Genomic Analysis of the Potential for Aromatic Compounds

Home , Achromobacter piechaudii, Alicycliphilus, Burkholderia ambifaria, Burkholderia lata, Oxalobacter, Thiomonas intermedia

bs_bs_banner

Environmental Microbiology (2012) 14(5), 1091–1117 doi:10.1111/j.1462-2920.2011.02613.x

Minireview

Genomic analysis of the potential for aromatic

compounds biodegradation in Burkholderialesemi_2613 1091..1117

Danilo Pérez-Pantoja,1 Raúl Donoso,1,2 in the catabolic clusters of these pathways indicating Loreine Agulló,3 Macarena Córdova,3 recent events in its evolutionary history. In addition, a Michael Seeger,3 Dietmar H. Pieper4 and significant bias towards secondary chromosomes, Bernardo González1,2* now termed chromids, is observed in the distribution 1Center for Advanced Studies in Ecology and of catabolic genes across multipartite genomes, Biodiversity. Millennium Nucleus in Plant Functional which is consistent with a genus-specific character. Genomics. Facultad de Ciencias Biológicas, P. Strains isolated from environmental sources such as Universidad Católica de Chile. Santiago, Chile. soil, rhizosphere, sediment or sludge show a higher 2Facultad de Ingeniería y Ciencias, Universidad Adolfo content of catabolic genes in their genomes com- Ibáñez. Santiago, Chile. pared with strains isolated from human, animal or 3Laboratorio de Microbiología Molecular y Biotecnología plant hosts, but no significant difference is found Ambiental, Departamento de Química, Center for among Alcaligenaceae, Burkholderiaceae and Coma- Nanotechnology and Systems Biology, Universidad monadaceae families, indicating that habitat is more Técnica Federico Santa María, Valparaíso, Chile. of a determinant than phylogenetic origin in shaping 4Microbial Interactions and Processes Research Group, aromatic catabolic versatility. Department of Medical Microbiology, HZI – Helmholtz Centre for Infection Research. Braunschweig, Germany. Introduction Aromatic compounds are widespread in nature, being Summary found as lignin and petroleum components, xenobiotic The relevance of the b-proteobacterial Burkholderi- chemicals, aromatic amino acids and constituents of plant ales order in the degradation of a vast array of exudates, among other sources. The aerobic degradation aromatic compounds, including several priority pol- of aromatic compounds and their halogenated derivatives lutants, has been largely assumed. In this review, the by bacteria has been well studied (Pieper et al., 2010; presence and organization of genes encoding oxyge- Pérez-Pantoja et al., 2010a). The general principle of aro- nases involved in aromatics biodegradation in 80 matics degradation indicates that a broad range of periph- Burkholderiales genomes is analysed. This genomic eral reactions transforms a huge variety of compounds to analysis underscores the impressive catabolic poten- a restricted set of central intermediates, which are subject tial of this bacterial lineage, comprising nearly all of to ring-cleavage and subsequent funnelling into the Krebs the central ring-cleavage pathways reported so far in cycle. Typically, peripheral reactions consist in activation bacteria and most of the peripheral pathways of the aromatic ring through oxygenases and/or CoA involved in channelling of a broad diversity of aro- ligases generating di- or trihydroxylated intermediates matic compounds. The more widespread pathways and/or dearomatized CoA derivatives (Fig. 1) (Pérez- in Burkholderiales include protocatechuate ortho Pantoja et al., 2010a; Pieper et al., 2010). The activation ring-cleavage, catechol ortho ring-cleavage, homo- of the aromatic ring through hydroxylation (Fig. 1, outer gentisate ring-cleavage and phenylacetyl-CoA ring- circle) is commonly catalysed by members of one of three cleavage pathways found in at least 60% of genomes oxygenase families: the Rieske non-haem iron oxygena- analysed. In general, a genus-specific pattern of posi- ses which frequently catalyse the incorporation of two tional ordering of biodegradative genes is observed oxygen atoms (Gibson and Parales, 2000), the flavoprotein monooxygenases (van Berkel et al., 2006) and the soluble diiron multicomponent oxygenases (Leahy et al., Received 20 May, 2011; accepted 11 September, 2011. *For corre- spondence. E-mail [email protected]; Tel. (+56) 2 3311619; 2003). The subsequent ring-cleavage of di- or trihydroxy- Fax (+56) 2 3311906. lated intermediates (Fig. 1, inner circle) can be catalysed

HOOC

COOH OH 3-Hydroxypheny lacetic acid 4-Hydroxyphenylpyruvic acid NH2 Mha6H COOH Gallic acid COOH Phenylacetyl-CoA O COOH Anthranilic acid Benzoyl-CoA COSCoA HOOC HppDO OH COSCoA AntDO Cl HO Benzoic acid COSCoA HO OH COOH Cl OH OH HO BenDO NH OH OH 2 Paa Homogentisic acid Box 2,4-Dichlorophenol Hge Cl Gal HO Cl Sal1H HOOC 2-Aminobenzoyl-CoA COSCoA COSCoA 3,5-Dichlorocatechol Salicylic acid Cl OH O COOH OH OH OH Cca OH 2-Halobenzoic acid Cl OH Cl Cl COOH Ab Cl OH c OH COOH COOH cp4H OH Cl OhbDO T 2,4,6-Trichlorophenol O COOH Cat12 COSCoA Hxq OH HOOC IaaDO Catechol HOOC COOH NH2 6-Chlorohydroxyquinol C HOOC Cl at HO COOH Indole-3-acetic 23 COOH HO OH acid O 4-Hydroxyphenylacetic acid BphDO CHO O COOH N COOH H CHO Hpc HO COOH COOH COOH OH OH OH Homoprotocatechuic Cca COOH Cl acid HO Central Salicylic OH Biphenyl COOH COOH COOH O COOH acid OH Me t abolism HOOC COOH Cl Cl OH 3-Chlorocatechol H C OH COOH 3 CH3 Gen Tmo O Mhb6H COOH zDO COOH HO OH Cb Benzene Phenol HO O COOH CHO Gentisic acid 3-Hydroxybenzoic acid Dhc HOOC OHC COOH HOOC COOH Cl OH HOOC Chlorobenzene HO HOOC Isophthalic COOH HOOC OH DO acid mt COOH COOH C H C CH NH 2 CHO 3 3 COOH COOH 2,3-Dihydroxy-p-cumic acid HOOC OH HOOC NH2 NH 2 HO Protocatechuic O O COOH COOH COOH acid COOH P Phthalic acid h CHO b OH COOH 2-Aminophenol OH 3H COOH H3C CH 3 COOH p -Cumic acid Dhp COOH Hxq HO OH COOH HO HO OH HO OH Hydroquinone HOOC O CH OH H2N COOH 3 Vanillic COOH 3-Hydroxyanthranilic OH OH O CH3 Hydroxyquinol OH acid Isovanillic Terephthalic acid 2,3-Dihydroxyphenylpropionic 4-Hydroxybenzoic acid acid acid acid Res4H DapDO PcmH O HO H C OH Mhp2H HO 3 OH OH p-Cresol HOOC 2,4’-Dihydroxyacetophenone OH Resorcinol 3-Hydroxyphenylpropionic acid

Fig. 1. Overview of peripheral and ring-cleavage pathways for bacterial catabolism of aromatic compounds. The inner circle includes the structures of dearomatized and ring-cleavage products. The outer circle includes the structures of aryl-CoA and dihydroxylated ring-cleavage intermediates. Dotted lines indicate multiple steps. Gene markers listed in Table 1 are bolded. by either intradiol or extradiol dioxygenases. While all bacterial lineage whose genomic sequences have been intradiol dioxygenases described thus far belong to the completed (Chain et al., 2006; Gross et al., 2008; Mattes same superfamily, members of at least three different et al., 2008; Pérez-Pantoja et al., 2008). Based on the families are reported to be involved in the extradiol ring- available literature in the biodegradation ﬁeld, the excep- cleavage of hydroxylated aromatics. Altogether, the inven- tional catabolic potential of Burkholderiales revealed tory of oxygenases involved in activation and cleavage by genomic analysis is not surprising because of the of the aromatic ring is extensive and phylogenetically large number of strains belonging to this order that diverse, including several different families that have have been isolated by its biodegradative abilities (a been recently compiled for phylogenomic studies (Pérez- non-comprehensive list restricted to genus Burkholderia Pantoja et al., 2010b). Analysis of the distribution of these is found in Denef, 2007). The relevance of Burkholderiales aromatic oxygenases-encoding genes among available for environmental biotechnology has been largely pro- bacterial genome sequences has shown that members of posed based on these culture-dependent methods the Burkholderiales order of the b-proteobacteria harbour (O’Sullivan and Mahenthiralingam, 2005; Chiarini et al., the most impressive potential for aromatic compounds 2006; Denef, 2007). However, the unequivocal conﬁrma- catabolism. These biodegradative abilities become out- tion of Burkholderiales preponderance in bioremediation standing in highly specialized degraders belonging to this processes occurring in multiple environmental systems

© 2011 Society for Applied Microbiology and Blackwell Publishing Ltd, Environmental Microbiology, 14, 1091–1117 Degradation of aromatic compounds in Burkholderiales 1093 has been raised with the emergence of stable isotope families in these 80 sequenced genomes is not homo- probing techniques that links functional activity to specific geneous since 48 strains belong to the Burkholderiaceae members of microbial communities (Madsen, 2006). family. The other strains are distributed among the Using this in situ approach, a key biodegradative role for Comamonadaceae (16), Alcaligenaceae (8) and Oxalo- Burkholderiales has been reported in bioremediation of bacteraceae (5) families (Fig. 2). Three unclassified multiple aromatic compounds including: polychloroby- Burkholderiales species, Thiomonas intermedia K12, phenyls (Tillmann et al., 2005; Uhlik et al., 2009), toluene Methylibium petroleiphilum PM1 and Leptothrix cholodnii (Sun et al., 2010), benzene (Liou et al., 2008; Xie et al., SP-6 were also analysed. 2011), benzoate (Pumphrey and Madsen, 2008), salicy- The genome database for Burkholderiales includes late (Singleton et al., 2005), phenol (Manefield et al., bacteria that can be classified, somewhat artificially, 2005), pentachlorophenol (Mahmood et al., 2005), 2,4- based on their source of isolation: (i) human host, mainly dichlorophenoxyacetate (Cupples and Sims, 2007), naph- opportunistic pathogens (18 strains, designated H in thalene (Jeon et al., 2003; Singleton et al., 2005) and Table S1), (ii) animal host, mainly pathogenic (4 strains, phenanthrene (Singleton et al., 2005). Such multiple evi- Z), (iii) plant pathogens (6 strains, P), (iv) rhizosphere and dences consolidate Burkholderiales as the main player in root nodules (12 strains, R), (v) soil (11 strains, S), (vi) microbial ecology of bioremediation treatments for aro- wastewater and sludge (12 strains, W), (vii) sediment (7 matic decontamination. strains, D), (viii) endosymbionts (3 strains, E) and (ix) The order Burkholderiales belongs to the b subclass of miscellaneous sources (7 strains, U). It should be kept in proteobacteria and comprises the families Burkholderi- mind that the isolation source does not necessarily reflect aceae, Oxalobacteraceae, Alcaligenaceae and Coma- the actual habitat/ecological function of a strain in nature. monadaceae including strictly aerobic and facultative For instance, different strains of a given species have anaerobic chemoorganotrophs, obligate and facultative been isolated from different environments (e.g. Burkhold- chemolithotrophs, nitrogen-fixing organisms, as well as eria thailandensis) and may thus be considered to plant, animal and human pathogens (Garrity et al., 2005); thrive in different environmental niches. Then, correlations therefore constituting a phenotypic, metabolic and eco- between the habitat and the presence/absence of cata- logically diverse bacterial lineage. The molecular basis for bolic properties (see the end of this section), as well as this multi level, remarkable versatility has started to be with other phenotypic traits, should be taken with caution. unveiled by the large quantity of sequenced genomes To address the distribution of aromatic degradation now publicly available, allowing comparative genomic capabilities throughout the genomes of sequenced analyses of metabolic traits. This prompted us to carry out members of the Burkholderiales order, 48 oxygenase- a detailed genomic analysis on the presence, organiza- encoding gene markers covering aromatic ring activating tion and phylogenetic relationships among oxygenase- reactions and central ring-cleavage pathways were encoding genes involved in aromatic catabolism in the surveyed (Table 1). Well-characterized key oxygenase Burkholderiales order. This study improves our under- sequences were used as initial seeds to search for standing of the biodegradative potential of this bacterial orthologous sequences. The distribution of aromatic cata- group, revealing the importance of aromatic catabolism bolic functions among the Burkholderiales members is for the habitat adaptability of bacteria. highly variable (Fig. 2). The number of detected catabolic sequence markers ranges from zero in seven strains (Herminimonas arsenicoxydans, Janthinobacterium sp. Distribution of catabolic functions Marseille, Oxalobacter formigenes HOxBLS, O. formi- in Burkholderiales genes OxCC13, Taylorella equigenitalis MCE9, Poly- Genomes belonging to proteobacterial strains (nearly nucleobacter sp. QLW-P1DMWA-1 and Polynucleobacter 1200) make up the more abundant (almost half) bacte- necessarius ssp. necessarius STIR 1) to 25 in Burkhold- rial phylum in the database of genomic sequences. eria lata 383, 30 in Burkholderia xenovorans LB400 and The sequenced proteobacterial genomes comprise 33 in Cupriavidus pinatubonensis JMP134, formerly Ral- 167 genomes of b-proteobacteria that in turn contain stonia eutropha and Cupriavidus necator (Sato et al., 110 genomes belonging to strains of the Burkholderiales 2006). The number of genomes scoring positive for order. Thus, Burkholderiales is the second most repre- the presence of a particular marker is also highly variable. sented proteobacterial order in the genome database While gene markers tracking the phenylacetyl- following Enterobacteriales. The 80 Burkholderiales CoA oxygenase/reductase (Paa), homogentisate 1,2- genomes analysed (Fig. 2, Table S1) comprise represen- dioxygenase (Hge), 4-hydroxyphenylpyruvate dioxyge- tatives of all four Burkholderiales families and include at nase (HppDO), homoprotocatechuate 2,3-dioxygenase least 48 species from 23 genera. However, the distribution (Hpc), benzoate 1,2-dioxygenase (BenDO), catechol 1,2- of members belonging to the different Burkholderiales dioxygenase (Cat12), 4-hydroxybenzoate 3-hydroxylase

Burkholderiaceae Burkholderia multivorans CGD1 DDDD D,DDDDDD D D D D DDDDD 20 Burkholderia multivorans CGD2 DDD D,DDDDDDD DD DDDDDD 19 Burkholderia multivorans ATCC 17616 222222,2222232 33 22111 20 Burkholderia ubonensis Bu DD D DD D DDD 9 Burkholderia vietnamiensis G4 22222,3P32P 222 111 16 Burkholderia glumae BGR1 222 22222 2 2211 13 Burkholderia dolosa AUO158 DDD D DD D DDD 10 Burkholderia cenocepacia MC0-3 2222 2,3 322 1 2 2 2211 1 17 Burkholderia cenocepacia AU1054 2222 1,2 12 2 2211 1 14 Burkholderia cenocepacia HI2424 2222 2,3 322 2 2 2211 1 16 Burkholderia ambifaria AMMD 22222 2,32322 2,22 2 1111 19 Burkholderia ambifaria MC40-6 22222 2,32322 2 2 2 11 1 17 Burkholderia ambifaria IOP40-10 DDDDD D,DDDDD D DDD DDDD 19 Burkholderia lata 383 22222 2,32,223222 1 2 2 3 2211131 25 Burkholderia oklahomensis C6786 DD D DDD D DDD 10 Burkholderia oklahomensis E0147 DD D DDD D DDD 10 Burkholderia thailandensis TXDOH DD D DDD D DDD 10 Burkholderia mallei ATCC 23344 22 2 22 2 111 9 Burkholderia mallei NCTC 10247 22 2 22 2 111 9 Burkholderia mallei SAVP1 22 2 22 11 1,1 9 Burkholderia pseudomallei 1106a 22 2 22 2 111 9 Burkholderia pseudomallei K96243 22 2 22 2 111 9 Burkholderia pseudomallei 668 22 2 22 2 111 9 Burkholderia thailandensis E264 22 2 222 2 111 10 Burkholderia thailandensis MSMB43 DD D DD D DDD 9 Burkholderia rhizoxinica HKI 0454 12124 Burkholderia phymatum STM815 222,23 232 3 2233 11211219 Burkholderia sp. CCGE1002 2132212P2 322232 1,21,321,31224 Burkholderia sp. H160 D DDD DD D D D D,D D DDD D DDD DDD 22 Burkholderia sp. CCGE1001 DDD D D D DDD DDDDDDDDD 18 Burkholderia sp. CCGE1003 DD D DD D DDDDD11 Burkholderia graminis C4D1M DD D D D DD D DDDDDD14 Burkholderia phytofirmans PsJN 2 1 1,1 1 1 2 2 1121112 15 1,1 Burkholderia xenovorans LB400 2 1 21 322,3111 2221 111,1 3,3 1 1,2 1 1,3 2 30 Burkholderia sp. Ch1-1 D D DDD DDD DDDDDD DDDDDD20 Limnobacter sp. MED105 DD 2 Ralstonia sp. 5_7_47FAA DD D D D D DDD9 Ralstonia pickettii 12D 11 1, P12122 1 212214 Ralstonia pickettii 12J 1 1 12122 1 21 2212 Ralstonia solanacearum PSI07 112 111 2 1212212 Ralstonia solanacearum CFBP2957 112 1 1 21 2 2 9 Ralstonia solanacearum GMI1000 112 21,1121 21,212 Cupriavidus metallidurans CH34 2 2,2 1,2 1 2 1 1,1 2 2 2 2 2 1 1 1 18 Cupriavidus pinatubonensis JMP134 2 2,2 1,2 2 2 2 1,2 P,P P,P 1,2 1 1 1 2 2,2 2 2,2 2 2 2 2 1,2 1 1 33 Cupriavidus taiwanensis LMG19424 P12221,21,2110 Cupriavidus necator H16 22 1212 2 2122221,21,2118 Polynucleobacter necessarius STIR1 0 Polynucleobacter sp. QLW-P1DMWA-1 0 Oxalobacter formigenes OxCC13 0 Oxalobacter formigenes HOxBLS 0 Herbaspirillum seropedicae SmR1 11 1,1 111 1 11 1 11 Janthinobacterium sp. Marseille 0 Oxalobacteraceae Herminiimonas arsenicoxydans ULPAs1 0 Taylorella equigenitalis MCE9 0 Bordetella petrii DSM 12804 1 1 1 1 1 1,1 1 1,1 1 1,1 1 1 1 1 1 1 19 Bordetella parapertussis 12822 11 1 1 4 Bordetella pertussis Tohama I 11 1 3 Bordetella bronchiseptica RB50 111115 Bordetella avium 197N 111111118 Achromobacter piechaudii ATCC 43553 DD D D DDDDDDDDDD 14 Alcaligenaceae Achromobacter xylosoxidans A8 1PP1111111,1111111118 Thiomonas intermedia K12 11 2 Methylibium petroleiphilum PM1 1,1 1 1,1 11 1 8 Leptothrix cholodnii SP-6 1,1 1 1 1 1 1 1,1 1 1 1 1,1 1 1 16 Rhodoferax ferrireducens T118 1,1 1 1 1 5 Polaromonas sp. JS666 11 1 11 1,111111P1115 Polaromonas naphthalenivorans CJ2 1,1 1 1 P 1 1 P 1 1,1,1 11 1 11 17 Variovorax paradoxus S110 112 2,2212 1211111116 Variovorax paradoxus EPS 1 1 1 1,1 1 1 1,1 1,1 1 1 1 1 15 Comamonas testosteroni CNB-2 1,1 1,1 1 1 1,1 1 1 1 1 11111118 Comamonas testosteroni S44 DDDDD DD D DD D DDDD 15 Comamonas testosteroni KF-1 DDDDDDDDD D DD D DD DD 17 Delftia acidovorans SPH-1 1,1 1 1 1 1 11 1111113 Verminephrobacter eiseniae EF01-2 11 1 1 1 1 1 1111112 Acidovorax delafieldii 2AN D 1 Acidovorax avenae subsp. citrulli AAC00-1 11 1 1 4 Acidovorax avenae subsp. avenae ATCC 19860 DD DDDD6 Alicycliphilus denitrificans BC 11 1 1 4 Acidovorax sp. JS42 11,1 1 1 5 Comamonadaceae Acidovorax ebreus TPSY 1 1 2 48 14 11 60 6 27 9 14 3 4 3 66 19 13 2 40 31 7 17 21 2 1 6 4 30 1 3 39 2 8 11 17 2 3 14 1 2 1 37 15 21 61 72 26 9 65 31 37

(Phb3H) and protocatechuate 3,4-dioxygenase (Pca34) only twice in these genomes (Fig. 2). The overall distri- key catabolic activities are highly represented in bution of catabolic genes in the Burkholderiales order Burkholderiales genomes (the number of genomes pos- is more heterogeneous than in other bacterial orders sessing such markers ranged between 39 and 72), four (Pérez-Pantoja et al., 2010b), which might reﬂect the eco- oxygenase-encoding genes are found only once and ﬁve logical diversity of this group.

Fig. 2. Distribution of catabolic gene markers through genomes of Burkholderiales strains taxonomically ordered on a 16S rRNA phylogenetic tree. Maximum likelihood phylogenetic tree (left), for 16S rRNA sequences extracted from the 80 Burkholderiales genomes constructed using MEGA 5 software (Tamura et al., 2011). The tree constructed based on the general time-reversible model of nucleotide substitution including a gamma function with ﬁve categories as proposed by Akaike information criterion for selection of appropriate evolutionary model using jModeltest software (Posada and Buckley, 2004). Peripheral and central pathways markers (right), linked by ring-cleavage intermediates, are grouped in shaded or non-shaded interleaved squares. Numbers denote presence of the gene marker sequence in the chromosome (1) or chromids (2, 3) when genomic structure is available. P: plasmid location; D: draft genome sequence, avoiding the precise location of gene markers. The total count of gene markers for each strain is indicated at the external right column. The total count per gene marker in all strains is indicated on the bottom row. Protein similarity searches performed with BLASTP program from the NCBI website using the default parameters (Altschul et al., 1997).

Table 1. Catabolic enzyme markers.

Enzyme Family Enzyme function Abbreviation

Protocatechuate 3,4-dioxygenase Intradiol dioxygenase Intradiol cleavage Pca34 Catechol 1,2-dioxygenase Intradiol dioxygenase Intradiol cleavage Cat12 Hydroxyquinol 1,2-dioxygenase Intradiol dioxygenase Intradiol cleavage Hxq Chlorocatechol 1,2-dioxygenase Intradiol dioxygenase Intradiol cleavage Cca Catechol 2,3-dioxygenase Type I extradiol dioxygenase Extradiol cleavage Cat23 2,3-Dihydroxy-p-cumate-3,4-dioxygenase Type I extradiol dioxygenase Extradiol cleavage Dhc Chlorohydroquinone 1,2-dioxygenase Type I extradiol dioxygenase Extradiol cleavage Chq 2,4,5-Trihydroxytoluene dioxygenase Type I extradiol dioxygenase Extradiol cleavage Tht Protocatechuate 4,5-dioxygenase LigB type dioxygenase Extradiol cleavage Pca45 Gallate 4,5-dioxygenase LigB type dioxygenase Extradiol cleavage Gal Homoprotocatechuate 2,3-dioxygenase LigB type dioxygenase Extradiol cleavage Hpc 2,3-Dihydroxyphenylpropionate LigB type dioxygenase Extradiol cleavage Dhp 1,2-dioxygenase 2-Aminophenol 1,6-dioxygenase LigB type dioxygenase Extradiol cleavage Amn Hydroquinone 1,2-dioxygenase Type IV extradiol dioxygenase Extradiol cleavage Hqo Gentisate 1,2-dioxygenase Cupin superfamily dioxygenase Extradiol cleavage Gen Homogentisate 1,2-dioxygenase Cupin superfamily dioxygenase Extradiol cleavage Hge 3-Hydroxyanthranilate 3,4-dioxygenase Cupin superfamily dioxygenase Extradiol cleavage Han Benzoyl-CoA oxygenase/reductase Soluble diiron oxygenase Dearomatization Box Phenylacetyl-CoA oxygenase/reductase Soluble diiron oxygenase Dearomatization Paa 2-Aminobenzoyl-CoA monooxygenase/ Class A flavoprotein monooxygenase Dearomatization Abc reductase 4-Hydroxybenzoate 3-hydroxylase Class A flavoprotein monooxygenase Forming protocatechuate Phb3H 3-Hydroxybenzoate 4-hydroxylase Class A flavoprotein monooxygenase Forming protocatechuate Mhb4H Salicylate 1-hydroxylase Class A flavoprotein monooxygenase Forming catechol Sal1H 3-Hydroxybenzoate 6-hydroxylase Class A flavoprotein monooxygenase Forming gentisate Mhb6H 3-Hydroxyphenylpropionate 2-hydroxylase Class A flavoprotein monooxygenase Forming 2,3-dihydroxyphenylpropionate Mhp2H 3-Hydroxyphenylacetate 6-hydroxylase Class A flavoprotein monooxygenase Forming homogentisate Mha6H (di)Chlorophenol hydroxylase Class A flavoprotein monooxygenase Forming (di)chlorocatechol CphH 4-Hydroxyphenylacetate 3-hydroxylase Class D flavoprotein monooxygenase Forming homoprotocatechuate Pha3H Trichlorophenol 4-hydroxylase Class D flavoprotein monooxygenase Forming chlorohydroxyquinol Tcp4H Resorcinol 4-hydroxylase Class D flavoprotein monooxygenase Forming hydroxyquinol Res4H Vanillate 3-O demethylase Rieske non-haem iron oxygenase Channelling to protocatechuate VanOD Isovanillate 4-O demethylase Rieske non-haem iron oxygenase Channelling to protocatechuate IvaOD Terephthalate 1,2-dioxygenase Rieske non-haem iron oxygenase Channelling to protocatechuate TphDO Phthalate 3,4-dioxygenase Rieske non-haem iron oxygenase Channelling to protocatechuate PhtDO Isophthalate 3,4-dioxygenase Rieske non-haem iron oxygenase Channelling to protocatechuate IphDO Anthranilate 1,2-dioxygenase Rieske non-haem iron oxygenase Channelling to catechol AntDO Benzoate 1,2-dioxygenase Rieske non-haem iron oxygenase Channelling to catechol BenDO 2-Halobenzoate 1,2-dioxygenase Rieske non-haem iron oxygenase Channelling to catechol OhbDO Biphenyl 2,3-dioxygenase Rieske non-haem iron oxygenase Channelling to catechol BphDO Indole 3-acetate dioxygenase Rieske non-haem iron oxygenase Channelling to catechol IaaDO Chlorobenzene dioxygenase Rieske non-haem iron oxygenase Channelling to chlorocatechol CbzDO p-Cumate 2,3-dioxygenase Rieske non-haem iron oxygenase Channelling to 2,3-dihydroxy-p-cumate CmtDO Salicylate 5-hydroxylase Rieske non-haem iron oxygenase Channelling to gentisate Sal5H Toluene/benzene monooxygenase Soluble diiron oxygenase Forming phenol Tmo Phenol monooxygenase Soluble diiron oxygenase Forming catechol Pmo 4-Hydroxyphenylpyruvate dioxygenase a-Keto acid-dependent dioxygenase Forming homogentisate HppDO 2,4′-Dihydroxyacetophenone dioxygenase Cupin superfamily dioxygenase Forming 4-hydroxybenzoate DapDO p-Cresol methylhydroxylase Flavocytochrome C Forming 4-hydroxybenzoate PcmH

Significantly different presence/absence patterns are replicons have a distinct and consistent set of properties observed within members of Burkholderiales. Genomes (Harrison et al., 2010). By its utility we adopted this of members of the three main genera of the Burkholderi- nomenclature to analyse several interesting facts con- aceae family (Burkholderia, Cupriavidus and Ralstonia) cerning the distribution of catabolic gene markers among encode a significant number of catabolic functions multipartite genomes (numbered as ‘2’ and ‘3’ in Fig. 2). (usually more than nine). An average of 14 catabolic oxy- The distribution of the relative content of catabolic genases are encoded in the 35 Burkholderia strains genes in the main Burkholderiales genera harbouring analysed. As an exception, in the genome of Burkholderia multipartite genomic structure, i.e. the number of these rhizoxinica HKI 0454 only four of these proteins are oxygenases-encoding genes relative to the total number detected. This may reflect the fact that this strain is a of genes in each replicon type, show a strong bias fungal endosymbiont whose genome is significantly towards chromids against the chromosome (Fig. 3A). Sta- smaller than that of the other Burkholderia species tistical analysis (t-test) shows that these differences in (Table S1), and consequently shows a reduced metabolic catabolic content among replicon types are all significant versatility (Lackner et al., 2011). The Pca34, Phb3H, at a confidence level of 95%. Examples are Pca34 and Cat12, Hge, HppDO and Paa encoding genes are present Hpc proteins, which are encoded on chromids in in all Burkholderia species, with the exception of strain all Burkholderia and Cupriavidus strains (except HKI 0454 that only has the paa genes. Members of the B. rhizoxinica). Chromids also predominantly encode phylogenetic lineage formed by Burkholderia phymatum, Cat12 proteins. Noteworthy, only two of the ring-cleavage B. graminis, B. phytofirmans, B. xenovorans, Burkhold- oxygenases analysed are preferentially encoded in the eria sp. CCGE1001, CCGE1002, CCGE1003, H160 and chromosome: homogentisate 1,2-dioxygenase (Hge) and Ch1-1 (Fig. 2), have catabolic properties different from phenylacetyl-CoA oxygenase/reductase (Paa). Hge is those of the so-called ‘Burkholderia cepacia complex’ and typically encoded in the chromosome, except in Burkhold- the so-called ‘Pseudomallei group’, and are the only eria glumae BGR1 and members of Cupriavidus and Ral- Burkholderia species that harbour genes encoding the stonia genera. In strains containing multiple gene copies Abc and Box proteins (Fig. 2), both comprising CoA- of Hge (e.g. B. xenovorans LB400 and Burkholderia sp. dependent oxidative pathways. The overall metabolic CCGE1002), at least one gene is located on the chromo- potential of these nine Burkholderia strains is, in general, some. The Paa protein is also commonly encoded in the more similar to that of strains of the Cupriavidus and chromosome, except in Ralstonia genomes. Again, addi- Ralstonia genera than to strains of the ‘B. cepacia tional copies are localized elsewhere in some genomes complex’. The aromatic metabolic potential of Burkhold- such as those of C. necator H16, C. pinatubonensis eriales strains not belonging to the genera Burkholderia, JMP134, C. taiwanensis LMG19424 and Burkholderia sp. Cupriavidus and Ralstonia is highly heterogeneous. For CCGE1002. Coincident with its preferential chromosomal instance, the Pca34, BenDO and Hpc markers, frequently location, Hge and Paa are the ring-cleavage oxygenases found in Burkholderia, Cupriavidus and Ralstonia, are that show the broadest distribution in Burkholderiales encoded only in three to five genomes of other Burkhold- genomes. Both features suggest that these catabolic eriales strains; and AntDO and IaaDO proteins, commonly routes are ‘core’ pathways in Burkholderiales, which is encoded in Burkholderia genomes, are absent in all of the additionally supported by their involvement in metaboliza- rest, except in Herbaspirillum seropedicae SmR1. In con- tion of aromatic amino acids. On the other hand, the only trast, several representatives with unique metabolic prop- central catabolic markers encoded in plasmids are Cat23 erties were observed in this heterogeneous group. The in Burkholderia vietnamiensis G4, Han in Polaromonas p-cresol methylhydroxylase (PcmH)-encoding genes are sp. JS666 and Cca in Achromobacter xylosooxidans A8 exclusively observed in two of the Comamonas test- and C. pinatubonensis JMP134 (designated as ‘P’ in osteroni genomes and in Polaromonas naphthalenivorans Fig. 2), in accordance with the fact that these pathways CJ2; and C. testosteroni KF-1 is the only strain whose have often been localized on catabolic megaplasmids genome encodes a 2,4,5-trihydroxytoluene dioxygenase (Vedler, 2009). A few trends can also be observed (Tht). when the distribution of peripheral reactions markers Multiple replicons (multipartite) genomes are found in among multipartite genomes is analysed (Fig. 2). all Burkholderia, Cupriavidus and Ralstonia strains, The Phb3H, VanOD, IvaOD, AntDO, IaaDO, DapDO, including the small genome of B. rhizoxinica, and in nine Sal1H, Mhp2H and Mha6H catabolic markers are pre- other Burkholderiales strains. About 70% of completed dominantly encoded in chromids. The HppDO marker genomes contain more than one replicon (Table S1). The encoding 4-hydroxyphenylpyruvate dioxygenase that term ‘chromid’ has been recently proposed for secondary forms homogentisate, the most widely distributed marker chromosomes to distinguish them from both chromo- found in this survey, is always located in the chromosome, somes and plasmids, according to the fact that these except in the four Cupriavidus strains where both this

A Fig. 3. Relative content of catabolic gene 0,80.8 markers in Burkholderiales strains according Chromosome to replicon composition and genome size. A. Per cent of aromatic oxygenase-encoding Chromids genes per total genes encoded in different replicon types of Burkholderiales genera 0,60.6 harbouring multipartite genomes. B. Relation between genome size and percentage of aromatic oxygenase-encoding genes and total enzyme-encoding genes in 0,40.4 Burkholderiales strains harbouring at least one catabolic gene marker. The numbers of enzyme-encoding genes obtained from Integrated Microbial Genomes (IMG) system.

% Aromatic Oxygenase-encoding % Aromatic genes Oxygenase-encoding 0,20.2

0,0 Burkholderia Cupriavidus Ralstonia B 40 0.50,5 Aromatic oxygenase-encoding genes JMP134 enes g

Enzyme-encoding genes g 0.40,4 30 DSMZ 12804 SP-6 CNB-2 CJ2

0.30,3 enase-encodin yg

% Enzyme-encoding genes % Enzyme-encoding 20

0.20,2 % Aromatic % Aromatic ox 10 0.10,1

0 0,0 345678910 Genome size (Mb) peripheral marker and the corresponding Hge central ivorans CJ2, show a disproportionately high content of marker are encoded in chromids. Interestingly, the con- aromatic oxygenases; and especially remarkable is the figuration of these catabolically linked oxygenases being catabolic specialization of C. pinatubonensis JMP134 encoded in different replicons is only observed in the (Lykidis et al., 2010). Ralstonia strains and in B. glumae BGR1. This genetic Can some conclusions about the presence of these arrangement would correspond to an intermediate stage aromatic catabolism properties on the ecological or ‘lif- in the ‘migration’ of this pathway from chromosome to estyle’ groups defined above be drawn? This would be, of chromid that has occurred in the Cupriavidus genus course, one of the relevant outcomes of these analyses. (Fig. 2). In general, within the Burkholderiaceae family, human The relation between genome size and content of (Table S1, H) and/or zoonotic pathogens (Table S1, Z) catabolic oxygenases was compared in Burkholderiales have more aromatic catabolism potential than expected (Fig. 3B). Interestingly, the percentage of aromatic for animal host-associated species. This may be oxygenases-encoding genes increases with genome size, explained by the fact that these strains are predominantly but the relative content of enzymes remains fixed, sug- opportunistic pathogens, and, therefore thrive as free- gesting that the catabolism of aromatic compounds has living organisms. Alternatively, a role for aromatic catabo- increased relevance in Burkholderiales metabolisms, as lism in pathogenesis has been proposed in Burkholderia their genomes get larger. Additionally, it should be noted (Law et al., 2008) as expanded upon in the last section. that some strains as L. cholodnii SP-6, Bordetella petrii Interestingly, the lineage comprising strains B. phymatum, DSMZ 12804, C. testosteroni CNB-2 and P. naphthalen- B. graminis, B. xenovorans, B. phytofirmans, Burkhold-

© 2011 Society for Applied Microbiology and Blackwell Publishing Ltd, Environmental Microbiology, 14, 1091–1117 1098 D. Pérez-Pantoja et al. eria sp. CCGE1001, CCGE1002, CCGE1003, Ch1-1 and content shows disparity when grouped by isolation source H160, which seems to be composed of bacteria that have and not when clustered by taxonomical family, suggests in common the ability to interact beneficially with plants that shaping of the catabolic potential is more influenced (Table S1) (Balandreau and Mavingui, 2006), possesses by habitat than by phylogenetic origin. a different and greater catabolic potential towards aro- The overall genomic distribution of catabolic matic compounds (Fig. 2). The Cupriavidus and Ralstonia oxygenases-encoding genes of four main ring-cleavage genera, also belonging to the Burkholderiaceae family, pathways (catechol, protocatechuate, gentisate and have catabolic properties closely related to that divergent homogentisate) and their peripheral reactions was more Burkholderia group, and are also able to establish benefi- thoroughly analysed in six selected Burkholderiales cial interactions with plants, except for Ralstonia solan- strains (Fig. S1). The strains B. lata 383, B. xenovorans acearum strains (Amadou et al., 2008; Kataoka and Futai, LB400, C. pinatubonensis JMP134, Ralstonia pickettii 2009). The broader catabolic abilities of rhizobacterial 12J, Delftia acidovorans SPH-1 and B. petrii DSM 12804 strains compared with pathogenic members of Burkhold- were selected based on their superior catabolic proper- eriaceae family is reliably due to the nutrient-rich niche ties, representing six relevant taxonomic groups within generated as the result of exudation of a large diversity of Burkholderiales. In general, a supraoperonic clustering compounds, a lot of these having aromatic structures of catabolic genes that encode enzymes channelling (Bais et al., 2006). different aromatic compounds into a common central Although a high heterogeneity is observed among pathway (catabolic islands), as has been reported in members of other Burkholderiales families, some gross the g-proteobacterium Acinetobacter baylyi ADP1, a well- trends can also be detected (Fig. 2). In contrast to known catabolic strain belonging to Pseudomonadales members of the Burkholderiaceae family that seem to (Elsemore and Ornston, 1994), is not observed on the be relatively similar in their metabolic properties, Coma- genomes of the six strains (Fig. S1). On the other hand, monas, Variovorax, Deftia and Polaromonas strains, the genomes of B. petrii DSM 12804 and B. xenovorans members of the Comamonadaceae family, have a LB400 possess several catabolic genes included in quite distinguishable catabolic pattern, whereas the other genomic islands, suggesting that horizontal gene transfer genera of this family, such as Acidovorax, Alicycliphilus played a significant role in establishing the catabolic abili- or Methylibium have a restricted aromatic catabolic poten- ties of these strains (Chain et al., 2006; Gross et al., tial. The Alcaligenaceae family harbours two Achromo- 2008). In order to roughly evaluate the role of horizontal bacter species with interesting catabolic properties, gene transfer in the establishment of the catabolic whereas among Bordetella species, mainly human and markers in these six selected Burkholderiales strains, the animal pathogens, only the sediment isolate B. petrii G+C content of the biodegradative genes and their has remarkable catabolic abilities, which however are genomes were compared. The G+C content of the cata- typically encoded on genomic islands (Gross et al., 2008). bolic gene markers shows values not significantly differ- Members of the Oxalobacteraceae family encode very ent from the mean G+C content of all genes in the few aromatic catabolism markers (Fig. 2), with the sole respective replicons, suggesting that most catabolic exception of H. seropedicae, which is the only represen- genes were ‘trapped’ within their genomes over a long tative associated with plants (Table S1). period of evolution or, alternatively, being acquired from The relative content of aromatic oxygenases-encoding genomes having a similar G+C content (Fig. S1). Addi- genes belonging to peripheral and central catabolic path- tional sequence analyses, using IslandViewer (Langille ways was compared in Burkholderiales genomes classi- and Brinkman, 2009), clarify a putative allochthonous fied by isolation source in three gross groups: H/Z (animal origin for these catabolic markers. Except in a few cases, hosts), R/S (bulk and rhizospheric soils) or W/D (waste- these analyses confirm that most of catabolic genes con- water or sludge and sediment); or by its phylogenetic sidered are not included in genomic islands putatively affiliation considering the three main families showing acquired by horizontal transfer (Fig. S1). It should be catabolic potential: Burkholderiaceae, Comamonadaceae noted that the catabolic pathways considered in the pre- and Alcaligenaceae (Fig. 4). It can be observed that the ceding analyses are the most widely distributed and con- mean of the relative content of catabolic oxygenases is sequently does not constitute an ‘exotic’ feature, as the nearly similar among the three families but a bias towards routes previously reported to be acquired by horizontal non-host-associated strains is found. The patterns of gene transfer in some of these bacteria (Chain et al., catabolic gene content among groups are similar for 2006; Gross et al., 2008). Finally, regarding the overall peripheral or central pathways. Statistical analysis (t-test) genomic distribution of catabolic markers, no evident confirms that these differences in biodegradative potential similarities or patterns in gene distribution through the regarding host or non-host association are all significant selected genomes are observed among these six strains. at a confidence level of 95%. The fact that oxygenase Noteworthy, even the catabolic gene scattering in the

© 2011 Society for Applied Microbiology and Blackwell Publishing Ltd, Environmental Microbiology, 14, 1091–1117 Degradation of aromatic compounds in Burkholderiales 1099 A Fig. 4. Relative content of peripheral and 0,00200.0020 central catabolic markers per total genes encoded in genomes of Burkholderiales Peripheral oxygenases strains grouped by (A) isolation sources Central oxygenases including human and animal hosts (H/Z), soil and rhizosphere (S/R) and wastewater or 0,00150.0015 sludge and sediment (W/D); and (B) taxonomic families harbouring sizeable biodegradative properties.

0,00100.0010

0,00050.0005 Aromatic oxygenase-encoding genes genes per total genes oxygenase-encoding Aromatic 0,0000 H/Z S/R W/D B 0.00200,002 Peripheral oxygenases Central oxygenases

0.00150,002

0.00100,001 -encoding genes per per total genes genes -encoding

0.00050,001 Aromatic oxygenase Aromatic 0,000 Burkholderiaceae Alcaligenaceae Comamonadaceae

Burkholderia strains considered (both possessing similar been reported that Pca is the ring-cleavage substrate tripartite genomic structure) differs drastically (Fig. S1). in the degradation of 3-hydroxybenzoate (Chang and Due to its widespread distribution and essential role in Zylstra, 2008), 4-hydroxybenzoate (Zylstra et al., 1989), catabolism of a wide array of compounds, an in-depth isovanillate, vanillate, veratrate, phthalate, isophthalate analysis on the phylogenetic relationship and gene clus- (Chang and Zylstra, 1998; Providenti et al., 2001; 2006a), tering of the central pathways channelling four key ring- terephthalate (Sasoh et al., 2006), 3-chlorobenzoate cleavage intermediates, i.e. protocatechuate, catechol, (Nakatsu and Wyndham, 1993) and 3-nitrobenzoate homogentisate and gentisate, was performed. The results (Providenti et al., 2006b), among others. The ring- of this analysis are presented in the following sections. cleavage of Pca can be performed by two distinct mecha- nisms, either through intradiol cleavage by Pca34DO, a member of the intradiol dioxygenase superfamily, or Protocatechuate ring-cleavage pathways and through extradiol cleavage by Pca45DO, a member of the peripheral routes LigB superfamily. Pca34DO is an enzyme composed of Protocatechuate (Pca) is a key intermediate in the degra- two different subunits which, however, share substantial dation of the plant biopolymer lignin and several other amino acid identity (Yoshida et al., 1976). Typically adja- aromatic compounds (Fig. 1). In Burkholderiales, it has cently located pcaGH genes, broadly distributed in

Fig. 5. Maximum likelihood phylogenetic tree for a-subunit of protocatechuate 3,4-dioxygenase (Pca34) and gene clusters in which the respective dioxygenase-encoding genes are found. After initial identification of homologous sequences considering similarity by the chance expectation values and scores of BLAST probing, the whole sets of homologous proteins were subject to phylogenetic analysis including proteins with well-known function to refine genuine orthologous identification. The deduced protein sequences were aligned with MUSCLE using default values (Edgar, 2004), and alignments were manually checked and further edited to remove misaligned regions. The alignments of each data set of homologous proteins were used to evaluate models of sequence evolution using Prottest v.2.4 (Abascal et al., 2005). The Akaike information criterion was used to select the most appropriate evolutionary model for each data set. Maximum likelihood phylogenetic trees were constructed based on the selected models using PhyML 3.0 (Guindon et al., 2010) and MEGA 5 (Tamura et al., 2011) software. Confidence in phylogenetic inference was assessed using non-parametric bootstrap resampling. Branches tagged with dots indicate sequences with confirmed functionality not belonging to Burkholderiales strains that have been incorporated for additional phylogenetic comparisons. The genes encoding both subunits of Pca34 are aligned to one another. The functions of the different types of genes included in the clusters are indicated on the right. Dotted arrows indicate the presence of additional genes related to peripheral pathways channelling to protocatechuate. The sizes of genes and intergenic regions are to scale.

Cupriavidus, Ralstonia and Burkholderia strains, encode In Burkholderia and Ralstonia strains, pcaGH genes these subunits (Fig. 5). Analysis of the phylogenetic are not clustered with any other genes encoding enzymes relationship among the a-subunits of Pca34DO (pcaG performing peripheral reactions generating Pca, or with gene) indicates that the Pca34DO gene products from genes encoding enzymes performing subsequent reac- Ralstonia strains are more closely related to those of an tions of the Pca ortho cleavage pathway (Fig. 5). In a- and g-proteobacteria than to the Pca34DO gene prod- contrast, in all three Cupriavidus strains, which harbour ucts from Burkholderia spp. (Fig. 5). a protocatechuate intradiol cleavage pathway, pcaGH

© 2011 Society for Applied Microbiology and Blackwell Publishing Ltd, Environmental Microbiology, 14, 1091–1117 Degradation of aromatic compounds in Burkholderiales 1101 genes are clustered with genes coding for Phb3H (pobA) as well as in B. glumae BGR1 and B. phymatum STM815. transforming 4-hydroxybenzoate to protocatechuate, The gene order is conserved with genes encoding Phb3H as well as with genes encoding 3-carboxy-cis,cis- being followed by genes encoding the a- and b-subunits muconate cycloisomerase (pcaB), a fusion enzyme com- of succinyl-CoA:3-oxoadipate CoA transferase, 3- posed of 4-carboxymuconolactone decarboxylase and carboxy-cis,cis-muconate cycloisomerase, 3-oxoadipate 3-oxoadipate enol-lactone hydrolase domains (pcaL) and enol-lactone hydrolase and 4-carboxymuconolactone a putative transporter for Pca (Fig. 5). It should be noted decarboxylase and preceded by a gene encoding a LysR that the clustering of Pca34DO-encoding genes with sub- type transcriptional regulator. A gene encoding a Major sequent Pca ortho cleavage pathway genes and/or genes Facilitator Superfamily protein is additionally encoded in encoding enzymes of peripheral pathways channelling to all such gene clusters of ‘B. cepacia complex’ strains. Pca is commonly observed in a- and g-proteobacterial This contrasts the situation in Ralstonia strains, where the strains, although a different organization is found in Phb3H-encoding gene is located separately from Pca Cupriavidus strains (Fig. 5). ortho cleavage pathway-encoding genes. Here, the genes It is remarkable that only two Burkholderiaceae encoding enzymes for catabolizing the Pca ortho cleavage strains C. taiwanensis LMG19424 and B. rhizoxinica product are organized in a cluster that also contains HKI 0454 lack Pca34DO. Since even pathogenic 3-oxoadipyl-CoA thiolase and thus encode enzymes strains maintain such genes in their genomes, this for the transformation of 3-carboxy-cis,cis-muconate to pathway can be assumed to be highly relevant for the succinyl-CoA and acetyl-CoA, in the order: a- and environmental fitness of free-living strains. Cupriavidus b-subunits of succinyl-CoA:3-oxoadipate CoA transferase, taiwanensis LMG19424 is a nitrogen-fixing symbiont of 3-oxoadipyl-CoA thiolase, 3-carboxy-cis,cis-muconate legumes and B. rhizoxinica HKI 0454 is an intracellular cycloisomerase, 3-oxoadipate enol-lactone hydrolase and symbiont of the phytopathogenic fungus Rhizopus 4-carboxymuconolactone decarboxylase. microsporus (Table S1). Both strains are noticeably Two different oxygenases channelling substrates to less metabolically versatile than other Cupriavidus or 4-hydroxybenzoate (Fig. 1) are observed in Burkholderi- Burkholderia strains (Fig. 2) (Amadou et al., 2008; ales: DapDO found in 10 Burkholderia strains and in Lackner et al., 2011), indicating that during the evolution C. testosteroni KF-1, and PcmH found in P. naphthalen- of their genomes to the symbiotic lifestyle, many cata- ivorans CJ2 and C. testosteroni strains CNB-2 and S44 bolic traits were lost. (Fig. 2). Other peripheral enzymes channelling substrates Only five Burkholderiales strains outside of the directly to Pca (Fig. 1), where encoding genes are Burkholderiaceae family contain Pca34DO-encoding observed in 3, 4, 8, 13 and 27 Burkholderiales strains are genes (Fig. 2): strains Polaromonas sp. JS666, Variovo- IphDO, ThpDO, IvaOD, PhtDO and VanOD, respectively, rax paradoxus S110 and V. paradoxus EPS of the predominantly in Burkholderia and Comamonadaceae Comamonadaceae family (these strains are exceptional (Fig. 2). The broad diversity of peripheral oxygenases since members of this family usually perform a meta channelling lignin-derived compounds to protocatechuate cleavage of Pca, as described below), Alcaligenes in Burkholderia strains, which only comprise a Pca ortho piechaudii ATCC43553 and B. petrii DSM 12804. In the cleavage pathway, strongly suggests a key role of latter strain, Pca34DO genes are clustered with genes Burkholderia for the recycling of plant-derived carbon of encoding subsequent Pca ortho cleavage pathway steps aromatic compounds. (Fig. 5), and are located in the proximity of genes encod- The Pca meta cleavage pathway has been mainly ing phthalate 3,4-dioxygenase (PhtDO). The acquisition of described in strains of the Comamonadaceae family the Pca ortho cleavage pathway through horizontal trans- (Providenti et al., 2001; Mampel et al., 2005). This cleav- fer in B. petrii DSM 12804 is suggested by the location of age reaction is catalysed by a heteromultimeric Pca45DO the respective gene cluster on a genomic island (Gross composed by two unrelated subunits with the encoding et al., 2008). genes typically located adjacent to each other (Fig. 6). With the sole exception of B. petrii DSM 12804 (Fig. 2), Genome surveys show such genes to be predominantly all Burkholderiales strains harbouring Pca34DO-encoding localized in genomes of members of the Comamona- genes also have a Phb3H-encoding gene, indicating daceae family with seven out of 16 strains where the that 4-hydroxybenzoate is probably the major substrate complete genome sequence is available harbouring such degraded through the Pca ortho cleavage pathway. This is genes (Fig. 2). In contrast to the typical organization as supported by the fact that downstream enzymes of the Pca a heterodimer, a fusion protein is encoded in the genome ortho cleavage pathway (3-carboxymuconate cis,cis of Verminephrobacter eiseniae EF01-2 (Fig. 6). Only two cycloisomerase, 4-carboxymuconolactone decarboxylase Burkholderiales strains outside the Comamonadaceae and enol-lactone hydrolase) are encoded in the same gene family are found to harbour a Pca45DO gene, i.e. cluster as Phb3H in all strains of the ‘B. cepacia complex’ L. cholodnii SP-6, and B. phymatum STM815. The last

© 2011 Society for Applied Microbiology and Blackwell Publishing Ltd, Environmental Microbiology, 14, 1091–1117 1102 .Pérez-Pantoja D. 01SceyfrApidMcoilg n lcwl ulsigLtd, Publishing Blackwell and Microbiology Applied for Society 2011 © tal et . niomna Microbiology Environmental , 14 Fig. 6. Maximum likelihood phylogenetic tree for gallate dioxygenase (Gal) and protocatechuate 4,5-dioxygenase (Pca45) and gene clusters in which the respective dioxygenase-encoding 1091–1117 , genes are found. Branches tagged with dots indicate sequences with confirmed functionality not belonging to Burkholderiales, or belonging to Burkholderiales strains not included in this study, that have been incorporated for additional phylogenetic comparisons. The genes encoding Gal and Pca45 are aligned to one another. The functions of the different types of genes included in the clusters are indicated at the bottom. Other details as indicated in legend of Fig. 5. Degradation of aromatic compounds in Burkholderiales 1103 case is remarkable since B. phymatum STM815 also pos- hydrolase (Mhb4H)-encoding genes are seldom found in sesses a Pca ortho cleavage pathway (Fig. 2, previous Burkholderiales and among strains harbouring a Pca meta section, Fig. 5). cleavage, are found only in the three sequenced C. test- Analysis of the phylogenetic relationship among osteroni strains, two Burkholderia spp. strains and B. phy- Pca45DO gene products shows that enzymes encoded matum STM815 (Fig. 2). Vanillate 3-O-demethylase within Burkholderiales genomes are closely related (VanOD) and phthalate 3,4-dioxygenase (PhtDO) are to each other and different from those of the a- observed in eight and five of the Burkholderiales strains proteobacterial Sphingomonadaceae family (represented owning a Pca meta cleavage pathway respectively. It is by the enzyme from Sphingomonas paucimobilis SYK-6, noteworthy that the full set of five Rieske-type oxygenases Fig. 6), the other proteobacterial family where Pca meta acting on the three phthalate isomers, vanillate and isova- cleavage has been well described (Noda et al., 1990). nillate are found in C. testosteroni KF-1 (Fig. 2) and four or Exceptional are the enzymes of B. phymatum STM815 two of these in C. testosteroni CNB-2 and SA44, respec- and one of Pca45DO proteins encoded in the genomes of tively, in agreement with previous reports on other Coma- C. testosteroni strains CNB-2 and S44, which are only monas strains (Providenti et al., 2006a; Sasoh et al., 2006; distantly related to both Pca45DO of other Burkholderi- Fukuhara et al., 2010). ales and those of Sphingomonadaceae and thus can be Gallate dioxygenases, like Pca45DO belonging to the assumed to be of a different origin (Fig. 6). LigB superfamily of extradiol dioxygenases, have so far Some strains harbour multiple Pca meta cleavage been described in S. paucimobilis SYK-6 (Kasai et al., gene clusters. The full set of genes encoding enzymes 2005) and Pseudomonas putida KT2440 (Nogales et al., for the transformation of the Pca 4,5-dioxygenolytic 2005), and are specific for this substrate and do not trans- ring-cleavage product, 4-carboxy-2-hydroxymuconate- form Pca. Both gallate dioxygenases have sizes signifi- 6-semialdehyde to pyruvate and oxaloacetate (4- cantly larger than those of the b-subunits of Pca45DO and carboxy-2-hydroxymuconate semialdehyde dehy- were suggested to have evolved from the fusion of large drogenase, 2-pyrone-4,6-dicarboxylate hydrolase, 4- and small subunits. Respective proteins are encoded in oxalomesaconate hydratase and 4-carboxy-4-hydroxy-2- 13 Burkholderiales genomes and based on their similari- oxoadipate aldolase) are typically clustered with the ties could be grouped into two types (GI and GII in Fig. 6). Pca45DO-encoding gene in one gene cluster (Fig. 6), The ring-cleavage of gallate gives 4-oxalomesaconate, except in V. eiseniae EF01-2, where the cluster lacks an and thus, the transformation of gallate to pyruvate aldolase-encoding gene. Typically, the organization of and oxaloacetate requires only a 4-oxalomesaconate these gene clusters in Burkholderiales strains follows a hydratase and a 4-carboxy-4-hydroxy-2-oxoadipate aldo- hydratase-aldolase-hydrolase-Pca45DO-dehydrogenase lase as downstream enzymes. Accordingly, all gene clus- gene order, with the insertion of genes of unknown function ters encoding a gallate dioxygenase of type II are devoid in a few cases. This gene organization clearly differs of genes encoding 4-carboxy-2-hydroxymuconic semial- from the organization of Pca meta cleavage gene clusters dehyde dehydrogenase and 2-pyrone-4,6-dicarboxylate in Sphingomonas strains (Fig. 6). Recently, a putative hydrolase (Fig. 6), enzymes necessary only for Pca deg- 4-oxalomesaconate tautomerase required for Pca degra- radation. This supports the notion that these gene clusters dation in Comamonas sp. strain E6 has been reported have evolved for gallate degradation (Nogales et al., (Kamimura et al., 2010). Respective genes are also 2011). Gene clusters encoding a type I gallate dioxyge- observed in most Pca45 meta cleavage gene clusters nase encode only the ring-cleavage activity. In this case, (Fig. 6). a second gene cluster comprising genes encoding, Several oxygenases have been described to transform among functional genes, a 4-oxalomesaconate hydratase lignin-derived compounds to Pca followed by Pca meta and a 4-carboxy-4-hydroxy-2-oxoadipate aldolase, are cleavage (Fig. 1), especially in Comamonadaceae strains located elsewhere in the chromosome. It should be noted (Fig. 2). All Burkholderiales strains harbouring a Pca that genes encoding putative 4-oxalomesaconate tau- meta cleavage pathway also own a gene encoding tomerases are also usually comprised in gallate catabolic the 4-hydroxybenzoate hydroxylase (Phb3H) (Fig. 2), gene clusters; however, their function remains to be although Phb3H-encoding genes are not included in Pca elucidated. meta cleavage gene clusters. An exception to this are the gene clusters of Rhodoferax ferrireducens T118 and one of Catechol ring-cleavage pathways and the gene clusters of the C. testosteroni strains CNB-2 and peripheral routes S44 (Fig. 6), where one cluster comprises genes encoding for Phb3H and Pca45DO, and a second cluster comprising Major work on elucidation of catechol degradation has so genes for the full downstream catabolic pathway. In con- far been performed in Pseudomonas strains (Harwood trast to Phb3H-encoding genes, 3-hydroxybenzoate and Parales, 1996; Jiménez et al., 2002). These organ-

© 2011 Society for Applied Microbiology and Blackwell Publishing Ltd, Environmental Microbiology, 14, 1091–1117 1104 D. Pérez-Pantoja et al. isms typically harbour a catRBCA gene cluster, encoding multivorans and B. ambifaria, a complete catBCA cluster a LysR type regulator (catR), a catechol 1,2-dioxygenase is connected to genes encoding for subunits of BenDO. (Cat12DO) forming muconate (catA), a muconate cycloi- This organization resembles that previously observed in somerase forming muconolactone (catB) and a mucono- the g-proteobacterium Klebsiella pneumoniae 342 (Fouts lactone isomerase forming enol-lactone (catC), which is et al., 2008), and the cat gene organization is also similar recognized as a common intermediate in the degradation to that typically observed in Pseudomonads (Jiménez of catechol and protocatechuate, and then hydrolysed by et al., 2002). Accordingly, the catA gene products of an enol-lactone hydrolase typically not encoded in the cat these type IV clusters are more closely related to gene cluster. The catBCA cluster (cat operon) thus com- g-proteobacterial catA gene products than to type I catA prises all genes exclusively responsible for catechol deg- gene products (Fig. 7), indicating a common origin. radation forming the catechol branch of the 3-oxoadipate All ‘B. cepacia complex’ strains encoding type IV cat pathway (Fig. 1). A complete catechol branch is frequent gene cluster also harbour a complete type I cat gene only in Burkholderiaceae, as nearly all Burkholderia, cluster. For this reason, it can be argued that a second Cupriavidus and Ralstonia genomes analysed so far complete gene cluster may be an environmental burden. possess such genes (Fig. 2). All strains of the ‘B. cepacia Accordingly, different variants of the type IV cluster har- complex’ harbour a catRBAC gene cluster, contrasting the bouring catBA or even only catA genes are observed typical catRBCA order observed in Pseudomonads (Fig. 7). This indicates the presence of a second gene (Fig. 7). These closely related gene clusters (type Ia) are copy of catA to be of a selective advantage. A reason for preceded by genes encoding an AntDO anthranilate this could be the necessity to avoid the accumulation of dioxygenase (Fig. 7), indicating the major role of the cat- highly toxic catechol intermediates (Schweigert et al., echol branch to channel anthranilate formed during the 2001). Incorporation of catA in a benzoate gene cluster degradation of tryptophan into the 3-oxoadipate pathway. would avoid such accumulation, but directly produce the Even though the same gene order is conserved in cat non-toxic cis,cis-muconate, which is reported as an operons of B. phymatum, B. phytofirmans, B. xenovorans inducer of the type I cat gene cluster of Burkholderia sp. and Burkholderia sp. Ch1-1, the respective genes are not TH2 (Suzuki et al., 2002) and of Pseudomonas cat gene located in a cluster with AntDO (Fig. 7, type Ib). However, clusters (Parsek et al., 1992). Besides strains of the both Burkholderia sp. Ch1-1 and B. xenovorans harbour ‘B. cepacia complex’, B. graminis, B. glumae, B. lata, an AntDO localized elsewhere (Fig. 2). Despite the B. phytofirmans, as well as various Burkholderia strains absence of AntDO-encoding genes in some strains (uncharacterized to the species level) and R. pickettii outside the B. cepacia complex, they are also capable of strains harbour gene clusters comprising a catA gene degrading anthranilate, due to the presence of a gene phylogenetically related to those of the type IV cluster. cluster allowing degradation via aminobenzoyl-CoA (Abc). Such genes are absent from B. xenovorans and B. phy- This pathway is typically absent from Burkholderia strains matum (Fig. 7). of the ‘B. cepacia complex’ (Fig. 2). Like Burkholderia The localization of cat genes in a cluster with ben genes strains outside of the B. cepacia complex, all Cupriavidus, is also observed in three of the four sequenced Cupriavi- Ralstonia, four out of five Bordetella spp., as well as dus strains and in two of the three R. solanacearum the majority of members of the Comamonadaceae strains (Fig. 7, type III). They are absent from the family harbour the anthranilate degradative pathway genomes of C. taiwanensis, which is devoid of most path- via aminobenzoyl-CoA, whereas AntDO is restricted to ways involving dihydroxylated intermediates, and also of members of the Burkholderia genus. The only exception R. solanacearum GMI1000. Even though the structure of to this is H. seropedicae SmR1, where an AntDO- these gene clusters is similar to type IV gene clusters encoding gene precedes the muconate cyloisomerase- defined above, it should be noted that these catA gene encoding gene (Fig. 7). Interestingly, in B. phytofirmans, products are phylogenetically more closely related to type B. xenovorans and Burkholderia sp. Ch1-1 the catBAC I catA gene products (Fig. 7), indicating the association of gene cluster is framed by genes encoding a Rieske non- cat genes into ben gene clusters to have occurred at least haem iron oxygenase, where the deduced sequences of twice in evolution. the a-subunits share > 60% of identity with that of IacC Both C. pinatubonensis JMP134 and C. metallidurans gene supposed to be involved in indole 3-acetate degra- CH34 carry a second catA gene copy (Fig. 7, type II). In dation by P. putida 1290 (Leveau and Gerards, 2008). both strains, genes encoding a multicomponent phenol Besides the above mentioned type Ia gene cluster, a monooxygenase (PMO) precede the second copy. Note- subset of strains of the ‘B. cepacia complex’ (Fig. 2) com- worthy, in strains so far described as capable of degrading prises a second cat gene cluster (type IV, Fig. 7), which is phenols via PMO, degradation of the catechol intermedi- typically connected with genes coding for a benzoate ate occurs by extradiol cleavage through catechol dioxygenase (BenDO). As an example, in Burkholderia 2,3-dioxygenases (Cat23DO) (Nordlund and Shingler,

Fig. 7. Maximum likelihood phylogenetic tree for catechol 1,2-dioxygenase (Cat12) and gene clusters in which the respective dioxygenase-encoding genes are found. Branches tagged with dots indicate sequences with conﬁrmed functionality not belonging to Burkholderiales, or belonging to Burkholderiales strains not included in this study, that have been incorporated for additional phylogenetic comparisons. The genes encoding Cat12 are aligned to one another. Dotted arrows indicate the presence of additional genes related to peripheral pathways channelling to catechol. A bidirectional arrow indicates truncated sequences. Other details as indicated in legend of Fig. 5.

1990; Nordlund et al., 1993; Merimaa et al., 2006). In fact, cally of broad substrate speciﬁcity and transform meth- both strains harbour two copies of PMO-encoding genes, ylphenols with high activity (Powlowski and Shingler, with the second copy of the genes localized in a gene 1994; Watanabe et al., 1998). Such activity would channel cluster with a Cat23DO. Multicomponent PMO are typi- methylphenols in the ortho cleavage pathway, which is

© 2011 Society for Applied Microbiology and Blackwell Publishing Ltd, Environmental Microbiology, 14, 1091–1117 1106 D. Pérez-Pantoja et al. usually not suited to mineralize them (Catelani et al., degrade phenol. Noteworthy, such a gene cluster is the 1971; Knackmuss et al., 1976). However, early studies only aromatic catabolic trait observed in Limnobacter sp. on C. pinatubonensis JMP134 have indicated a high MED105 (Fig. 2). Cat12DO activity only after growth with phenol and not Cat23DO of Burkholderiales can be phylogenetically with methylphenols (Pieper et al., 1989), indicating a spe- grouped into three distinct clusters (Fig. 8), with the major- cific regulation that restricts the flux of methylaromatics ity of them being related to Cat23DO CbzE of P. putida into the ortho cleavage pathway. In addition, strain GJ31 (Kaschabek et al., 1998) or TbuE of R. pickettii JMP134 is one of the few strains reported thus far which PK01 (Kukor and Olsen, 1996) involved in chlorobenzene can mineralize methylaromatics via a catechol intradiol or toluene degradation (subfamily I.2.C according to the cleavage pathway due to the presence of a catabolic gene classification by Eltis and Bolin, 1996). However, the pro- module encoding enzymes for 4-methylmuconolactone teins of Limnobacter sp. MED105, and C. necator H16 as transformation (Erb et al., 1998), which is otherwise a well as one of the proteins of M. petroleiphilum PM1 dead-end product for various strains (Catelani et al., belong to a distinct protein cluster more similar to subfam- 1971; Knackmuss et al., 1976). Interestingly, such genes, ily I.2.A proteins, such as XylE of P. putida mt-2 (Fig. 8), which could have been recruited in strain JMP134 for indicating that Cat23DO phylogeny does not follow strain 4-methylphenol metabolism via the ortho cleavage phylogeny. Three further Cat23DO proteins (Fig. 8) are pathway, are absent from the genome of C. metallidurans related to LapB involved in alkylphenol degradation by CH34, but observed on the pHG1 megaplasmid of Pseudomonas sp. strain KL28 (Jeong et al., 2003). C. necator H16 that harbours only a single phenol Interestingly, gene clusters comprising subfamily I.2.C hydroxylase gene cluster containing a Cat23DO gene. Cat23DO highly similar to the ones found in Burkholderi- Out of the 16 Comamonadaceae strains sequenced, ales genomes, have also been reported in previous only D. acidovorans SPH-1, Acidovorax sp. JS42, Aci- studies to be frequently clustered with genes encoding dovorax ebreus TPSY, Alicyclophilus denitrificans BC and aniline dioxygenase. One such case is Delftia sp. AN3 P. naphthalenivorans CJ2 harbour a cat gene cluster (Liu et al., 2002). However, such an organization is not (Figs 2 and 7), indicating such activity is not a common observed in any of the sequenced genomes. The cluster- trait in this family. Similarly, in the Alcaligenaceae, only ing of highly similar meta cleavage pathways with distinct both sequenced Achromobacter species and B. petrii peripheral reactions may indicate that these pathways harbour a catechol ortho cleavage pathway. The genome have been formed recently by the combination of appro- sequence of B. petrii, however, indicates this gene cluster priate pathway gene modules. to be located on a genomic island and thus to have been Four of the gene clusters also contain genes encoding recruited from foreign DNA (Gross et al., 2008). for a benzene/toluene monooxygenase (Leahy et al., The extradiol ring-cleavage of catechol and methylsub- 2003), which should allow the hosts, in concert with stituted catechols is typically catalysed by family I.2 type I phenol monooxygenase, to grow on benzene/toluene extradiol dioxygenases (Eltis and Bolin, 1996). Contrast- via two successive monooxygenations. However, some ing genes encoding Cat12DO, which are observed in phenol monooxygenases, among them toluene 2- nearly all Burkholderiaceae, Cat23DO genes are found monooxygenase of strain G4, sequentially oxidize toluene only in eight Burkholderiaceae, among them three of to 2-methylphenol and further to 3-methylcatechol the four sequenced Cupriavidus strains (Figs 2 and 8). (Newman and Wackett, 1995), even though, as evident Outside the Burkholderiaceae, Cat23DO genes are for phenol monooxygenase of Pseudomonas stutzeri observed in five Comamonadaceae, including A. denitrifi- OX1, phenol seems to be the highly preferred substrate cans BC proposed to grow on benzene via extradiol (Cafaro et al., 2004). Thus, the fact that A. denitrificans cleavage of catechol using oxygen generated during BC is capable of growth on benzene in the absence of a chlorate respiration (Weelink et al., 2008), as well as in benzene/toluene monooxygenase (and a Rieske non- Limnobacter sp. MED105, L. cholodnii SP-6 and M. pe- haem iron oxygenase capable to dioxygenate benzene) troleiphilum PM1. may be due to the broad substrate specificity of the Interestingly, out of the 19 Cat23DO-encoding genes, respective phenol monooxygenase. which are observed in 18 Burkholderiales gene clusters The degradation of catechol and 4-methylcatechol (the gene cluster of Acidovorax sp. JS42 comprises only (and thus of phenol and 4-methylphenol) requires the two distantly related genes encoding Cat23DO), all but presence of the dehydrogenase branch of the meta- one are observed in gene clusters encoding a multicom- cleavage pathway suited for metabolism and specifically ponent PMO (Fig. 8). Such a high abundance of respec- a 2-hydroxymuconic semialdehyde dehydrogenase for tive gene clusters in Burkholderiales is astonishing, rapid dehydrogenation of the ring-cleavage products 2- especially taking into account that these microorganisms hydroxymuconic or 2-hydroxy-4-methylmuconic semialde- are not typically isolated based on their capability to hyde (Harayama et al., 1987), followed by isomerization

© 2011 Society for Applied Microbiology and Blackwell Publishing Ltd, Environmental Microbiology, 14, 1091–1117 01SceyfrApidMcoilg n lcwl ulsigLtd, Publishing Blackwell and Microbiology Applied for Society 2011 © erdto faoai opud in compounds aromatic of Degradation niomna Microbiology Environmental , 14 1091–1117 , ukodrae 1107 Burkholderiales

Fig. 8. Maximum likelihood phylogenetic tree for catechol 2,3-dioxygenase (Cat23) and gene clusters in which the respective dioxygenase-encoding genes are found. Branches tagged with dots indicate sequences with confirmed functionality not belonging to Burkholderiales, or belonging to Burkholderiales strains not included in this study, that have been incorporated for additional phylogenetic comparisons. The genes encoding Cat23 are aligned to one another. The functions of the different types of genes included in the clusters are indicated at the bottom. A bidirectional arrow indicates truncated sequences. Other details as indicated in legend of Fig. 5. 1108 D. Pérez-Pantoja et al. and decarboxylation to yield 2-hydroxypent-2,4-dienoate. genase has been described in P. putida strain U (Arias- In contrast, the degradation of 3-methylcatechol Barrau et al., 2005) as being composed of the hydroxylase (formed from 2-methylphenol) requires the presence of and a small coupling protein, constituting a novel type of a 2-hydroxymuconic semialdehyde hydrolase for rapid two-component hydroxylase, distinct from the classical hydrolysis of the ketonic ring-cleavage product 2-hydroxy- two-component flavoprotein monooxygenases. 6-oxo-2,4-heptadienoate to yield 2-hydroxypent-2, The homogentisate pathway converts homogentisate 4-dienoate (Harayama et al., 1987). This central interme- into fumarate and acetoacetate (Arias-Barrau et al., diate of the hydrolytic and dehydrogenase branch is 2004). The homogentisate dioxygenase (Hge), an further dissembled by a hydratase giving 4-hydroxy-2- enzyme of the cupin superfamily of extradiol dioxygena- oxovalerate, and an aldolase giving pyruvate and acetal- ses (Dunwell et al., 2000) cleaves the aromatic ring dehyde. All respective genes are, for example, present between the acetyl substituent and the proximal hydroxyl in the archetype meta cleavage pathway gene cluster group, yielding maleylacetoacetate. This product is localized on the pWW0 TOL plasmid of P. putida mt-2 isomerized by a GSH-dependent maleylacetoacetate (Harayama and Rekik, 1990), together with an acetalde- isomerase (Suemori et al., 1995) into fumarylacetoac- hyde dehydrogenase-encoding gene. Moreover, typically etate, which in turn is hydrolysed by a fumarylacetoac- a ferredoxin-encoding gene is observed upstream of etate hydrolase to give fumarate and acetoacetate. the gene encoding Cat23DO, and the respective proteins However, in some bacteria maleylacetoacetate can be have been reported to have a reactivating function through directly hydrolysed into maleate and acetoacetate (Craw- reduction of the iron atom in the active site of the enzyme ford, 1976). (Hugo et al., 1998). Similar ferredoxins are observed to In Burkholderiales, the homogentisate pathway is be encoded upstream of the Cat23DO genes in Burkhold- highly represented (Figs 2 and 9) and can be observed in eriales (Fig. 8), indicating the gene product to be of a most sequenced members of the families Alcaligenaceae significant selective advantage. Also typically, both a (except for Taylorella equigenitalis) and Burkholderiaceae 2-hydroxymuconic semialdehyde hydrolase and enzymes (except for B. rhizoxinica, Poynucleobacter spp. and Lim- of the dehydrogenase branch are encoded in meta cleav- nobacter sp.), but is absent from most Comamonadaceae age pathway gene clusters of Burkholderiales, together (with the exception of Polaromonas sp. JS666, where with genes encoding downstream enzymes (Fig. 8). such a pathway is plasmid encoded, and Variovorax Exceptions to this are the gene clusters of C. necator H16, spp.). The catabolic operons (hmg gene clusters) typically which should thus not be capable of growth on comprise genes encoding Hge (hmgA) and fumary- 2-methylphenol via 3-methylcatechol and one of the two lacetoacetate hydrolase (hmgB). Fumarylacetoacetate gene clusters of M. petroleiphilum PM1. As is the case for hydrolases form a monophyletic group and share > 45% B. lata 383, where one of the gene clusters is devoid of of sequence identity and > 50% of sequence identity with genes encoding proteins of the dehydrogenase branch, fumarylacetoacetate hydrolase from P. putida U. Genes also M. petroleiphilum PM1 harbours one complete gene encoding maleylacetoacetate isomerase are typically cluster, thus allowing degradation of differently substituted absent from the hmg gene clusters. This is in contrast to catechols. the organization in other proteobacteria such as P. putida, P. fluorescens, Sinorhizobium meliloti and Silicibacter pomeroyi (Fig. 9), where hmg gene clusters typically com- Homogentisate pathway and peripheral reactions prise a maleylacetoacetate isomerase-encoding gene. Homogentisate is a central metabolite formed during the Different types of hmg gene clusters can be easily aerobic catabolism of aromatic amino acids such as phe- defined. Members of the Burkholderiaceae family usually nylalanine and tyrosine in many bacteria (Arias-Barrau comprise hmgAB genes framed by genes encoding up to et al., 2004). These amino acids are converted to three transporters of the major facilitator superfamily 4-hydroxyphenylpyruvate and subsequently transformed (MFS) where the proteins encoded by genes preceding into homogentisate by a 4-hydroxyphenylpyruvate dioxy- hmgA share significant similarity to the PcaK 4- genase (HppDO) (Fitzpatrick, 2003). This enzyme is a Fe2+ hydroxybenzoate transporter of P. putida PRS2000 dependent, non-haem oxygenase that catalyses the (> 30% sequence identity) (Harwood et al., 1994). A gene conversion of 4-hydroxyphenylpyruvate to homogentisate encoding a flavoprotein monooxygenase typically pre- in a reaction involving decarboxylation, substituent migra- cedes the hmg gene clusters of Burkholderia strains, tion and aromatic oxygenation (Morán, 2005) (Fig. 1). and also in various other Burkholderiales such as Borde- 3-Hydroxyphenylacetate is also channelled into the tella spp., R. pickettii or Polaromonas sp. JS666. In a homogentisate pathway, as described in Pseudomonas phylogenetic analysis, the flavoprotein monooxygenases and Burkholderia species (Arias-Barrau et al., 2005; encoded by Polaromonas sp. JS666 and B. xenovorans Méndez et al., 2011). 3-Hydroxyphenylacetate monooxy- LB400 (BxeA2725) cluster with those encoding

Fig. 9. Maximum likelihood phylogenetic tree for homogentisate dioxygenase (Hge) and gene clusters in which the respective dioxygenase-encoding genes are found. Branches tagged with dots indicate sequences with conﬁrmed functionality not belonging to Burkholderiales that have been incorporated for additional phylogenetic comparisons. The genes encoding Hge are aligned to one another. A bidirectional arrow indicates truncated sequences. Other details as indicated in legend of Fig. 5. © 2011 Society for Applied Microbiology and Blackwell Publishing Ltd, Environmental Microbiology, 14, 1091–1117 1110 D. Pérez-Pantoja et al.

3-hydroxyphenylacetate 6-hydroxylase of P. putida of several polycyclic aromatic compounds such as naph- U (44% and 47% of identity respectively) and can thalene and phenantrene (Lee et al., 2011; Tittabutr be assumed to encode 3-hydroxyphenylacetate 6- et al., 2011). In the gentisate central pathway, gentisate hydroxylases. The other flavoprotein monooxygenases 1,2-dioxygenase (Gen), a member of the cupin super- encoded in hmg gene clusters share > 63% of sequence family, cleaves the aromatic ring between the carboxyl identity but are distantly related to 3-hydroxyphenylacetate substituent and the proximal hydroxyl group to yield 6-hydroxylases (34–44% of sequence identity). Their con- maleylpyruvate. Maleylpyruvate is isomerized by a served inclusion in the hmg gene cluster indicates that they glutathione-dependent maleylpyruvate isomerase to channel substrates to homogentisate; however, their fumarylpyruvate. A fumarylpyruvate hydrolase hydroly- actual function remains to be elucidated. ses fumarylpyruvate to fumarate and pyruvate. Maley- At first glance, the organization of hmg gene clusters lpyruvate can also be converted into pyruvate and outside of members of the Burkholderia genus seems to be maleate by a maleylpyruvate hydrolase in a glutathione- relatively conserved as far as a transcriptional regulator independent way (Crawford and Frick, 1977; Bayly et al., precedes the hmgAB genes. In P. putida U, a divergently 1980). transcribed IclR family regulator regulates hmgAB cata- In contrast to the homogentisate pathway, the gentisate bolic genes (Arias-Barrau et al., 2004). Such an organiza- pathway is present only in a few Burkholderia strains tion seems the exception in Burkholderiales, since in (Fig. 2), namely Burkholderia cenocepacia, B. multi- sequenced members of the Alcaligenaceae family and vorans and B. lata 383 (but interestingly absent from Ralstonia sp. the cluster is preceded by a LysR type closely related strains such as B. ambifaria). It is also regulator-encoding gene, and in Cupriavidus sp. by a TetR found in Burkholderia sp. CGE1001 and 1003 as well as regulator-encoding gene. The gene cluster of Polaromo- in B. xenovorans LB400. The gentisate pathway is nas JS666 comprises a MarR type regulator-encoding widespread in Comamonadaceae and Alcaligenaceae, gene and one of the additional gene clusters of B. xeno- and is present in Achromobacter and some Bordetella vorans LB400 (BxeC0994) a LuxR type regulator- strains (Fig. 2). The gentisate catabolic gene clusters in encoding gene. Typically, only one hmg gene cluster is Burkholderiales (Fig. 10) typically contain all three cata- present, with the exception of Burkholderia sp. CCGE1002 bolic genes, an organization also observed in other bac- (two clusters), Burkholderia sp. Ch1-1 (two clusters and teria. Gentisate dioxygenases of Burkholderiales are quite one containing a truncated Hge-encoding gene) and diverse and share only down to > 30% of sequence iden- B. xenovorans LB400 containing an impressive four such tity. As it is shown in Fig. 10, they can be grouped into at gene clusters. Recently it has been reported that B. xeno- least five distinct clusters, where the identity among vorans LB400 degrades 3-hydroxypenylacetate and members of each cluster is > 50%. A similar diversity 4-hydroxyphenylacetate through both homogentisate and is seen in fumarylpyruvate hydrolases, which may also homoprotocatechuate pathways, suggesting a very be sorted into five distinct groups. Most putative complex regulation of these four Hge genes (Méndez fumarylpyruvate hydrolases encoded in group I gentisate et al., 2011). gene clusters are, among validly described fumarylpyru- Among pathways channelling to homogentisate, a vate hydrolases, most closely related to that of K. pneu- homogentisate forming 4-hydroxyphenylpyruvate dioxy- moniae M5a1 (Jones and Cooper, 1990), whereas genase is present in all strains harbouring an hmg fumarylpyruvate hydrolases Bpro_0982 and Lcho_0798 gene cluster but it is always located elsewhere on the of Polaromonas sp. JS666 and L. cholodnii SP-6 are most genome. Also, a tyrosine forming phenylalanine hydroxy- closely related to that of P. putida AK5 (ACO92376). Inter- lase is present in nearly all such strains except Achro- estingly, fumarylpyruvate hydrolases related to that of mobacter and Bordetella strains. Interestingly, such Ralstonia sp. U2 are not only encoded by most gentisate genes are also observed in strains where the hmg gene clusters of group III but also by gentisate gene pathway is absent, such as Acidovorax or Comamonas clusters of Comamonas strains (group V). While all of strains. the above mentioned fumarylpyruvate hydrolases share > 50% of sequence identity, hydrolase enzymes encoded by group II and IV gene clusters are only distantly related Gentisate pathway and peripheral reactions to archetype fumarylpyruvate hydrolases, sharing with Gentisate (2,5-dihydroxybenzoate) is a central metabo- them < 30% of sequence identity. Interestingly, such a lite formed during the catabolism of diverse aromatic novel fumarylpyruvate hydrolase is also encoded in one of compounds such as salicylate (2-hydroxybenzoate), the gentisate gene clusters of P. naphthalenivorans CJ2 3-hydroxybenzoate and some phenolic compounds (Pnap2454), even though the encoded gentisate dioxyge- including m-cresol, 2,5-xylenol and 3,5-xylenol (Poh and nase shares significant sequence identity with that of Bayly, 1980), and even has a key role in the degradation Ralstonia sp. U2.

Fig. 10. Maximum likelihood phylogenetic tree for gentisate dioxygenase (Gen) and gene clusters in which the respective dioxygenase-encoding genes are found. Branches tagged with dots 1111 indicate sequences with conﬁrmed functionality not belonging to Burkholderiales, or belonging to Burkholderiales strains not included in this study, that have been incorporated for additional phylogenetic comparisons. The genes encoding Gen are aligned to one another. Dotted arrows indicate the presence of additional genes related to peripheral pathways channelling to gentisate. Other details as indicated in legend of Fig. 5. 1112 D. Pérez-Pantoja et al.

Enzyme diversity also extends to maleylpyruvate this may prevent the accumulation of toxic intracellular isomerases encoded by gentisate gene clusters. In levels of gentisate due its higher affinity for this substrate contrast to homogentisate gene clusters where an (Lee et al., 2011). isomerase-encoding gene is usually absent, most genti- In three of the group IV catabolic gene clusters (Fig. 10), sate gene clusters comprise such a gene. Gentisate gene the Gen-encoding gene is preceded by genes encoding clusters I, III and V typically comprise genes encoding flavoprotein monooxygenases with significant similarity to a glutathione-dependent maleylpyruvate isomerase of 2-aminobenzoyl-CoA monooxygenase/reductase from the class zeta subfamily of glutathione S-transferases Azoarcus evansii which transforms aminobenzoyl-CoA with high similarity to maleylpyruvate isomerases of vali- into 2-amino-5-oxo-cyclohex-1-ene-1-carbonyl-CoA (60– dated function (Jones and Cooper, 1990; Marsh et al., 62% of sequence identity) but also with similarity to 2008). The phylogeny follows that described above for salicylyl-CoA monooxygenase from Streptomyces WA46 fumarylpyruvate hydrolases. Cognate genes are absent (40–42% of sequence identity) (Ishiyama et al., 2004). As from type II gentisate gene clusters, while type IV clusters the gene products can be supposed to channel to genti- typically comprise a gene encoding a putative maleate sate, it may be reasoned that they encode salicylyl- cis-trans isomerase with high similarity (> 60% sequence CoA monooxygenases. Nevertheless it remains to be identity) to maleate isomerase from Alcaligenes faecalis determined if genes encoding enzymes with similarity (Hatakeyama et al., 1997). It could be reasoned that a to indolepyruvate ferredoxin oxidoreductase (Fig. 10) glutathione-independent maleate isomerase is recruited actually encode enzymes that can catalyse the oxidative for gentisate degradation in these gene clusters. decarboxylation of aromatic compounds such as 2- The function of the majority of gentisate gene clusters hydroxyphenylglyoxylate. Overall, the analysis of genti- is obviously the degradation of 3-hydroxybenzoate, sate catabolic gene clusters indicates a high diversity of as respective genes encoding proteins with > 45% of gene organization and a high rate of gene rearrangement. sequence identity with 3-hydroxybenzoate 6-hydroxylase Pnap3144 from P. naphthalenivorans CJ2 (Park et al., Research needs 2007) are also found here. However, genes of type IV and V clusters usually do not comprise 3-hydroxybenzoate In general terms, this analysis highlights the amazing 6-hydroxylase-encoding genes. Whether the genes catabolic potential of Burkholderiales towards aromatic observed in Comamonadaceae, as encoding proteins with compounds. One evident aspect of these traits is the 33–38% identity with 3-hydroxybenzoate 6-hydroxylase potential utilization in environmental biotechnology for the Pnap3144 are functional remains to be elucidated. A biodegradation of aromatic pollutants. This topic has been second function of the gentisate gene cluster is in the addressed in recent reviews (O’Sullivan and Mahenthiral- degradation of salicylate after 5-hydroxylation through a ingam, 2005; Denef, 2007) and is not discussed here. multicomponent Rieske non-haem iron oxygenase. Six of However, some emerging topics could attract additional the gentisate catabolic gene clusters identified here com- interest on the biodegradation of aromatic compounds by prise genes encoding a salicylate 5-hydroxylase (Fig. 10). Burkholderiales. Additionally, genes encoding salicylate 5-hydroxylases, One of these aspects is the possible role of aromatic indicating the capability of the host to grow on salicylate via catabolic genes for pathogenicity of some members of the gentisate as an intermediate, are found elsewhere in the ‘B. cepacia complex’. In B. cenocepacia, the requirement genomes of both V. paradoxus strains, both Achromo- of a functional phenylacetic acid catabolic pathway for full bacter strains, three of four Acidovorax strains, two Cupria- pathogenicity in the Caenorhabditis elegans host model vidus strains, as well as in B. petrii, indicating such has recently been shown (Law et al., 2008) as well as the capability to be widespread in Burkholderiales. The turn- induction of this route during growth in synthetic cystic over of salicylate through gentisate has been usually fibrosis medium (Hamlin et al., 2009; Yudistira et al., related to the degradation of polycyclic compounds in 2011). It is pertinent to remember that the phenylacetate Burkholderiales (Lee et al., 2011; Tittabutr et al., 2011); catabolic route is one of the most widespread catabolic however, no Phn and Nag-like dioxygenases involved in pathways in Burkholderiales (Fig. 2) and it has been sug- the degradation of phenanthrene and naphthalene, gested that this compound or a derived catabolite may act respectively, seem to be encoded in the Burkholderiales as a signal that triggers multiple physiological processes genomes analysed with the exception of P. naphthalen- (Patrauchan et al., 2011), probably indicating a key role in ivorans CJ2 isolated by its ability to degrade naphthalene the ecological fitness of this group in multiple environ- (Jeon et al., 2006). In this bacterium, one (Pnap_3145) of ments. Further research is needed to understand the spe- three Gen-encoding genes included in type III catabolic cific role of this pathway in host–pathogen interactions. clusters (Fig. 10), has been shown to have a key role in An additional interesting aspect of aromatic catabolism naphthalene degradation and it has been proposed that in Burkholderiales is its potential role in the degradation of

© 2011 Society for Applied Microbiology and Blackwell Publishing Ltd, Environmental Microbiology, 14, 1091–1117 Degradation of aromatic compounds in Burkholderiales 1113 antibiotics. In a recent study, hundreds of soil strains were Bais, H.P., Weir, T.L., Perry, L.G., Gilroy, S., and Vivanco, isolated for their ability to grow on antibiotics as a sole J.M. (2006) The role of root exudates in rhizosphere inter- carbon source. This study included 18 different antibiotics actions with plants and other organisms. Annu Rev Plant Biol 57: 233–266. representing eight major classes of natural and synthetic Balandreau, J., and Mavingui, P. (2006) Beneficial interac- origin, the structures of various of which comprised an tions of Burkholderia spp. with plants. In Burkholderia: aromatic moiety (Dantas et al., 2008). Phylogenetic pro- Molecular Microbiology and Genomics. Coeyne, T., and filing of the isolates revealed a diverse set of species Vandamme, P. (eds). Wymondham, Norfolk: Horizon Bio- belonging to 11 different orders; however, 41% of the science, pp. 129–151. species isolated belonged to the Burkholderiales, fol- Bayly, R.C., Chapman, P.J., Dagley, S., and Di Berardino, D. lowed by Pseudomonadales (24%) (Dantas et al., 2008). (1980) Purification and some properties of maleylpyruvate hydrolase and fumarylpyruvate hydrolase from Pseudomo- It seems probable that the importance of Burkholderiales nas alcaligenes. J Bacteriol 143: 70–77. in antibiotics degradation is related to the broad catabolic van Berkel, W.J., Kamerbeek, N.M., and Fraaije, M.W. (2006) potential for aromatic degradation. In this sense, it should Flavoprotein monooxygenases, a diverse class of oxidative be mentioned that during the mining of catabolic oxyge- biocatalysts. J Biotechnol 124: 670–689. nases performed for the present analysis, a huge number Cafaro, V., Izzo, V., Scognamiglio, R., Notomista, E., of genes putatively encoding aromatic oxygenases Capasso, P., Casbarra, A., et al. (2004) Phenol hydroxy- remain with unassigned substrates/functions, constituting lase and toluene/o-xylene monooxygenase from Pseudomonas stutzeri OX1: interplay between two candidates to be tested in search of the genes respon- enzymes. Appl Environ Microbiol 70: 2211–2219. sible for antibiotic degradation, among other possibilities. Catelani, D., Fiecchi, A., and Galli, E. (1971) Dextro-gamma- carboxymethyl-gamma-methyl-delta-alpha-butenolide. A 1,2-ring-fission product of 4-methylcatechol by Pseudomo- Acknowledgements nas desmolyticum. Biochem J 121: 89–92. Support provided by the research programme FONDAP 1501 Chain, P.S., Denef, V.J., Konstantinidis, K.T., Vergez, L.M., 0001, programme 7, the Millennium Scientific Initiative Agullo, L., Reyes, V.L., et al. (2006) Burkholderia xeno- through Grants P/04-007-F and P06-009-F; FONDECYT vorans LB400 harbors a multi-replicon, 9.73-Mbp genome Grants 1070343, 1070507, 7080148, 7090079, USM Grants shaped for versatility. Proc Natl Acad Sci USA 103: 15280– 13110 and 130948; BMBF-CONICYT 2007-2009 and PCBT- 15287. Network (R-12). D.P.-P., R.A.D. and MC acknowledge Chang, H.K., and Zylstra, G.J. (1998) Novel organization of CONICYT fellowships. D.H.P. acknowledges support by the the genes for phthalate degradation from Burkholderia projects BACSIN and MAGICPAH from the European Com- cepacia DBO1. J Bacteriol 180: 6529–6537. mission. Authors acknowledge Verónica Morgante for valu- Chang, H.K., and Zylstra, G.J. (2008) Examination and able contributions to composition of Fig. 2 and Max Chavarría expansion of the substrate range of m-hydroxybenzoate for helpful support in statistical analysis. hydroxylase. Biochem Biophys Res Commun 371: 149– 153. Chiarini, L., Bevivino, A., Dalmastri, C., Tabacchioni, S., and References Visca, P. (2006) Burkholderia cepacia complex species: health hazards and biotechnological potential. Trends Abascal, F., Zardoya, R., and Posada, D. (2005) ProtTest: Microbiol 14: 277–286. selection of best-fit models of protein evolution. Bioinfor- Crawford, R.L. (1976) Degradation of homogentisate by matics 21: 2104–2105. strains of Bacillus and Moraxella. Can J Microbiol 22: 276– Altschul, S.F., Madden, T.L., Schaffer, A.A., Zhang, J., Zhang, 280. Z., Miller, W., and Lipman, D.J. (1997) Gapped BLAST and Crawford, R.L., and Frick, T.D. (1977) Rapid spectro- PSI-BLAST: a new generation of protein database search photometric differentiation between glutathione- programs. Nucleic Acids Res 25: 3389–3402. dependent and glutathione-independent gentisate and Amadou, C., Pascal, G., Mangenot, S., Glew, M., Bontemps, homogentisate pathways. Appl Environ Microbiol 34: 170– C., Capela, D., et al. (2008) Genome sequence of the 174. beta-rhizobium Cupriavidus taiwanensis and comparative Cupples, A.M., and Sims, G.K. (2007) Identification of in situ genomics of rhizobia. Genome Res 18: 1472–1483. 2,4-dichlorophenoxyacetic acid-degrading soil microorgan- Arias-Barrau, E., Olivera, E.R., Luengo, J.M., Fernández, C., isms using DNA-stable isotope probing. Soil Biol Biochem Galán, B., García, J.L., et al. (2004) The homogenti- 39: 232–238. sate pathway: a central catabolic pathway involved Dantas, G., Sommer, M.O., Oluwasegun, R.D., and Church, in the degradation of L-phenylalanine, L-tyrosine, and G.M. (2008) Bacteria subsisting on antibiotics. Science 3-hydroxyphenylacetate in Pseudomonas putida. 320: 100–103. J Bacteriol 186: 5062–5077. Denef, V.J. (2007) Biodegradation of organic anthropogenic Arias-Barrau, E., Sandoval, A., Naharro, G., Olivera, E.R., pollutants by Burkholderia species. In Burkholderia: and Luengo, J.M. (2005) A two-component hydroxylase Molecular Microbiology and Genomics. Coenye, T., and involved in the assimilation of 3-hydroxyphenyl acetate in Vandamme, P. (eds). Wymondham, Norfolk: Horizon Bio- Pseudomonas putida. J Biol Chem 280: 26435–26447. science, pp. 177–201.

Dunwell, J.M., Khuri, S., and Gane, P.J. (2000) Microbial Harwood, C.S., and Parales, R.E. (1996) The beta- relatives of the seed storage proteins of higher plants: ketoadipate pathway and the biology of self-identity. Annu conservation of structure and diversiﬁcation of function Rev Microbiol 50: 553–590. during evolution of the cupin superfamily. Microbiol Mol Harwood, C.S., Nichols, N.N., Kim, M.K., Ditty, J.L., and Biol Rev 64: 153–179. Parales, R.E. (1994) Identiﬁcation of the pcaRKF gene Edgar, R.C. (2004) MUSCLE: a multiple sequence alignment cluster from Pseudomonas putida: involvement in chemo- method with reduced time and space complexity. BMC taxis, biodegradation, and transport of 4-hydroxybenzoate. Bioinformatics 5: 113. J Bacteriol 176: 6479–6488. Elsemore, D.A., and Ornston, L.N. (1994) The pca–pob Hatakeyama, K., Asai, Y., Uchida, Y., Kobayashi, M., supraoperonic cluster of Acinetobacter calcoaceticus con- Terasawa, M., and Yukawa, H. (1997) Gene cloning and tains quiA, the structural gene for quinate-shikimate dehy- characterization of maleate cis-trans isomerase from drogenase. J Bacteriol 176: 7659–7666. Alcaligenes faecalis. Biochem Biophys Res Commun 239: Eltis, L.D., and Bolin, J.T. (1996) Evolutionary relationships 74–79. among extradiol dioxygenases. J Bacteriol 178: 5930– Hugo, N., Armengaud, J., Gaillard, J., Timmis, K.N., and 5937. Jouanneau, Y. (1998) A novel [2Fe–2S] ferredoxin from Erb, R.W., Timmis, K.N., and Pieper, D.H. (1998) Character- Pseudomonas putida mt-2 promotes the reductive reacti- ization of a gene cluster from Ralstonia eutropha JMP134 vation of catechol 2,3-dioxygenase. J Biol Chem 273: encoding metabolism of 4-methylmuconolactone. Gene 9622–9629. 206: 53–62. Ishiyama, D., Vujaklija, D., and Davies, J. (2004) Novel Fitzpatrick, P.F. (2003) Mechanism of aromatic amino acid pathway of salicylate degradation by Streptomyces sp. hydroxylation. Biochemistry 42: 14083–14091. strain WA46. Appl Environ Microbiol 70: 1297–1306. Fouts, D.E., Tyler, H.L., DeBoy, R.T., Daugherty, S., Ren, Q., Jeon, C.O., Park, W., Padmanabhan, P., DeRito, C., Snape, Badger, J.H., et al. (2008) Complete genome sequence of J.R., and Madsen, E.L. (2003) Discovery of a bacterium,

the N2-fixing broad host range endophyte Klebsiella pneu- with distinctive dioxygenase, that is responsible for in situ moniae 342 and virulence predictions verified in mice. biodegradation in contaminated sediment. Proc Natl Acad PLoS Genet 4: e1000141. Sci USA 100: 13591–13596. Fukuhara, Y., Inakazu, K., Kodama, N., Kamimura, N., Kasai, Jeon, C.O., Park, M., Ro, H.S., Park, W., and Madsen, E.L. D., Katayama, Y., et al. (2010) Characterization of the (2006) The naphthalene catabolic (nag) genes of isophthalate degradation genes of Comamonas sp. strain Polaromonas naphthalenivorans CJ2: evolutionary impli- E6. Appl Environ Microbiol 76: 519–527. cations for two gene clusters and novel regulatory control. Garrity, G.M., Bell, J.A., and Lilburn, T. (2005) Order I. Appl Environ Microbiol 72: 1086–1095. Burkholderiales ord. nov. In Bergey’s Manual of Systematic Jeong, J.J., Kim, J.H., Kim, C.K., Hwang, I., and Lee, K. Bacteriology. Brenner, D.J., Krieg, N.R., Staley, J.T., and (2003) 3- and 4-alkylphenol degradation pathway in Garrity, G.M. (eds). New York, USA: Springer, p. 575. Pseudomonas sp. strain KL28: genetic organization of the Gibson, D.T., and Parales, R.E. (2000) Aromatic hydrocarbon lap gene cluster and substrate specificities of phenol dioxygenases in environmental biotechnology. Curr Opin hydroxylase and catechol 2,3-dioxygenase. Microbiology Biotechnol 11: 236–243. 149: 3265–3277. Gross, R., Guzmán, C.A., Sebaihia, M., dos Santos, V.A., Jiménez, J.I., Miñambres, B., Garcia, J.L., and Díaz, E. Pieper, D.H., Koebnik, R., et al. (2008) The missing link: (2002) Genomic analysis of the aromatic catabolic path- Bordetella petrii is endowed with both the metabolic versa- ways from Pseudomonas putida KT2440. Environ Micro- tility of environmental bacteria and virulence traits of patho- biol 4: 824–841. genic Bordetellae. BMC Genomics 9: 449. Jones, D.C., and Cooper, R.A. (1990) Catabolism of Guindon, S., Dufayard, J.F., Lefort, V., Anisimova, M., 3-hydroxybenzoate by the gentisate pathway in Klebsiella Hordijk, W., and Gascuel, O. (2010) New algorithms and pneumoniae M5a1. Arch Microbiol 154: 489–495. methods to estimate maximum-likelihood phylogenies: Kamimura, N., Aoyama, T., Yoshida, R., Takahashi, K., Kasai, assessing the performance of PhyML 3.0. Syst Biol 59: D., Abe, T., et al. (2010) Characterization of the protocat- 307–321. echuate 4,5-cleavage pathway operon in Comamonas sp. Hamlin, J.N., Bloodworth, R.A., and Cardona, S.T. (2009) strain E6 and discovery of a novel pathway gene. Appl Regulation of phenylacetic acid degradation genes of Environ Microbiol 76: 8093–8101. Burkholderia cenocepacia K56-2. BMC Microbiol 9: Kasai, D., Masai, E., Miyauchi, K., Katayama, Y., and Fukuda, 222. M. (2005) Characterization of the gallate dioxygenase Harayama, S., and Rekik, M. (1990) The meta cleavage gene: three distinct ring cleavage dioxygenases are operon of TOL degradative plasmid pWW0 comprises 13 involved in syringate degradation by Sphingomonas pauci- genes. Mol Gen Genet 221: 113–120. mobilis SYK-6. J Bacteriol 187: 5067–5074. Harayama, S., Mermod, N., Rekik, M., Lehrbach, P.R., and Kaschabek, S.R., Kasberg, T., Muller, D., Mars, A.E., Timmis, K.N. (1987) Roles of the divergent branches of the Janssen, D.B., and Reineke, W. (1998) Degradation of meta-cleavage pathway in the degradation of benzoate chloroaromatics: purification and characterization of a and substituted benzoates. J Bacteriol 169: 558–564. novel type of chlorocatechol 2,3-dioxygenase of Harrison, P.W., Lower, R.P., Kim, N.K., and Young, J.P. Pseudomonas putida GJ31. J Bacteriol 180: 296–302. (2010) Introducing the bacterial ‘chromid’: not a chromo- Kataoka, R., and Futai, K. (2009) A new mycorrhizal helper some, not a plasmid. Trends Microbiol 18: 141–148. bacterium, Ralstonia species, in the ectomycorrhizal sym-

biosis between Pinus thunbergii and Suillus granulatus. new subgroup within class III of extradiol dioxygenases. Biol Fertil Soils 45: 315–320. Arch Microbiol 183: 130–139. Knackmuss, H.-J., Hellwig, M., Lackner, H., and Otting, W. Manefield, M., Griffiths, R.I., Leigh, M.B., Fisher, R., and (1976) Cometabolism of 3-methylbenzoate and methylcat- Whiteley, A.S. (2005) Functional and compositional com- echols by a 3-chlorobenzoate utilizing Pseudomonas: parison of two activated sludge communities remediating accumulation of (+)-2,5-dihydro-4-methyl- and (+)-2,5- coking effluent. Environ Microbiol 7: 715–722. dihydro-2-methyl-5-oxo-furan-2-acetic acid. Eur J Appl Marsh, M., Shoemark, D.K., Jacob, A., Robinson, C., Cahill, Microbiol 2: 267–276. B., Zhou, N.Y., et al. (2008) Structure of bacterial Kukor, J.J., and Olsen, R.H. (1996) Catechol 2,3- glutathione-S-transferase maleyl pyruvate isomerase and dioxygenases functional in oxygen-limited (hypoxic) envi- implications for mechanism of isomerisation. J Mol Biol ronments. Appl Environ Microbiol 62: 1728–1740. 384: 165–177. Lackner, G., Moebius, N., Partida-Martínez, L., and Mattes, T.E., Alexander, A.K., Richardson, P.M., Munk, A.C., Hertweck, C. (2011) Complete genome sequence of Han, C.S., Stothard, P., and Coleman, N.V. (2008) The Burkholderia rhizoxinica, an endosymbiont of Rhizopus genome of Polaromonas sp. strain JS666: insights into the microsporus. J Bacteriol 193: 783–784. evolution of a hydrocarbon- and xenobiotic-degrading bac- Langille, M.G., and Brinkman, F.S. (2009) IslandViewer: an terium, and features of relevance to biotechnology. Appl integrated interface for computational identification and Environ Microbiol 74: 6405–6416. visualization of genomic islands. Bioinformatics 25: 664– Méndez, V., Agulló, L., González, M., and Seeger, M. (2011) 665. The homogentisate and homoprotocatechuate central Law, R.J., Hamlin, J.N., Sivro, A., McCorrister, S.J., pathways are involved in 3- and 4-hydroxyphenylacetate Cardama, G.A., and Cardona, S.T. (2008) A functional degradation by Burkholderia xenovorans LB400. PLoS phenylacetic acid catabolic pathway is required for full ONE 6: e17583. pathogenicity of Burkholderia cenocepacia in the Cae- Merimaa, M., Heinaru, E., Liivak, M., Vedler, E., and Heinaru, norhabditis elegans host model. J Bacteriol 190: 7209– A. (2006) Grouping of phenol hydroxylase and catechol 7218. 2,3-dioxygenase genes among phenol- and p-cresol- Leahy, J.G., Batchelor, P.J., and Morcomb, S.M. (2003) Evo- degrading Pseudomonas species and biotypes. Arch lution of the soluble diiron monooxygenases. FEMS Micro- Microbiol 186: 287–296. biol Rev 27: 449–479. Morán, G.R. (2005) 4-Hydroxyphenylpyruvate dioxygenase. Lee, H.J., Kim, J.M., Lee, S.H., Park, M., Madsen, E.L., and Arch Biochem Biophys 433: 117–128. Jeon, C.O. (2011) Characterization of gentisate-1,2- Nakatsu, C.H., and Wyndham, R.C. (1993) Cloning and Dioxygenase in the third naphthalene catabolic gene expression of the transposable chlorobenzoate-3,4- cluster of Polaromonas naphthalenivorans CJ2: role in dioxygenase genes of Alcaligenes sp. strain BR60. Appl naphthalene degradation. Microbiology 157: 2891– Environ Microbiol 59: 3625–3633. 2903. Newman, L.M., and Wackett, L.P. (1995) Purification and Leveau, J.H., and Gerards, S. (2008) Discovery of a bacterial characterization of toluene 2-monooxygenase from gene cluster for catabolism of the plant hormone indole Burkholderia cepacia G4. Biochemistry 34: 14066– 3-acetic acid. FEMS Microbiol Ecol 65: 238–250. 14076. Liou, J.S.C., DeRito, C.M., and Madsen, E.L. (2008) Field- Noda, Y., Nishikawa, S., Shiozuka, K., Kadokura, H., Naka- based and laboratory stable isotope probing surveys of the jima, H., Yoda, K., et al. (1990) Molecular cloning of the identities of both aerobic and anaerobic benzene- protocatechuate 4,5-dioxygenase genes of Pseudomonas metabolizing microorganisms in freshwater sediment. paucimobilis. J Bacteriol 172: 2704–2709. Environ Microbiol 10: 1964–1977. Nogales, J., Canales, A., Jiménez-Barbero, J., García, J.L., Liu, Z., Yang, H., Huang, Z., Zhou, P., and Liu, S.J. (2002) and Díaz, E. (2005) Molecular characterization of the Degradation of aniline by newly isolated, extremely aniline- gallate dioxygenase from Pseudomonas putida KT2440. tolerant Delftia sp. AN3. Appl Microbiol Biotechnol 58: 679– The prototype of a new subgroup of extradiol dioxygena- 682. ses. J Biol Chem 280: 35382–35390. Lykidis, A., Pérez-Pantoja, D., Ledger, T., Mavromatis, K., Nogales, J., Canales, A., Jiménez-Barbero, J., Serra, B., Anderson, I.J., Ivanova, N.N., et al. (2010) The complete Pingarrón, J.M., García, J.L., and Díaz, E. (2011) Unravel- multipartite genome sequence of Cupriavidus necator ling the gallic acid degradation pathway in bacteria: the gal JMP134, a versatile pollutant degrader. PLoS ONE 5: cluster from Pseudomonas putida. Mol Microbiol 79: 359– e9729. 374. Madsen, E.L. (2006) The use of stable isotope probing tech- Nordlund, I., and Shingler, V. (1990) Nucleotide sequences of niques in bioreactor and field studies on bioremediation. the meta-cleavage pathway enzymes 2-hydroxymuconic Curr Opin Biotechnol 17: 92–97. semialdehyde dehydrogenase and 2-hydroxymuconic Mahmood, S., Paton, G.I., and Prosser, J.I. (2005) semialdehyde hydrolase from Pseudomonas CF600. Cultivation-independent in situ molecular analysis of bac- Biochim Biophys Acta 1049: 227–330. teria involved in degradation of pentachlorophenol in soil. Nordlund, I., Powlowski, J., Hagström, A., and Shingler, V. Environ Microbiol 7: 1349–1360. (1993) Conservation of regulatory and structural genes for Mampel, J., Providenti, M.A., and Cook, A.M. (2005) Proto- a multi-component phenol hydroxylase within phenol- catechuate 4,5-dioxygenase from Comamonas test- catabolizing bacteria that utilize a meta-cleavage pathway. osteroni T-2: biochemical and molecular properties of a J Gen Microbiol 139: 2695–2703.

O’Sullivan, L.A., and Mahenthiralingam, E. (2005) Biotechno- Providenti, M.A., Shaye, R.E., Lynes, K.D., McKenna, N.T., logical potential within the genus Burkholderia. Lett Appl O’Brien, J.M., Rosolen, S., et al. (2006b) The locus coding Microbiol 41: 8–11. for the 3-nitrobenzoate dioxygenase of Comamonas sp. Park, M., Jeon, Y., Jang, H.H., Ro, H.S., Park, W., Madsen, strain JS46 is ﬂanked by IS1071 elements and is subject to E.L., and Jeon, C.O. (2007) Molecular and biochemical deletion and inversion events. Appl Environ Microbiol 72: characterization of 3-hydroxybenzoate 6-hydroxylase from 2651–2660. Polaromonas naphthalenivorans CJ2. Appl Environ Micro- Pumphrey, G.M., and Madsen, E.L. (2008) Field-based biol 73: 5146–5152. stable isotope probing reveals the identities of benzoic Parsek, M.R., Shinabarger, D.L., Rothmel, R.K., and Chakra- acid-metabolizing microorganisms and their in situ growth barty, A.M. (1992) Roles of CatR and cis,cis-muconate in in agricultural soil. Appl Environ Microbiol 74: 4111– activation of the catBC operon, which is involved in ben- 4118. zoate degradation in Pseudomonas putida. J Bacteriol Sasoh, M., Masai, E., Ishibashi, S., Hara, H., Kamimura, N., 174: 7798–7806. Miyauchi, K., and Fukuda, M. (2006) Characterization of Patrauchan, M.A., Parnell, J.J., McLeod, M.P., Florizone, C., the terephthalate degradation genes of Comamonas sp. Tiedje, J.M., and Eltis, L.D. (2011) Genomic analysis of the strain E6. Appl Environ Microbiol 72: 1825–1832. phenylacetyl-CoA pathway in Burkholderia xenovorans Sato, Y., Nishihara, H., Yoshida, M., Watanabe, M., Rondal, LB400. Arch Microbiol 193: 641–650. J.D., Concepcion, R.N., and Ohta, H. (2006) Cupriavidus Pérez-Pantoja, D., De la Iglesia, R., Pieper, D.H., and pinatubonensis sp. nov. and Cupriavidus laharis sp. nov., González, B. (2008) Metabolic reconstruction of aromatic novel hydrogen-oxidizing, facultatively chemolithotrophic compounds degradation from the genome of the amazing bacteria isolated from volcanic mudﬂow deposits from Mt. pollutant-degrading bacterium Cupriavidus necator Pinatubo in the Philippines. Int J Syst Evol Microbiol 56: JMP134. FEMS Microbiol Rev 32: 736–794. 973–978. Pérez-Pantoja, D., González, B., and Pieper, D.H. (2010a) Schweigert, N., Zehnder, A.J., and Eggen, R.I. (2001) Chemi- Aerobic degradation of aromatic hydrocarbons. In Hand- cal properties of catechols and their molecular modes of book of Hydrocarbon and Lipid Microbiology. Timmis, K.N. toxic action in cells, from microorganisms to mammals. (ed.). Berlin, Germany: Springer-Verlag Berlin Heidelberg, Environ Microbiol 3: 81–91. pp. 799–837. Singleton, D.R., Powell, S.N., Sangaiah, R., Gold, A., Ball, Pérez-Pantoja, D., Donoso, R., Junca, H., González, B., and L.M., and Aitken, M.D. (2005) Stable-isotope probing of Pieper, D.H. (2010b) Phylogenomics of aerobic bacterial bacteria capable of degrading salicylate, naphthalene, or degradation of aromatics. In Handbook of Hydrocarbon phenanthrene in a bioreactor treating contaminated soil. and Lipid Microbiology. Timmis, K.N. (ed.). Berlin, Appl Environ Microbiol 71: 1202–1209. Germany: Springer-Verlag Berlin Heidelberg, pp. 1355– Suemori, A., Nakajima, K., Kurane, R., and Nakamura, Y. 1397. (1995) o-, m- and p-hydroxybenzoate degradative path- Pieper, D.H., Engesser, K.H., and Knackmuss, H.-J. (1989) ways in Rhodococcus erythropolis. FEMS Microbiol Lett Regulation of catabolic pathways of phenoxyacetic acids 125: 31–35. and phenols in Alcaligenes eutrophus JMP134. Arch Sun, W., Xie, S., Luo, C., and Cupples, A.M. (2010) Direct link Microbiol 151: 365–371. between toluene degradation in contaminated-site micro- Pieper, D.H., González, B., Cámara, B., Pérez-Pantoja, D., cosms and a Polaromonas strain. Appl Environ Microbiol and Reineke, W. (2010) Aerobic degradation of chloroaro- 76: 956–959. matics. In Handbook of Hydrocarbon and Lipid Microbiol- Suzuki, K., Ichimura, A., Ogawa, N., Hasebe, A., and ogy. Timmis, K.N. (ed.). Berlin, Germany: Springer-Verlag Miyashita, K. (2002) Differential expression of two catechol Berlin Heidelberg, pp. 839–864. 1,2-dioxygenases in Burkholderia sp. strain TH2. Poh, C.L., and Bayly, R.C. (1980) Evidence for isofunctional J Bacteriol 184: 5714–5722. enzymes used in m-cresol and 2,5-xylenol degradation via Tamura, K., Peterson, D., Peterson, N., Stecher, G., Nei, M., the gentisate pathway in Pseudomonas alcaligenes. and Kumar, S. (2011) MEGA5: molecular evolutionary J Bacteriol 143: 59–69. genetics analysis using maximum likelihood, evolutionary Posada, D., and Buckley, T.R. (2004) Model selection and distance, and maximum parsimony methods. Mol Biol Evol model averaging in phylogenetics: advantages of the AIC 28: 2731–2739. and Bayesian approaches over likelihood ratio tests. Syst Tillmann, S., Strompl, C., Timmis, K.N., and Abraham, W.R. Biol 53: 793–808. (2005) Stable isotope probing reveals the dominant role of Powlowski, J., and Shingler, V. (1994) Genetics and bio- Burkholderia species in aerobic degradation of PCBs. chemistry of phenol degradation by Pseudomonas sp. FEMS Microbiol Ecol 52: 207–217. CF600. Biodegradation 5: 219–236. Tittabutr, P., Cho, I.K., and Li, Q.X. (2011) Phn and Nag-like Providenti, M.A., Mampel, J., MacSween, S., Cook, A.M., and dioxygenases metabolize polycyclic aromatic hydrocar- Wyndham, R.C. (2001) Comamonas testosteroni BR6020 bons in Burkholderia sp. C3. Biodegradation doi:10.1007/ possesses a single genetic locus for extradiol cleavage of s10532-011-9468-y. protocatechuate. Microbiology 147: 2157–2167. Uhlik, O., Jecna, K., Mackova, M., Vlcek, C., Hroudova, M., Providenti, M.A., O’Brien, J.M., Ruff, J., Cook, A.M., and Demnerova, K., et al. (2009) Biphenyl-metabolizing bacte- Lambert, I.B. (2006a) Metabolism of isovanillate, vanillate, ria in the rhizosphere of horseradish and bulk soil contami- and veratrate by Comamonas testosteroni strain BR6020. nated by polychlorinated biphenyls as revealed by stable J Bacteriol 188: 3862–3869. isotope probing. Appl Environ Microbiol 75: 6471–6477.

Vedler, E. (2009) Megaplasmids and the degradation of aro- protocatechuate 3,4-dioxygenase genes. J Bacteriol 171: matic compounds by soil bacteria. In Microbial Megaplas- 5907–5914. mids. Schwartz, E. (ed.). Berlin, Germany: Springer-Verlag Berlin Heidelberg, pp. 33–53. Watanabe, K., Teramoto, M., Futamata, H., and Harayama, Supporting information S. (1998) Molecular detection, isolation, and physiological characterization of functionally dominant phenol-degrading Additional Supporting Information may be found in the online bacteria in activated sludge. Appl Environ Microbiol 64: version of this article: 4396–4402. Fig. S1. Chromosomal location of catabolic gene markers Weelink, S.A.B., Tan, N.C.G., ten Broeke, H., van den belonging to peripheral and central routes linked by catechol, Kieboom, C., van Doesburg, W., Langenhoff, A.A.M., et al. protocatechuate, gentisate and homogentisate as ring- (2008) Isolation and characterization of Alicycliphilus deni- cleavage intermediates, in the genomes of selected Burkhold- trificans strain BC, which grows on benzene with chlorate eriales strains. Replicons are represented linearly with as the electron acceptor. Appl Environ Microbiol 74: 6672– end-points corresponding to those in the Entrez Genome 6681. database. The genomic size and mean G+C content of total Xie, S., Sun, W., Luo, C., and Cupples, A.M. (2011) Novel genes including the standard deviation are indicated at the aerobic benzene degrading microorganisms identified in right and left end-points of each replicon respectively. The three soils by stable isotope probing. Biodegradation 22: G+C content of each gene marker is indicated next to the 71–81. symbol representing it. Catabolic gene markers included in Yoshida, R., Hori, K., Fujiwara, M., Saeki, Y., and Kag- genomic islands as predicted by IslandViewer (Langille and amiyama, H. (1976) Nonidentical subunits of protocat- Brinkman, 2009) are indicated with an asterisk next to the echuate 3,4-dioxygenase. Biochemistry 15: 4048–4053. symbol representing it. Yudistira, H., McClarty, L., Bloodworth, R.A., Hammond, Table S1. Burkholderiales strains information and genome S.A., Butcher, H., Mark, B.L., and Cardona, S.T. (2011) statistics. Phenylalanine induces Burkholderia cenocepacia phenylacetic acid catabolism through degradation to Please note: Wiley-Blackwell are not responsible for phenylacetyl-CoA in synthetic cystic fibrosis sputum the content or functionality of any supporting materials sup- medium. Microb Pathog 51: 186–193. plied by the authors. Any queries (other than missing mate- Zylstra, G.J., Olsen, R.H., and Ballou, D.P. (1989) Cloning, rial) should be directed to the corresponding author for the expression, and regulation of the Pseudomonas cepacia article.