified 229 protein spots that were differentially expressed in the presence of OPP and/or BA compared to succinate (Fig. 3). Among these, 97 and 35 proteins showed differential expression only in the presence of OPP or BA respectively, compared to succinate. In addition 97 proteins showed differential expression in the presence of OPP and BA vs succinate (Fig. 3). Differentially expressed protein spots were excised, sequenced, and annotated. Quantitative and sequencing data of the proteins identified in the proteome of strain P3 are given in Supplementary Tables S1 and S2, respectively, and the position of each protein spot in the 2D-gels is given in Supplementary Fig. S4.

Figure 3
figure 3

(a) 2D gels of the proteome of S. haloaromaticamans cells grown in MSMN + CA + ortho-phenylphenol (OPP), + benzoic acid (BA) or + Succinate. Colored frames on the 2D gels indicate regions where relevant catabolic proteins are located and the enlarged frames offer a focus on the intensity of spots associated to selected catabolic proteins among the different treatments; (b) Venn diagram representing the number of protein spots which showed differential expression in BA- vs succinate - growing cells (35) or in OPP - compared to succinate - growing cells (97). Furthermore, 97 common proteins showed differential expression in OPP- and BA-growing cells vs succinate-growing cells.

OPP catabolic proteins

Thirteen spots were associated with proteins having a putative role in OPP transformation. They were all up-regulated in the presence of OPP and/or BA and showed homology to translated genes from operons 2, 3 and 4 (Table 2). Proteins which were highly up-regulated only in the presence of OPP included (i) a flavin-dependent monoxygenase OppA1, (ii) the meta-ring fusion product hydrolases OppD1 and BphD and (iii) a 4-hydroxy-2-oxovalerate aldolase BphI (Fig. 3). These enzymes are involved either in the upper part of the OPP metabolic pathway (OppA1, OppD1, BphD) or in the transformation of 2-hydroxypenta-2,4-dienoate (BphI), both being modules of the OPP pathway that do not involve BA (Fig. 2a). BenA and BenB were up-regulated only in the presence of BA (Fig. 3), while CatA, PcaI and PcaF were up-regulated in the presence of both OPP and BA (Fig. 3), in line with their role in the downstream metabolism of BA in the OPP (BA is an intermediate metabolite) and BA treatments.

Table 2 Differentially expressed proteins with a putative role in the catabolism of ortho-phenylphenol (OPP) and benzoic acid (BA) in the proteome of S. haloaromaticamans grown on OPP or BA compared to cells grown on Succinate (Succ).

Stress-related proteins

Several proteins involved in bacterial stress response were up-regulated in OPP- and BA-grown cells (Supplementary Table S3). The significant up-regulation of alkyl hydroperoxide reductase and superoxide dismutase, major scavengers of hydroxyperoxides, in the OPP-growing cells suggests the activation of a mechanism to cope with the oxidative stress induced by OPP or its chemically reactive intermediates (i.e. 2,3-dihydroxybiphenyl and catechol) which are known to produce reactive oxygen species and damage cell membranes33. Alkyl hydroperoxide reductase was among the more strongly up-regulated proteins in P. putida KT2440 cells growing in BA, p-hydroxybenzoate, vaniline and phenylethylamine compared to its expression in succinate growing cells34. The concurrent up-regulation of chaperons and chaperonins in the OPP and BA-grown cells is in line with the parallel activation of a stress-response mechanism. Chaperons (DnaK) and chaperonins are essential for the survival of bacteria under stress conditions since they facilitate the correct folding of denaturated proteins in the cytosol35,36. The mobilization of stress response mechanisms by S. haloaromaticamans cells during degradation of OPP and BA suggests that these compounds are not preferred growth substrates. Previous proteomic analyses with other xenobiotic-degrading bacterial strains have also noted a stimulation of the stress response mechanisms during degradation of BA36, phenanthrene37, and phenylurea herbicides38.

Transporters and membrane proteins

Up-regulation of transporters (i.e. ABC transporter) and proteins involved in membrane permeability and stability (i.e. TonB, pesticin receptor, YceI) was detected in the proteome of OPP-grown cells (Supplementary Table S3). This is in line with previous studies with other xenobiotics-degrading strains like Pseudomonas putida KT244039,40. Dominquez-Cuevas et al.33 showed that the primary effect of toluene to P. putida KT2440 is at the cell envelope level, which then leads to reciprocal oxidative damage and mobilization of the stress response system of the bacterial cell. Flagella-domain related proteins were also highly up-regulated in OPP- and BA-grown cells (Supplementary Table S3). Nikodinovic-Runic et al.41 also observed an up-regulation of the biosynthesis of flagella-associated proteins upon exposure of a P. putida to styrene in N starvation conditions. This was attributed to a general stimulation of bacterial motility in response to environmental perturbations like exposure to organic pollutants.

Proteins involved in energy production

A large number of proteins involved in energy production (i.e. ATP synthases, ubiquinol-cytochrome c reductase, electron transfer flavoprotein, NADH dehydrogenase/NAD(P)H nitroreductase, 2,3-bisphosphoglycerate-dependent phosphoglycerate mutase) and in the synthesis of biomolecules (i.e. aspartate aminotransferase, 30S and 50S ribosomal proteins, elongation factors) were up-regulated in OPP- and/or BA-grown cells (Supplementary Table S3). Previous studies with P. putida KT2440 also noted a significant up-regulation of proteins with a similar biosynthetic role in cells grown with xenobiotics34,42. This is probably a response for the de novo synthesis of proteins to counterbalance the toxicity of OPP or its intermediates and the parallel increase in needs of the growing cells. Jang et al.43 investigated the cellular response of Staphylococcus aureus to OPP and observed a strong over-expression of genes encoding ribosomal proteins probably related to an overall stimulation of the translation process triggered by stress conditions.

Overall, proteomic analysis further supported the metabolic pathway deduced by the genomic analysis and showed that S. haloaromaticamans, despite its high degradation efficiency, mobilizes its stress-related cellular mechanisms as a response to the potential toxicity of OPP or its transformation intermediates to bacterial cells.

Transcription analysis of catabolic genes

The expression pattern of all putative catabolic genes identified in the genome and especially those in the proteome of S. haloaromaticamans was determined via RT-q-PCR. This allowed further verification of the role of each of these genes in the metabolism of OPP. Thus, the bacterium was grown in OPP or BA and the expression of these selected genes was determined, along with the degradation of these compounds, and compared to their expression in succinate-grown cells.

Upper OPP pathway

The expression of all genes with a putative catabolic role in the upper part of the OPP pathway (oppD1A1CD2, oppA2) showed similar patterns with significantly higher expression (p < 0.05) in the presence of OPP compared to BA and succinate (Fig. 4 and Supplementary Fig. S5). Their expression significantly increased up to 12-h, which coincided with the near complete degradation of OPP (Supplementary Fig. S3) and dropped to levels similar to the other two treatments by the end of the study (27 h) (Fig. 4). The only exception was oppA2, whose expression did not differ between the different treatments (Supplementary Fig. S5). These findings suggest that from the two flavin-dependent monoxygenases found in the genome of S. haloaromaticamans it is OppA1 that drives the initial hydroxylation of OPP. OppA2 might be involved in the hydroxylation of aromatic compounds other than OPP and BA, however its function to date remains unknown. Among genes of the upper OPP pathway oppC showed the highest expression levels (Fig. 4). This might point to a metabolic strategy aiming to prevent the accumulation of the toxic intermediate 2,3-dihydroxybiphenyl as suggested above. A similar transcription regulation strategy was observed in the other OPP-degrading strain P. azelaica HBP16.

Figure 4
figure 4

The transcription patterns of selected genes involved in the upper metabolic pathway of ortho-phenylphenol (OPP) (oppA1, oppC, oppD1), in the upper and lower bph pathway (bphA1,bphI, bphH1) and in the ortho cleavage of benzoic acid (BA) (ben/cat pathway) by Sphingomonas haloaromaticamans strain P3 growing on MSM + CA + OPP, + BA or + Succinate (Succ). Within each time point bars designated by the same letter are not significantly different at the 5% level. The transcription patterns of the remaining catabolic genes are given in Supplementary Figs S5S8.

Upper and lower biphenyl pathway

Genes bphA1A2A3A4, bphB and bphD (operon 4), showed a significantly higher expression in the presence of OPP compared to BA- and succinate-grown cells (Fig. 4, Supplementary Fig. S6). The expression levels of these genes were much lower compared to the respective genes of the upper OPP pathway (oppD1A1CD2), suggesting an auxiliary role in the degradation of OPP. This is in agreement with the low substrate specificity of biphenyl dioxygenases44,45 and the structural resemblance of OPP and biphenyl, whose transformation converges on the same intermediates acting as weak effectors of the bph pathway. In contrast to the other genes of the operon, bphD showed expression levels equivalent to oppD1 from operon 3 (Supplementary Fig. S6), in line with their significant up-regulation in the proteome of OPP-growing cells (Table 2). The high levels of co-expression of oppD1D2 and bphD might be a strategy for accelerating the transformation of the meta-cleavage product or equal affinity for the common substrate produced by the transformation of OPP and biphenyl.

Genes bphH1 and bphI showed significantly higher expression levels during the first 12-h of OPP degradation compared to their expression in BA- and succinate-grown cells (Fig. 4). The bphH2 gene (operon 4) showed significantly higher expression when grown on OPP (Supplementary Fig. S6), but still much lower compared to its ortholog (bphH1) in operon 3, further verifying the prime role of operon 3 in the transformation of OPP and its intermediates like 2-hydroxypenta-2,4-dienoate.

Figure 5
figure 5

SDS-PAGE analysis of the flavin-dependent monoxygenase of ortho-phenylphenol (OppA1) isolated by Sphingomonas haloaromaticamans strain P3. Lane 1: Molecular mass ladder; Lane 2: Elution from a Protino Glutathione Agarose 4B column, and after in column digestion with protease, where the bands of OppA1, protease and GST-tag are indicated with arrows; (b) OPP degradation by different starting amounts of the purified OppA1. Each value is the mean of three replicates ± the standard deviation.

Ben/cat pathway (lower OPP pathway)

All genes from the same ben/cat operon showed uniform expression patterns. Genes from both operons showed significantly higher expression in BA- (p < 0.05) compared to OPP- and succinate-grown cells (Fig. 4; Supplementary Fig. S7). Whereas OPP induced a significant increase (p < 0.05) in the expression levels only of the operon 2 genes. When the expression level of orthologous genes of the two ben/cat operons were compared, a significantly higher expression of all genes from operon 2 was evident, suggesting that S. haloaromaticamans activates operon 2 for the metabolism of BA formed by OPP transformation. Considering the localization of operon 1 in the bacterial chromosome, it is tempting to assume that this operon is utilized by the bacterium for the consumption of biogenic aromatic organic acids, like substituted benzoic and cinnamic acids with methoxy or hydroxy substituents46, which thrive in the soil environment. Ben/cat operons are ubiquitous in soil bacteria playing a central role in the degradation of biogenic aromatic compounds encountered in soil47. Alternatively, ben/cat operon 1 might be regulated differently depending on the concentration of BA or the growth stage of the bacterium. Denef et al.48 showed that the biphenyl-degrading strain Burkholderia xenovorans LB400 could concurrently utilize two or three pathways to transform BA depending on the initial substrate (BA or biphenyl) and its growth stage.

The presence of two peripheral (upper opp and upper bph) and two central catabolic pathways (ben/cat and lower bph) in S. haloaromaticamans is contrary to the modular construction of catabolic pathways found in most xenobiotic-degrading bacteria, where several peripheral metabolic pathways are funnelling their products into a single central catabolic pathway49. The phylogenetic classification of the genes of the two ben/cat operons and the flanking of operon 2 by transposable elements supports its lateral acquisition from a phylogenetically close bacterium through horizontal gene transfer, rather than through duplication and gradual evolution of operon 1, a strategy commonly utilized by other oligotrophs (i.e. Arthrobacters)50. Sphingomonads are known to utilize horizontal gene transfer as a major mechanism to expand or optimize their catabolic capacities17,18,21. We suggest that S. haloaromaticamans has evolved its capacity to metabolize OPP by gradual acquisition and assemblage of operons 2 and 3 from taxonomically close and distant bacteria, respectively. While the acquisition of operon 3 by S. haloaromaticamans was a key step towards the development of its capacity to transform OPP, the acquisition of operon 2 is most probably an optimization step towards a more efficient metabolism of BA.

Transcriptional regulatory proteins

BenR1 was significantly up-regulated only in the presence of BA, whereas benR2 was significantly up-regulated when grown in OPP and BA (Supplementary Fig. S8), in line with the expression patterns of catabolic genes in their respective operons. Up-regulation of both benR1 and benR2 was observed only at 27 h (after completion of BA degradation) indicating a transcriptional repressor activity. Vasely et al.51 reported a repressor activity of catR on the catABC locus of a Rhodococcus erythropolis strain, whereas other studies have shown that catR acts as an activator of the ortho-cleavage pathway52. The presence of a single regulatory protein BenR for both segments of the pathway (ben and cat) has been reported previously in xenobiotics-degrading sphingomonads31 and it was identified as an effective transcription regulation mechanism induced by both BA and cis, cis-muconate53.

The putative transcriptional regulatory genes of operons 3 and 4 showed higher expression levels (p<0.05) in the presence of OPP (Supplementary Fig. S8), although they exhibited diverse expression patterns. In operon 3, OppR1, a XylR_N-type σ54-dependent transcriptional regulator, showed a significant increase in its expression after completion of OPP degradation (27 h) (Supplementary Fig. S8), suggesting a repressor activity. In contrast, the hbpR of P. azelaica HBP1 was a transcriptional activator of hbpCAD26. In operon 4, bphR1 and bphR2, GntR-type and XylR_N-type σ54-dependent transcriptional regulators respectively, showed increasing expression levels (p<0.05) during degradation of OPP suggesting a transcriptional activator regulatory role (Supplementary Fig. S8). Two-component regulatory systems are a common feature of biphenyl-degrading bacteria composed usually by a GntR family transcriptional regulator, acting either as an activator27,54 or as a repressor28,55, and a second LysR-type transcriptional regulator.

Overall, transcription analysis further verified the metabolic pathway of OPP, as initially depicted by the genomic and proteomic analysis, clarified the role of orthologous enzymes in the metabolic pathway of OPP and provided novel insights into the genetic networking regulating the metabolic pathway of the fungicide by S. haloaromaticamans strain P3.

Heterologous expression of the flavin-dependent monoxygenase OppA1

Proteogenomic and transcription analysis pointed to oppA1 (BHE75_04573) as responsible for the initial hydroxylation of OPP. The flavin-dependent monoxygenase encoded by this gene was isolated via heterologous expression in E. coli and purified (Fig. 5a). The isolated protein had an estimated molecular mass of ca. 68 kDa as determined by SDS-PAGE, in line with its predicted molecular mass based on its amino acid sequence. The recombinant enzyme was functional and degraded OPP in vitro in less than 9 h when 32 or 96 μg were added in the reaction, and in less than 21 h when a lower enzyme amount was used in the in vitro test (9.6 μg; Fig. 5b). These results confirmed that the monoxygenase encoded by oppA1 is responsible for the initial hydroxylation of OPP in the metabolic pathway of OPP by S. haloaromaticamans. Further analysis will focus on the detailed characterization of OppA1, considering the high industrial importance of such monoxygenases in the hydroxylation of aromatic compounds whose hydroxylation by chemical means requires harsh conditions and toxic reagents.

Phylogenetic analysis of OppA1 revealed that it is closely affiliated to other 2,4-dichlorophenol-6-monoxygenases and formed a well-supported clade with the 2-hydroxybiphenyl-3-monoxygenase, responsible for the initial hydroxylation of OPP by P. azelaica HBP1 (formely known as Pseudomonas nitroreducens), and a dichlorophenol hydroxylase from another Sphingomonas isolate (Fig. 6). All 2,4-dichlorophenol-6-monoxygenases belong to the class A of flavoprotein monoxygenases (EC., are known to hydroxylate various aromatic compounds and participate in important biosynthetic and transformation pathways56. The 2-hydroxybiphenyl-3-monoxygenase of P. azelaica HBP1 is one of the most well studied enzymes of the class A flavin-dependent monoxygenases7,8 and it has been used for the industrial scale production of 3-substituted catechols57.

Figure 6
figure 6

Maximum likelihood tree of the two flavin-dependent monoxygenases OppA1 and OppA2 found in the genome of the S. haloaromaticamans using 1000 bootstrap replicates and the Le-Gascuel (LG67) model with gamma rate heterogeneity and accounting for invariable sites.

Concluding Remarks

The metabolic pathway of OPP by the S. haloaromaticamans strain P3 was elucidated using a combination of genomics and proteomics, whose results were further verified by transcription and chromatographic analyses. OPP is transformed, through the upper metabolic pathway, to BA and 2-hydroxypenta-2,4-dienoate. The former is further metabolized via the ortho cleavage pathway and the latter is transformed via the lower bph pathway, both to Krebs cycle intermediates. The key enzyme of the pathway, a flavin-dependent monoxygenase, was isolated and its activity against OPP was verified in vitro. Solid evidence suggest that the catabolic operons controlling the transformation of OPP are part of a 55-kb transposon which was probably acquired by S. haloaromaticamans via horizontal gene transfer. Proteomic analysis revealed the activation of a stress-related response by S. haloaromaticamans during degradation of OPP which was echoed in the up-regulation of associated functions like protein synthesis, energy production, motility and membrane transportation.

Materials and Methods

Bacterial strain, growth conditions and chemicals

The S. haloaromaticamans strain P3, using OPP as a carbon source, was routinely cultivated in a mineral salts medium supplemented with nitrogen and casamino acids (0.15 g L−1) (MSMN + CA)3. The medium was amended with filter sterilized aqueous solutions of OPP (200 mg L−1), BA (200 mg L−1) and succinate (2000 mg L−1). Bacterial growth was determined by optical density at 600 nm (OD600). Analytical standards of OPP (99.9%), BA (99.5%), 2,3-dihydroxybiphenyl (≥98%), catechol (≥99%) were purchased from Sigma-Aldrich (St Louis, USA) and succinate (99%) from PanReac-AppliChem (St. Louis, USA).

Genomic analysis of S. haloaromaticamans

Total DNA was extracted from a fresh culture of S. haloaromaticamans with the Purelink Genomic DNA Mini kit (Invitrogen Life Technologies, USA) and quantified by Qubit (Fisher Scientific, USA). Sequencing was performed by Illumina MiSeq with a 2 × 300 bp paired-end (insert ~550 bp) and a 2 × 300 bp mate-pair (insert ~3000 bp) runs. Genome assembly was performed with Allpaths-LG v5096058 using the default parameters. Genome completeness and purity was checked with the CheckM v0.9.6 software suite59 and annotation of the resulting contigs was performed with Prokka v1.1060 as described in details in the Supplementary Information (see Section SI 1.1.).

Phylogenetic analyses of catabolic enzymes

The translated sequences of the catabolic genes identified via genomic analysis were subjected to maximum likelihood phylogenies using the RAxML software v8.1.2461 as described in details in the Supplementary Information (see section SI 1.2.).

Proteomic analysis of S. haloaromaticamans

Experimental set up and crude protein extraction

Triplicate cultures of MSMN + CA + OPP (50 mg L−1), + BA (62 mg L−1) or + succinate (100 mg L−1) were inoculated with S. haloaromaticamans as described above, aiming to equivalent carbon concentration in all treatments (42.5 mg C L−1). Duplicate non inoculated MSMN + CA + OPP or MSMN + CA + BA were also prepared. All samples were incubated in an orbital shaker at180 rpm and 26 °C. The degradation of OPP and BA and the growth of S. haloaromaticamans was determined at regular intervals by HPLC and OD600 measurements, respectively. When bacterial growth in all treatments reached the mid-log phase (OD600 = 0.15–0.16) bacterial cells were harvested by centrifugation (5000 rpm, 10 min, 4 °C) and used for protein extraction. Bacterial pellet was kept on ice and re-suspended in cold buffer A (50 mM tris-base, 100 mM NaCl, 10% glycerol, pH 7.5). Upon addition of 0.1% TRITON and 0.5 mM PMSF the cells were ultrasonicated on ice three times for 15 sec. Cell debris were removed by centrifugation (12000 rpm, 45 min, 4 °C).

2-D proteomic analysis and protein identification by mass spectrometry

For 2D-PAGE separation crude protein extracts were further clarified, concentrated and protein pellet was solubilized at the desired volume of rehydration buffer as described before62. Protein concentration was determined according to Bradford63 using a Bio-Rad assay kit with BSA as standard. Protein extracts were analyzed by 2D-PAGE as described by Ainalidou et al.64. For each sample 30 μg of total soluble proteins were analysed. Proteins were first separated by isoelectric focusing using gel strips forming an immobilized non-linear pH gradient from 3 to 10 (pH 3–10 NL IPG strips, 11 cm; Bio-Rad) and then by SDS–PAGE using 12.5% Tris-HCl polyacrylamide gels (Bio-Rad) following standard procedures. For each treatment three biological replicates were run in parallel and silver stained. 2-DΕ gels were scanned with Bio-Rad GS-800 Calibrated Densitometer equipped with PDQuest Advanced 2-DΕ Gel Analysis software (version 8.1, Bio-Rad) as previously described64. Data were analyzed by one-way ANOVA (P ≤ 0.05) and means were compared using Student’s t-test (significance level 95%). The statistical significant differences were further combined by the quantitative 2-fold change of spot volume. Spots showing values in the ratios OPP/Succinate and BA/Succinate (volume intensity) lower than 0.5 or higher than 2 were excised from the 2D-PAGE gels and digested with trypsin (more details are available in the Supplementary Information section SI 1.3.).

Tryptic peptide mixtures were analyzed in a MALDI-TOF mass spectrometer (Autoflex-Speed, Bruker Daltonics). The protein identification was carried out by peptide mass fingerprinting on a locally installed Mascot-Server v 2.0 against the genome of S. haloaromaticamans P3 & Uniprot-Trembl databases. The mass error tolerance on the Mascot server was set to 25 ppm, methionine oxidation was considered as a variable modification and cystein carbamido methylation was considered as a fixed modification. Proteins not identified by MALDI-TOF analysis were reanalyzed by HPLC-tandem MS/MS (Thermo Scientific) as described in the Supplementary Information (see Section S.I. 1.3.).

Transcription analysis

Bacterial pellet collected at 4, 10, 12 and 27 h during the proteomics experiment was stored at −80 °C and used for transcription analysis of putative catabolic genes. RNA was extracted with the Nucleospin RNA II kit (Macherey-Nagel, Germany). In most cases a DNAse treatment step (DNAse I, Amplification Grade, Invitrogen Life Technologies) was essential to remove DNA residues from extracted RNA. DNA-free RNA was then reverse transcribed to obtain cDNA (Superscript II, Invitrogen Life Technologies) using random hexamers (Takara, Japan).

Primers for the amplification of all putative catabolic genes of S. haloaromaticamans were manually designed with the program PrimerSelect™ (Lasergene®, DNASTAR) based on the genomic analysis of the studied strain (Supplementary Table S4). The primers were further checked for the potential formation of secondary structures and their specificity was validated in silico (Primer-BLAST, and by PCR, using total DNA of strain P3, and sequencing of the PCR product obtained. Primers for the amplification of the gyrB gene of S. haloaromaticamans (reference gene in the transcription analysis) were also designed based on the genome of the strain P3 (Supplementary Table S4). RT-q-PCR thermocycling conditions and reagents are given in Supplementary Information (see section S.I. 1.4.). Quantification of gene expression was performed according to Pfaif65. Transcription analysis data were subjected to two-way ANOVA and significant differences were detected with the post-hoc Tukey test (p < 0.05). Statistical analysis was performed with the SPSS Statistics (IBM Corp. Version 21.0.) software.

Isolation and in vitro assessment of the activity of flavin-dependent monoxygenase OppA1

Primers were designed to amplify the full length sequence of oppA1 (oppAf 5′ACTCATGGATCCATGACTTCAGCAGTTCAAAAACCG3′ and oppAr 5′ACTCATCTCGAGTTAAGAAGCCGGCGTGAATTTTG3′). Underlined nucleotides indicate restriction sites for enzymes BamHI and XhoI to facilitate ligation into the plasmid vector pGEX-6P-1 (N-terminal GST tag). Amplification of the putative oppA was performed in 50 μl reactions containing 1X Polymerase Buffer, 1.5 mM MgCl2, 0.2 mM of each dNTP, 0.5 μΜ of each primer, 1 U of Thermo Scientific Phusion High-Fidelity DNA Polymerase and 1 μl of total bacterial DNA. Amplification conditions were 98 °C for 30 sec, 30 cycles of 98 °C for 10 sec, 67 °C for 30 sec and 72 °C for 1 min, and final extension at 72 °C for 10 min. The product was purified, digested with BamHI and XhoI and ligated into the pGEX vector digested with the same enzymes. Plasmids were transformed into Escherichia coli BL21(DH3) cells.

Eight hundred ml of E. coli transformed cell cultures were grown in LB + ampicillin (100 μg mL−1) in an orbital shaker (200 rpm) at 37 °C to an OD600 of 0.6. The recombinant enzyme was induced with IPTG (0.025 mM) and cells were harvested, 16 h after induction, by centrifugation at 8000 rpm at 4 °C for 7 min. Cells were re-suspended in Buffer A and lysed with ultrasonication (6 × 10 sec). The samples were centrifuged for 15 min at 14500 rcf at 4 °C. The soluble proteins in the supernatant were collected and purified by passage through pre-equilibrated Protino Glutathione Agarose 4B (Macherey-Nagel, Duren, Germany). The suspension was mixed gently at 4 °C for 2 h. The gel was centrifuged for 5 min at 500 xg and the supernatant was discarded. The gel was resuspended in 100 μl of buffer A, transferred to an appropriate chromatography column (Micro Bio-Spin™ Chromatography Columns) and washed successively with the solutions buffer A, buffer A + Triton (0.2%), buffer A and buffer A + DTT (10 mM). Then, the column outlet was closed with cap and the gel was incubated overnight at 4 °C with buffer A + 3 C protease (55 μg). After the incubation the eluate was collected and the purified protein was run with SDS-PAGE66. The gel was stained with Coomassie Brilliant Blue R-250 and destained with a solution of 30% methanol and 10% acetic acid in deionized water. The activity of the purified enzyme was tested in 1-ml reactions composed of 0.25 mM of OPP, 1 mM of NADH, 20 mM phosphate buffer (pH 7.5) and three protein amounts, 9.6, 32 and 96 μg. Triplicate reactions per protein level were prepared, while triplicate controls without protein were also included. The degradation of OPP was determined by HPLC-UV at 0, 9 and 21 h.

Analytical detection of key OPP transformation products

Triplicate MSMN + CA + OPP (50 mg L−1) cultures were inoculated with S. haloaromaticamans. Triplicate non-inoculated controls were also included as abiotic controls. Immediately after inoculation and at hourly intervals thereafter the degradation of OPP and the formation of the putative metabolites 2,3-dihydroxybiphenyl, BA and catechol were determined by HPLC-UV. All chemicals were extracted from 0.5-mL aliquots and analysed as described before3.

Data availability

The assembled genome of the strain P3 has been deposited at DDBJ/ENA/GenBank under the accession number MIPT00000000. All data generated or analysed during this study are included in this published article and its Supplementary Information files.