OUP user menu

Microbial biodegradation of polyaromatic hydrocarbons

Ri-He Peng, Ai-Sheng Xiong, Yong Xue, Xiao-Yan Fu, Feng Gao, Wei Zhao, Yong-Sheng Tian, Quan-Hong Yao
DOI: http://dx.doi.org/10.1111/j.1574-6976.2008.00127.x 927-955 First published online: 1 November 2008

Abstract

Polycyclic aromatic hydrocarbons (PAHs) are widespread in various ecosystems and are pollutants of great concern due to their potential toxicity, mutagenicity and carcinogenicity. Because of their hydrophobic nature, most PAHs bind to particulates in soil and sediments, rendering them less available for biological uptake. Microbial degradation represents the major mechanism responsible for the ecological recovery of PAH-contaminated sites. The goal of this review is to provide an outline of the current knowledge of microbial PAH catabolism. In the past decade, the genetic regulation of the pathway involved in naphthalene degradation by different gram-negative and gram-positive bacteria was studied in great detail. Based on both genomic and proteomic data, a deeper understanding of some high-molecular-weight PAH degradation pathways in bacteria was provided. The ability of nonligninolytic and ligninolytic fungi to transform or metabolize PAH pollutants has received considerable attention, and the biochemical principles underlying the degradation of PAHs were examined. In addition, this review summarizes the information known about the biochemical processes that determine the fate of the individual components of PAH mixtures in polluted ecosystems. A deeper understanding of the microorganism-mediated mechanisms of catalysis of PAHs will facilitate the development of new methods to enhance the bioremediation of PAH-contaminated sites.

Keywords
  • polycyclic aromatic hydrocarbons
  • biodegradation
  • microorganisms
  • pathway
  • genetic regulation

Introduction

Polycyclic aromatic hydrocarbons (PAHs) are an all-pervading group of hydrophobic organic compounds consisting of two or more combined benzene rings in linear, angular or cluster arrangements. PAHs are the most important components of crude oil, creosote, asphalt and coal tar. They contaminate the soil via many routes, including the burning of fossil fuels, the manufacture of gas and coal tar, wood processing, escaped automobile gasoline and the incineration of waste (Harvey, 1991). There are more than 100 diverse PAH compounds, and most of them persist in the ecosystem for many years owing to their low aqueous solubility and their absorption to solid particles (Volkering et al., 1992, 1993; Bosma et al., 1997). The presence of PAHs in contaminated soil and sediment poses a significant risk to the soil, because many PAH compounds are known or suspected to be toxic, mutagenic and, in some cases, carcinogenic (Cerniglia & Heitkamp, 1989; Patnaik, 1992). On the basis of their abundance and toxicity, 16 PAH compounds have been included in the United States Soilal Protection Agency's list of priority pollutants (Keith & Telliard, 1979).

The elimination of PAHs from contaminated soil is essential during its remediation according to relevant cleanup standards. The biological treatment of soil contaminated with PAHs should be a more efficient, financially affordable and adaptable choice than physicochemical treatment because it presents potential advantages such as the complete degradation of the pollutants, lower treatment cost, greater safety and less soil disturbance (Habe & Omori, 2003). Many bacteria have been discovered that degrade PAHs via either metabolism or cometabolism. The ordinary biochemical pathways of the bacterial metabolism of PAHs have been scrutinized. Low-molecular-weight (MW) PAHs, such as naphthalene, phenanthrene and anthracene, are usually readily degraded by bacteria in soil and under laboratory conditions (Cerniglia, 1984, 1992; Sutherland et al., 1995). Since the initial studies by Heitkamp et al. (1988a, b) on the bacterial degradation of pyrene, there have been numerous reports describing the microbial oxidation of four-ring PAH. Recently, a Sphingomonad endowed with the remarkable ability to grow on four-ring PAH chrysene as its sole carbon and energy source was isolated (Willison, 2004). However, less is known about the bacteria capable of utilizing PAHs containing five or more rings as a carbon and energy source, such as benzo[α]pyrene and benz[α]anthracene. The mechanisms involved in the degradation of PAHs with more than five rings are unclear.

Diverse fungi capable of utilizing PAHs have been investigated as well. Some filamentous fungi, basidiomycetes, white-rot fungi and deuteromycetes have been shown to remove PAHs more competently than bacteria. PAHs susceptible to fungal biodegradation include naphthalene, phenanthrene, anthracene, pyrene, benzo[α]pyrene, fluorene, dibenzothiophene, catechol, benzo[α]anthracene, benzo[ghi]-perylene, chrysene, benzo[b]fluoranthene and benzo[k]-fluoranthene (Cerniglia, 1997; Zheng & Obbard, 2002, 2003). At least two mechanisms are involved in PAH biodegradation: one utilizes the cytochrome P-450 system (Yadav et al., 2006) and the other uses the soluble extracellular enzymes of lignin catabolism, including lignin peroxidase, manganese peroxidase (Steffen et al., 2003) and laccase (Gianfreda et al., 1999).

In the past decade, more and more microorganisms capable of mineralizing PAH compounds have been isolated, and knowledge about many PAH-catabolic genes from bacteria and fungi is accumulating. However, until now, there have been no reviews summarizing this progress. In this review, we will summarize the recent advances in understanding the natural diversity and capabilities of bacteria and fungi for degrading PAH compounds.

Bacterial catabolism of polycyclic aromatic hydrocarbons

Numerous bacteria have been found that degrade PAHs, and some can utilize low-MW PAHs as their sole carbon source. The common biochemical pathways for the bacterial degradation of PAHs such as naphthalene (Resnick et al., 1996; Annweiler et al., 2000), phenanthrene (Menn et al., 1993; Kiyohara et al., 1994; Pinyakong et al., 2003a, b), anthracene and acenaphthene (Dean-Ross et al., 2001; Pinyakong et al., 2004) have been well investigated. Biodegradation mechanisms require the presence of molecular oxygen to initiate the enzymatic attack of PAH rings. In the initial step, dioxygenase-catalyzed oxidation of arenes generally takes place in aerobic bacterial systems to yield vicinal cis-dihydrodiols as the early bioproducts by a multicomponent enzyme system. These dihydroxylated intermediates may then be cleaved by intradiol or extradiol ring-cleaving dioxygenases through either an ortho-cleavage pathway or a meta-cleavage pathway, leading to central intermediates such as protocatechuates and catechols that are further converted to tricarboxylic acid (TCA) cycle intermediates (Cerniglia, 1992; Eaton & Chapman, 1992; Gibson & Parales, 2000).

Dioxygenases responsible for the formation of cis-dihydrodiols from arenes of PAH substrates appear to be the most ubiquitous in bacteria. They have usually been found to contain multicomponent enzyme systems involving several proteins, nonheme iron, and require NADH (Gibson & Subramanian, 1984; Gibson et al., 1990; Resnick et al., 1994). To date, dioxygenase-containing organisms have been grown on a range of carbon sources, and these enzymes have been classified accordingly, e.g. toluene (TDO), naphthalene (NDO) and biphenyl (BPDO). TDO is confirmed to be an extremely wide-ranging substrate enzyme. The TDO system is most suitable for cis-dihydroxylation of substituted benzene substrates and bicyclic arenes. However, there is a size limitation on substrate acceptability for TDO; bulky and bendy molecules fit poorly into the active site of this enzyme. For larger PAHs, both NDO and BPDO are capable of catalyzing dihydroxylation, with only the latter enzyme capable of metabolizing tetracyclic or larger examples. By contrast, the NDO enzyme appears to be generally incapable of catalyzing the cis-dihydroxylation of monocyclic aromatic rings (Boyd & Sheldrake, 1998).

Low-MW PAH dioxygenase system

The bacterial naphthalene dioxygenase system is particularly useful for oxidizing bi- and tri-cyclic PAH substrates, such as naphthalene, phenanthrene and anthracene. The naphthalene dioxygenase system is a multicomponent enzyme, generally including an NADH oxidoreductase, a ferredoxin and an oxygenase component that contains the active site. This catalytic portion of the naphthalene dioxygenase system, termed NDO, is composed of large and small subunits –α and β, respectively – that are in an α3β3 configuration. NDO is a member of a large family of oxygenases whose α subunits contain a Rieske [2Fe–2S] center and mononuclear nonheme iron, and determine the substrate specificity of NDO (Butler & Mason, 1997; Parales et al., 1998). Generally, two electrons from the reduced pyridine nucleotide are transferred via the reductase, the ferredoxin, and the Rieske center to the Fe (II) ion at the active site during a catalytic cycle. The reducing equivalents allow the activation of molecular oxygen, which is a prerequisite to dihydroxylation of the substrate (Carredano et al., 2000; Ferraro et al., 2005, 2006). Interestingly, the AaAb genes, which encode the ferredoxin and reductase components of the multicomponent PAH initial dioxygenase, are not present on the phn locus of the Burkholderia sp. strain RP007 (Laurie & Lloynd-Jones, 1999). Currently, we are trying to improve the efficiency of the electron transport functions of the nahAcAd gene products by DNA shuffling.

Naphthalene-degrading bacteria are broadly distributed in nature. So far, a few naphthalene dioxygenases have been purified and extensively characterized from different strains of bacteria, and genetic control of the pathways involved in naphthalene degradation has been studied in detail. The naphthalene catabolic genes (nah) of NAH7 are organized into two operons: the nal operon encoding the upper pathway enzymes involved in conversion of naphthalene to salicylate, and the sal operon encoding the lower pathway enzymes involved in the conversion of salicylate to pyruvate and acetyl coenzyme A. The two operons are closely genetically linked to each other and to their common regulatory gene, nahR. The nahR gene encodes a trans-acting positive regulator belonging to the LysR family of transcriptional regulators that are widely distributed in bacteria. NahR is required for the induction of the nah genes by salicylate and the high-level expression of those genes in bacteria (Yen & Gunsalus, 1982, 1985; Grund & Gunsalus, 1983).

The nucleotide sequences of genes encoding the upper pathway enzymes from several Pseudomonas strains have been reported: ndo genes from Pseudomonas putida NCIB 9816 (Yang et al., 1994), nah genes from P. putida G7 (Menn et al., 1993), NCIB9816-4 (Simon et al., 1993), ND6 (Li et al., 2004), BS202 (Kosheleva et al., 1986; Balashova et al., 2001), and P. putida sp. (Ono et al., 2007), dox genes from Pseudomonas sp.C18 (Denome et al., 1993), pah genes from P. putida OUS82 (Kiyohara et al., 1994; Takizawa et al., 1994) and Pseudomonas aeruginosa Pak1 (Takizawa et al., 1999), and nah genes from Pseudomonas stutzeri AN10 (Bosch et al., 1999a, b, 2000). All the nah structural and regulatory genes on the former plasmid are located on a 39-kb class II transposon, Tn4655 (Tsuda & Iino, 1990; Sota et al., 2006). The nucleotide sequences of the upper pathway genes are more than 90% identical. All these genes are arranged in a similar order: nahAa (ferredoxin reductase)- nahAb (NDO ferredoxin)- nahAc (the α-subunit of NDO)- nahAd (the β-subunit of NDO)- nahB (naphthalene-cis-dihydrodiol dehydrogenase)- nahF (salicylaldehyde dehydrogenase)- nahC (1,2-dihydroxynaphthalene dioxygenase)- nahQ (unknown gene)- nahE (trans-o-hydrobenzylidenepyruvate hydratase aldolase)- nahD (2-hydroxychromene-2-carboxylate isomerase) (Habe et al., 2003) (Fig. 1). The genetic regulation of this pathway was also studied in detail for different bacterial strains (Fig. 2). In Ralstonia sp. U2, the naphthalene dioxygenase genes (nag gene) contained all of the genes corresponding to the classical nah gene of Pseudomonas strains in the same order, with the exception of two extra genes inserted between the ferredoxin reductase gene and ferredoxin gene. The two additional genes, named nagG and nagH, are structural subunits of salicylate-5-hydroxylase and can help the host convert naphthalene to gentisate (Fuenmayor et al., 1998; Zhou et al., 2001). In the Comamonas testeroni strain GZ42, the upper pathway genes are similar to those of Ralstonia sp. U2 (Goyal & Zylstra, 1997). However, in C. testeroni strain GZ39, the pathway genes are quite different from those of the typical nah genes. In this strain, several genes (such as extradiol dioxygenase) are not within the cluster, but a glutathione-S-transferase gene was inserted into the cluster (Goyal & Zylstra, 1996; Zylstra et al., 1997).

Figure 1

Gene clusters for naphthalene/phenanthrene upper catabolic pathway in Pseudomonas strains. The pentagon represents the size, location, and direction of transcription of the ORF. The tnpI and tnpR genes encode the site-specific tyrosine and serine recombinases, respectively (Nash et al., 1996). The bold line represents the 84-bp sequence reported Eaton (1994). References for the sequences are as follows: nah genes from Pseudomonas putida G7 (NAH7) (AB237655), P. putida sp. (pSLX928-6) (AB255564), P. putida NCIB9816-4 (pDTG1) (AF491307), Pseudomonas sp. ND6 (pND6-1) (AY208917), Pseudomonas stutzeri AN10 (AF039533), and P. putida BS202 (AF010471); pah genes from Pseudomonas aeruginosa PaK1 (D84146) and P. putida OUS82 (AB004059); ndo genes from P. putida NCIB9816 (M23914); and dox genes from Pseudomonas sp. strain C18 (M60405). Enzyme designations are described in the text.

Figure 2

Proposed catabolic pathways of naphthalene by aerobic bacteria. The compounds are naphthalene (1), cis-1,2-dihydroxy-1,2-dihydronaphthalene (cis-naphthalene dihydrodiol) (2), 1,2-dihydroxynaphthalene (3), 2-hydroxy-2H-chromene-2-carboxylic acid (4), trans-o-hydroxybenzylidenepyruvic acid (5), salicylaldehyde (6), salicylic acid (7), gentisic acid (8), maleylpyruvic acid (9), fumarylpyruvic acid (10), pyruvic acid (11), fumaric acid (12), catechol (13), cis,cis-muconic acid (14), β-ketoadipic acid (15), β-ketoadipyl-CoA (16), succiny-CoA (17), acetyl-CoA (18), 2-hydroxymuconic-semialdehyde (19), 2-hydroxymuconic acid (20), 4-oxalocrotonic acid (21), 2-oxo-4-pentenoic acid (22), 4-hydroxy-2-oxovaleric acid (23), and acetaldehyde (24). The enzymes involved in each reaction step are naphthalene dioxygenase (NahAaAbAcAd) (step A1), cis-naphthalene dihydrodiol dehydrogenase (NahB) (A2), 1,2-dihydronaphthalene dioxygenase (NahC) (A3), 2-hydroxy-2H-chromene-2-carboxylate isomerase (NahD) (A4), trans-o-hydroxybenzylidenepyruvic hydratase-aldolase (NahE) (A5), salicylaldehyde dehydrogenase (NahF) (A6), salicylate 5-hydroxylase (NagGHAaAb) (A7), gentisate 1,2-dioxygenase (NagI) (A8), maleylpyruvate isomerase (NagL) (A9), fumarylpyruvate hydrolase (NagK) (A10), salicylylate hydroxylase (NahG) (A11), catechol 1,2-dioxygenase (A12), β-ketoadipate:succinyl-CoA transferase (A13), β-ketoadipyl-CoA thiolase (A14), catechol 2,3-dioxygenase (NahH) (A15), hydroxymuconic-semialdehyde dehydrogenase (NahI) (A16), 4-oxalocrotonate isomerase (NahJ) (A17), 4-oxalocrotonate decarboxylase (NahK) (A18), hydroxymuconic-semialdehyde hydrolase (NahN) (A19), 2-oxo-4-pentenoate hydratase (NahL) (A20), 2-oxo-4-hydroxypentenoate aldolase (NahM) (A21), and acetaldehyde dehydrogenase (NahO) (A22).

It was determined from the sequences of P. putida G7, NCIB9816-4, ND6, and P. stutzeri AN10 strains that the lower pathway genes consist of 11 genes organized in the order nahGTHINLOMKJY. nahY is not a catabolic gene, but a naphthalene chemotaxis gene (Grimm & Harwood, 1999; Boronin et al., 2001) (Fig. 2). In the strains AN10 and ND6, there is another salicylate hydroxylase gene (nahW) outside but close to the sal operon. It was suggested that isofunctional enzyme might be advantageous for the host to adjust its metabolism to extreme conditions (Bosch et al., 1999a, b; Li et al., 2004).

A number of PAH-degrading phenotypes cannot be explained genotypically by comparison with nah-like sequences in contaminated soils. The nah-like probe can only provide a monitor for 45% of strains isolated with a naphthalene-degrading phenotype, and yet the same probe was only useful for 15% of strains isolated with a phenanthrene-degrading phenotype (Lloyd-Jones et al., 1999, 2000). The enzymes involved in the conversion of naphthalene to salicylate can also degrade phenanthrene to naphthalene-1,2-diol. However, it was found that the sequence similarity and gene organization of the naphthalene catabolic genes were very different between the Burkholderia sp. RP007 strain and some classical naphthalene degradation strains. The strain RP007's naphthalene catabolic genes (phn gene) encode iron–sulfur protein α and β subunits of a PAH initial dioxygenase, but lack both ferredoxin and reductase components. The ferredoxin and reductase genes for electron transport functions of the phn gene may be located elsewhere in the RP007 genome or may be supplied by cellular housekeeping genes (Laurie & Lloyd-Jones, 1999, 2000).

Different routes of naphthalene degradation have also been illustrated. The Alcaligenes faecalis AFK2 strain makes use of phenanthrene as its sole carbon source through the o-phthalate pathway (Kiyohara et al., 1982). Some unique genes, such as 3,4-dihydroxyphenanthrene dioxygenase (phnC), 2-carboxybenzaldehyde dehydrogenase (phnI), and trans-2-carboxybenzal-pyruvate hydratase aldolase (phnH), appear in this pathway, which may make the strain prefer phenanthrene to naphthalene (Nagao et al., 1988; Kiyohara et al., 1990). The Nocardioides sp. KP7 strain degrades phenanthrene via the phthalate pathway as well (Iwabuchi & Harayama, 1997, 1998a, b). The phd genes of this strain belong to a new class of PAH-catabolic genes due to variation in gene organization and sequence similarity. The PhdA and PhdB genes, which encode the α- and β-subunit of phenanthrene dioxygenase, are essential for enzyme activity. Although the basic sequence features of each protein family are conserved, each of the subunits of the dioxygenase shows a modest (<60%) sequence identity to all of the known dioxygenase subunits. The ferredoxin component of the dioxygenase PhdC belongs to the [3Fe–4S] or [4Fe–4S] type of ferredoxin, but not to the [2Fe–2S] type found in most components of PAH dioxygenase (Iwabuchi & Harayama, 1998a, b; Saito et al., 1999). The dispensability of the ferredoxin and ferredoxin reductase components of Escherichia coli suggests the relatively low specificity of the electron transport systems toward the oxygenase components (Saito et al., 2000).

It has been observed that a high proportion of the PAH-degrading isolates in soil belong to the sphingomonads. Unlike other gram-negative bacteria strains, members of the genus Sphingomonas are able to degrade a wide range of natural and xenobiotic compounds, such as biphenyl, (substituted) naphthalene(s), fluorene, (substituted) phenanthrene(s), pyrene, (chlorinated) diphenylether(s), (chlorinated) furan(s), (chlorinated) dibenzo-p-dioxin(s), carbazole, estradiol, polyethylenglycols, chlorinated phenols and different herbicides and pesticides (Basta et al., 2005). It seems possible that PAH-degrading sphingomonads are adapted to the oligotrophic environment by having high-affinity uptake systems and an ability to simultaneously take up mixed substrates rather than being particular specialists for the degradation of hydrophobic aromatic compounds (Eguchi et al., 1996; Schut et al., 1997; Fredrickson et al., 1999; Pinhassi & Hagstrom, 2000). Many sphingomonads degrade naphthalene, phenanthalene, and anthracene via common pathways found in other gram-negative bacteria (Pinyakong et al., 2003a, b). It was demonstrated that large plasmids present in xenobiotic-degrading Sphingomonas strains are responsible for the degradative capabilities (Kim et al., 1996; Feng et al., 1997; Romine et al., 1999; Ogram et al., 2000; Basta et al., 2004). For example, a 40-kb DNA region was found to be related to aromatic catabolism in Sphingobium yanoikuyae strain B1, in which two dioxygenase genes are predicted to be required for conversion of PAHs and biphenyl to simple aromatic acid, and meta-cleavage genes are required for conversion of aromatic acids to the tricarboxylic acid cycle (TAC) intermediate (Yen & Serdar, 1988; Assinder & Williams, 1990; Zylstra & Kim, 1997; Kim & Zylstra, 1999). However, these aromatic-degradative genes for catabolic pathways are often localized in the plasmid separately from each other, or, at least, are not organized in coordinately regulated operons. There is no evident biochemical function of the individual gene clusters. In plasmid pNL1 of strain F199, there are at least 13 gene clusters predicted to encode enzymes associated with the degradation of aromatic compounds (Romine et al., 1999). Several parts of the DNA sequence in pNL1 regions encoding aromatic catabolic genes are similar to those in other Sphingomonas strains (Habe et al., 2003). The ‘flexible’ gene organization (e.g. different combinations of conserved gene clusters) could be one of the mechanisms that allows sphingomonads to adapt quickly and efficiently to novel compounds in the environment (Basta et al., 2005).

In contrast to Pseudomonas and other gram-negative bacteria whose structural genes required for naphthalene utilization are usually clustered, the gram-positive bacteria Rhodococcus strains NCIMB12038, P200 and P400 display only three structural genes required for naphthalene utilization (narAa, narAb and narB) in the region (Larkin et al., 1999; Kulakov et al., 2005). The narAa and narAb genes encode the α- and β-subunits of the naphthalene dioxygenase (NDO) catalytic component of iron–sulfur protein (ISPNAR). Both subunits of the NCIMB12038 NDO showed only 30% amino acid identity to the corresponding P. putida NDO subunits. The narB gene has 39% amino acid identity to NahB from the P. putida G7 strain. No genes were found in the Rhodococcus strains that corresponded to the genes encoding the electron transport components reductase and ferredoxin of NDO in Pseudomonas. It was found that the nar region is not organized into a single operon, but there are several transcription units that differ in the Rhodococcus strains. The narA and narB genes were found to be transcribed as a single unit through different start sites, and their transcription was induced by naphthalene. Putative regulatory genes (narR1 and narR2) are transcribed as a single mRNA in naphthalene-induced cells (Kulakov et al., 2000; Larkin et al., 2005).

As described above, many bacteria containing the enzymes involved in the conversion of naphthalene to salicylate can degrade phenanthrene as well. There are three pathways that can convert phenanthrene dihydrodiols to naphthalene-1,2-diol: initial dioxygenation at the 1,2-position, 3,4-position and 9,10-position of phenanthrene (Pinyakong et al., 2000; Seo et al., 2006) (Fig. 3). In studies of the Burkholderia sp. C3 strain, the initial dioxygenation at the 3,4-C positions may be dominant in the first three days of incubation because the concentrations of cis-phenanthrene-3,4-dihydrodiol were higher than cis-phenanthrene-1,2-dihydrodiol. However, cis-phenanthrene-1,2-dihydrodiol was more abundant than cis-phenanthrene-3,4-dihydrodiol at the 7th day. The result of a constant accumulation of 1-hydroxy-2-naphthoic acid rather than 2-hydroxy-1-naphthoic acid in the phenanthrene-supplemented media suggested that degradation of 2-hydroxy-1-naphthoic acid might not be a limiting step. Slow degradation of 1-hydroxy-2-naphthoic acid probably facilitated the 1,2-dioxygenation metabolic pathway but 3,4-dioxygenation. Naphthalene-1,2-diol was further degraded through meta- and ortho-cleavages. The meta-cleavage of naphthalene-1,2-diol yielded 2-hydroxybenzalpyruvic acid, which was then transformed to gentisic acid through salicylic aldehyde. This reaction was presumably catalyzed by hydratase–aldolase, which works in the naphthalene degradation pathway. The ortho-cleavage formed 2-carboxycinnamic acid, which was then converted to phthalic acid (Seo et al., 2007). The Arthrobacter sp. P1-1 strain can degrade phenanthrene through dioxygenation on the 1,2-, 3,4-, and 9,10-C positions. The intermediate products phenanthrene-1,2-, 3,4- and 9,10-diols mainly undergo meta-cleavage and convert into naphthalene-1,2-diol, except for a limited amount of ortho-cleavage that produces a trace amount of o-carboxyvinylnaphthoates and diphenic acid (Seo et al., 2006). As in the Burkholderia sp. C3 strain, the degradation of naphthalene-1,2-diol was branched again to yield a dominant ortho-cleavage pathway to produce phthalate and a minor meta-cleavage to salicylate. Phenanthrene-degrading bacteria using the ortho-cleavage pathway cannot grow when naphthalene is the sole carbon source (Keum et al., 2005; Xia et al., 2005).

Figure 3

Proposed catabolic pathways of phenanthrene by aerobic bacteria. The compounds are phenanthrene (1), cis-1,2-dihydroxy-1,2-dihydrophenanthrene (2), 1,2-dihydroxyphenanthrene (3), 2-[(E)-2-carboxyvinyl]-1-naphthoic acid (4), trans-4-(2-hydroxynaph -1-yl)-2-oxobut-3-enoic acid (5), 5,6-benzocoumarin (6), 2-hydroxy-1-naphthoic acid (7), naphthalene-1,2-dicarboxylic acid (8), cis-3,4-dihydroxy-3,4-dihydrophenanthrene (9), 3,4-dihydroxyphenanthrene (10), 1-[(E)-2-carboxyvinyl]-2-naphthoic acid (11), trans-4-(1-hydroxynaph-2-yl)-2-oxobut-3-enoic acid (12), 1-hydroxy-2-naphthoic acid (13), 7,8-benzocoumarin (14), 1,2-dihydroxynaphthalene (15), 2-hydroxy-2H-chromene-2-carboxylic acid (16), trans-o-hydroxybenzalpyruvic acid (17), salicylaldehyde (18), salicylic acid (19), trans-2-carboxybenzalpyruvic acid (20), 2-carboxybenzaldehyde (21), o-phthalic acid (22), protocatechuic acid (23), cis-9,10-dihydroxy-1,2-dihydrophenanthrene (24), and 2,2′-diphenic acid.

Anthracene is found in high amounts in PAH-contaminated environments, and a variety of bacterial species have the ability to utilize anthracene as their sole source of carbon and energy (Tongpim & Pickard, 1999; Wick et al., 2002). The initial reactions in the degradation of anthracene are catalyzed by multicomponent dioxygenases to produce cis-1,2-dihydrodiols. The enzymatic attack at the C-1 and C-2 positions of the anthracene moiety is similar to the naphthalene dioxygenase pathways. The product 1,2-dihydroxyanthracene can be cleaved by meta-ring cleavage and converted to 2-hydroxy-3-naphthaldehyde and then to 2-hydroxy-3-naphthoic acid, which is further metabolized to salicylate. Identification of metabolites by comparison with authentic standards and transient accumulation of o-phthalic acid by the Mycobacterium sp. strain during growth on anthracene suggest a pathway through o-phthalic acid and protocatechuic acid (van Herwijnen et al., 2003a, b). An alternate route of enzymatic attack by the Mycobacterium sp. strain PYR-1 at the C-9 and C-10 positions of anthracene was found to form the dead-end product 9, 10-anthraquinone. A small number of side product 6,7-benzocoumarin can also be detected during the meta-ring cleavage process. The accumulation of 1-methoxy-2-hydroxyanthracene provided further evidence for the dioxygenation of anthracene by the Mycobacterium sp. strain PYR-1 (Moody et al., 2001). On the other hand, ortho-cleavage of 1,2-dihydroxyanthracene into 3-(2-carboxyvinyl) naphthalene-2-carboxylic acid has also been reported for Mycobacterium sp. PYR-1 and a Rhodococcus species (Tongpim & Pickard, 1996; Dean-Ross et al., 2001; Moody et al., 2001) (Fig. 4).

Figure 4

Proposed catabolic pathways of anthracene by aerobic bacteria. The compounds are anthracene (1), anthracene-9,10-dihydrodiol (2), 9,10-dihydroxyanthracene (3), 9,10-anthraquinone (4), cis-1,2-dihydroxy-1,2-dihydroanthracene (5), 1,2-dihydroxy anthracene (6), cis-4-(2-hydroxynaphth-3-yl)-2-oxobut-3-enoic acid (7), 2-hydroxy-3-naphthoic acid (8), 6,7-benzocoumarin (9), o-phthalic acid (10), protocatechuic acid (11), 1-methoxy-2-hydroxyanthracene (12), and 3-(-2-carboxyvinyl)naphthalene-2-carboxylic acid (13).

Fluorene is a major component of fossil fuels and can be commonly identified in the atmosphere, fresh water, and both riverine and marine sediments (Grifoll et al., 1992, 1994, 1995). Three major degradative pathways for fluorine have been proposed (Casellas et al., 1997, 1998; Wattiau et al., 2001; Habe et al., 2004). Two of these pathways are initiated by a dioxygenation at the 1,2 or the 3,4 position. The corresponding cis dihydrodiols undergo dehydrogenation and then meta-cleavage. After the aldolase reaction and decarboxylation of the ring-fission product, the resulting indanones are substrates for a biological Baeyer–Villiger reaction, yielding the aromatic lactones 3-isochromanone and 3,4-dihydrocoumarin. Enzymatic hydrolysis of the latter results in the production of 3-(2-hydroxyphenyl) propionic acid, which is further metabolized via catechol, to connect this pathway with the central metabolism. The third route is initiated by monooxygenation at the C-9 position to produce 9-fluorenol, which is then dehydrogenated to 9-fluorenone (Fig. 5). Angular dioxygenation of 9-fluorenone leads to 1,1a-dihydroxy-1-hydro-9-fluorenone, which is dehydrogenated to 2′-carboxy-2,3-dihydroxybiphenyl and then spontaneously transformed to 8-hydroxy-3,4-benzocoumarin. Afterwards, the metabolite is catalyzed by a pathway analogous to biphenyl degradation, leading to the formation of phthalate (Engesser et al., 1989; Selifonov et al., 1993; Wattiau et al., 2001). The genes involved in the degradation of fluorene to phthalate were characterized in the fluorine-degrader Terrabacter sp. strain DBF63. A 14.7-kb region containing the genes flnRB-dbfA1A2-flnED1-ORF16 that carry out fluorene degradation is found in the DBF63 strain. The initial attack on both fluorene and 9-fluorenone is catalyzed by the angular dioxygenase (DbfA1 and DbfA2) to yield 9-fluorenol and 1,1a-dihydroxy-1-hydro-9-fluorenone, respectively. The FlnB protein exhibited activities against both 9-fluorenol and 1,1a-dihydroxy-1-hydro-9-fluorenone to produce 9-fluorenone and 2′-carboxy-2,3-dihydroxybiphenyl, respectively. FlnD1 is a heteromeric protein encoded by flnD1 and ORF16, and is a member of the class III two-subunit extradiol dioxygenases. FlnE was identified as a serine hydrolase for the metacleavage products that yield phthalate (Kasuga et al., 2001; Habe et al., 2003, 2004).

Figure 5

Proposed catabolic pathways of fluorene by aerobic bacteria. The compounds are fluorene (1), cis-1,2-dihydroxy-1,2-dihydrofluorene (2), 1,2-dihydroxy fluorene (3), 2-indanone (4), 3-isochromanone (5), cis-3,4-dihydroxy-3,4-dihydrofluorene (6), 3,4-dihydroxy fluorene (7), 1-indanone (8), 3,4-dihydrocoumarin (9), 3-(2-hydroxyphenyl) propionic acid (10), salicylic acid (11), 9-fluorenol (12), 9-fluorenone (13), 1,1a-dihydroxy-1-hydro-9-fluorenone (14), 2′-carboxy-2,3-dihydroxybiphenyl (15), phthalic acid (16), 4,5-dihydroxyphthalate (17), protocatechuic acid (18) 1,2-dihydro-1,2-dihydroxy-9-fluorenone (19), and 8-hydroxy-3,4-benzocoumarin (20).

High-MW PAH dioxygenase system

The presence of high-MW PAHs in contaminated soils continues to pose significant problems because the persistence and genotoxicity of PAHs increase with increasing molecule size (Cerniglia, 1992). The relationship between PAH environmental persistence and increasing numbers of benzene rings is in agreement with the knowledge of environmental biodegradation rates and molecule size (Bossert & Bartha, 1986; Heitkamp & Cerniglia, 1987). For example, half-lives of the three-ring molecule phenanthrene in soil and sediment may range from 16 to 126 days, while half-lives of the five-ring molecule benzo[α]pyrene (BαP) may range from 229 to 1400 days (Shuttleworth & Cerniglia, 1995). Electrochemical stability and hydrophobicity are two crucial factors for the accretion of high-MW PAHs in the environment (Volkering et al., 1992, 1993; Bosma et al., 1997; Harms & Bosma, 1997).

The first report of the biodegradation of high-MW PAHs was presented in 1975. Gibson et al. identified two products of BαP metabolism –cis-9,10-dihydroxy-9,10-dihydro-benzo[α]pyrene and cis-7,8-dihydroxy-7,8-dihydrobenzo[α]pyrene – from the culture solution of a mutant Beijerinckia sp. strain when it grew with succinate and biphenyl (Gibson et al., 1975). However, in 1988 the bacterium that could degrade PAHs with four aromatic rings was isolated from sediment below an oil field by Heitkamp and Cerniglia for the first time (Heitkamp & Cerniglia, 1988). Also in 1988, Mahaffey et al. (1988) investigated the further catabolism of benz[α]anthracene by the Beijerinckia sp. strain B1 (reclassified as S. yanoikuyae) after induction with biphenyl, m-xylene, or salicylate. Nuclear magnetic resonance (NMR) and mass spectral analyses led to the identification of one major metabolite – 1-hydroxy-2-anthranoic acid – and two minor metabolites – 2-hydroxy-3-phenanthroic acid and 3-hydroxy-2-phenanthroic acid. Later, Mueller et al. (1989, 1990) showed that a bacterial community isolated from a creosote waste site was capable of utilizing fluoranthene as its sole source of carbon and energy for bacterial growth.

Pyrene has often been used as a model compound for high-MW PAH biodegradation because it is structurally similar to several carcinogenic PAHs. Although a number of bacterial isolates have been reported to grow on or mineralize pyrene, the majority of these isolates are nocardioform actinomycetes, such as members of the genus Mycobacterium and Rhodococcus (Heitkamp & Cerniglia, 1988; Walter et al., 1991; Rehmann et al., 1998).

Multiple pathways of pyrene degradation have been offered for the Mycobacterium vanbaalenii PYR-1 strain, including typical dioxygenation and monooxygenation (Heitkamp & Cerniglia, 1988; Brezna et al., 2005). Although Mycobacteria strains can oxidize pyrene via initial dioxygenation at the 1,2-positions, the metabolite will form o-methylated derivatives of pyrene-1,2-diol or the dead end product 4-hydroxy-perinaphthenone. The chief pathway of pyrene degradation is dioxygenation at the 4,5-positions to produce both cis- and trans-4,5-pyrenedihydrodiol by dioxygenase and monooxygense, respectively. Rearomatization of the dihydrodiol and subsequent ortho-cleavage lead to the formation of 4,5-dicarboxyphenanthrene, which is further decarboxylated to 4-phenanthroate. Following another dioxygenation reaction, 4-phenanthroate forms cis-3,4-dihydroxyphenanthrene-4-carboxylate. Rearomatization of the metabolite yields 3,4-dihydroxyphenanthrene, which is further metabolized to 1-hydroxy-2-naphthoate. The subsequent enzymatic reactions, including intradiol ring cleavage dioxygenation, result in the production of o-phthalate. Then phthalate is further transformed to TCA cycle via the β-ketoadipate pathway (Wang et al., 2000; Kim et al., 2003) (Fig. 6).

Figure 6

Proposed catabolic pathways of pyrene by aerobic bacteria. The compounds are pyrene (1), cis-1,2-dihydroxy-1,2-dihydropyrene (2), 1,2-dihydroxy pyrene (3) 4-hydroxyperinaphthenone (4), 1,2-dimethoxypyrene (5), cis-4,5-dihydroxy-4,5-dihydropyrene (6), P2, 4,5-dihydroxypyrene (7), phenanthrene-4,5-dicarboxylate (8), phenanthrene-4-carboxylate (9), cis-3,4-dihydroxyphenanthrene-4-carboxylate (10), 3,4-dihydroxyphenanthrene (11), 1-hydroxy-2-naphthoate (12), trans-2′-carboxybenzalpyruvate (13), phthalate (14), 1,2-dihydroxynaphtharene (15), cinnamic acid (16) trans-4,5-dihydroxy-4,5-dihydropyrene (17), 6,6′-dihydroxy-2,2′-biphenyl dicarboxylic acid (18), and pyrene-4,5-dione (19).

Based on both genomic and proteomic data, Kim et al. have identified 27 enzymes necessary for constructing a complete pathway for pyrene degradation. Proteomic analysis also reveals that 18 enzymes in the pathway are upregulated more than twofold, as indicated by peptide counting when the organism was grown with pyrene. There are four ring-hydroxylating steps catalyzed by the corresponding dioxygenase NidAB2, MvanDraft_0817/0818, PhtAaAb, and NidA3B3. The electron transfer components for these four steps have been found: ferredoxin (PhtAc) and ferredoxin reductase (PhtAd). This gene arrangement of the β- and α-subunits of dioxygenase NidAB2 has rarely been found in other bacteria. NidA and NidB proteins are 40-56% similar to the corresponding isofunctional enzymes in the Nocardioides sp. strain KP7 and Pseudomonas sp. strain NCIMB12038. Three copies of the terminal subunits of ring-hydroxylating oxygenase (NidAB2, MvanDraft_0817/0818, and PhtAaAb), dihydrodiol dehydrogenase (MvanDraft_0815), and ring cleavage dioxygenase (MvanDraft_3242) were induced in pyrene-grown cells. However, no ORFs for enzymes such as 1-hydroxy-2-naphthoate hydroxylase, which is necessary for 1-hydroxy-2-naphthoate degradation through salicylate, were detected. ORFs for phthalate 4,5-dioxygenase and protocatechuate-4,5-dioxygenase, which are found in the optional routes for oxidation of phthalate and the meta cleavage of protocatechuate, respectively, were also not detected in the genome sequence (Kim et al., 2004a, 2007).

Besides the M. vanbaalenii PYR-1 strain, many other Mycobacterium sp. strains were identified and found to mineralize pyrene. Most of the PAH-degrading mycobacteria described are fast-growing species within the genus. The mycolic acid-rich cell walls of these soil bacteria may be an important factor in their utilization of hydrophobic substrates such as PAHs.

The Mycobacterium sp. strain RJGII-135 is capable of degrading a wide range of PAHs. Mineralization studies using [14C]pyrene added to contaminated soils, with and without the addition of this strain, indicated that this strain was able to grow on pyrene as its sole source of carbon and energy. During the mineralization of pyrene, three steady intermediates were formed within 4–8 h, which included two identified metabolites in strain PYR-1 – 4-phenanthrenecarboxylic acid and 4,5-pyrenedihydrodiol – and one special product only in the strain RJGII-135 – 4,5-phenanthrenedicarboxylic acid (Schneider et al., 1996).

The Mycobacterium sp. strain KMS was isolated from vadosezone soil at the Champion International Superfund site (Libby, MT), and is able to degrade pyrene and other PAHs. It was determined from 2-dimensional gel electrophoresis that the first aromatic ring hydroxylating dioxygenase that oxidizes pyrene to cis-4,5-pyrene-dihydrodiol includes the α and β subunits, 4Fe–4S ferredoxin, and the Rieske (2Fe–2S) region, which were all induced with the addition of pyrene to the cultures. Other proteins that further pyrene degradation, such as dihydrodiol dehydrogenase, oxidoreductase, and epoxide hydrolase, were also found to be notably induced with the addition of pyrene. There were five different β-subunits and two different Rieske (2Fe–2S) regions of the α-subunit of aromatic ring hydroxylating dioxygenase, indicating that there are multiple copies of dioxygenase expressed during pyrene degradation. Pyrene-4,5-dione was identified to be a pyrene degradation metabolite of the Mycobacterium sp. strain KMS. Pyrene-4,5-dione can then be degraded into two major intermediates – phenanthrene-4,5-dicarboxylic acid and 4-phenanthroic acid – by a flavoprotein-like, FAD-dependent oxidoreductase. Pyrene-4,5-dione can also be reduced back to 4,5-dihydroxypyrene by quinone reductase, as reported for M. vanbaalenii PYR-1 (Liang et al., 2006).

The Mycobacterium sp. strain AP1 was isolated from a pyrene enrichment culture and can grow with pyrene as its sole source of carbon and energy. However, the strain cannot remove the substrate completely, and its growth yield suggests an assimilation of about 10% of the pyrene carbon. A novel metabolite identified as 6,6-dihydroxy-2,2-biphenyl dicarboxylic acid reveals a new branch in the pathway that involves dioxygenation on 4,5 positions and 9,10 positions and cleavage of both central rings of the pyrene (Vila et al., 2001).

The Mycobacterium sp. strain KR2 was isolated from PAH-contaminated soil originating from the area of a former gaswork plant. This isolate metabolized up to 60% of the pyrene added (0.5 mg mL−1) within 8 days at 20 °C. Similar to pyrene degradation by strain PYR-1, pyrene metabolism in the KR2 strain seems to be controlled by metabolite induction. With the exception of 2-carboxybenzaldehyde and cis-3,4-phenanthrene dihydrodiol-4-carboxylic acid, all pyrene metabolites produced by strain KR2 have been detected in cultures of other pyrene-degrading bacteria. The formation of 4-phenanthrol from cis-3,4-phenanthrene dihydrodiol-4-carboxylic acid should be sterically favored under the assumption that the bulky carboxy group is extruded from the bay region, thus occupying a pseudoaxial position necessary for elimination (Rehmann et al., 1998).

The Mycobacterium sp. strain BB1 was isolated from a past coal gasification site. Exponential growth on pyrene was observed over a 35-fold increase of CO2 evolution in the fermentor. In addition, some nonsubstrate and harmless nonionic surfactants were found to enhance the degradation of pyrene (Boldrin et al., 1993; Fritzsche, 1994; Tiehm, 1994).

The Mycobacterium sp. strain CH1 was isolated from PAH-contaminated freshwater sediments and could utilize pyrene or a wide range of branched alkanes and n-alkanes as its sole carbon and energy sources. A lag phase of at least 3 days was detected when strain CH1 was grown in pyrene; however, this lag phase was decreased to <1 day when phenanthrene or fluoranthene was supplied in the culture media. Because there was no detectable hybridization with the nahAc gene, the enzyme system involved in pyrene degradation is unrelated to the naphthalene dioxygenase pathway. Furthermore, weak hybridization of the P. oleovorans alkB gene probe to strain CH1 DNA suggests slight homology between the genes involved in alkane oxidation. These observations suggest that the occurrence of both aromatic and aliphatic hydrocarbon-degradative capacities within a single strain may be more universal (Churchill et al., 1999).

The pathway describing the biodegradation of the four-ringed PAH fluoranthene by M. vanbaalenii PYR-1 has been discovered recently (Kweon et al., 2007) (Fig. 7). Thirty-seven fluoranthene metabolites including potential isomers were isolated from culture medium and analyzed using HPLC, GC-MS, and UV-visible absorption. Fifty-three enzymes were determined to likely be involved in fluoranthene degradation. Four proposed pathways have been proposed for the degradation of fluoranthene initiated by mono- and dioxygenation reactions. The dioxygenase attack occurs at the C-1,2, C-2,3, and C-7,8 positions of the fluoranthene. The C-1,2 and C-2,3 dioxygenation routes degrade fluoranthene via fluorene-type metabolites, whereas the C-7,8 route oxidizes fluoranthene via acenaphthylene-type metabolites by extra- and intradiol ring cleavages. The major site of dioxygenation is the C-2,3 dioxygenation route, which consists of 18 enzymatic steps via 9-fluorenone-1-carboxylic acid and phthalate with the initial ring-hydroxylating oxygenase, NidA3B3, oxidizing fluoranthene to fluoranthene cis-2,3-dihydrodiol. Six cytochrome P450 genes, including CYP51, are responsible for monooxygenation of fluoranthene. Then the monooxygenation product is transformed to monohydroxyfluoranthenes in the catalysis of ring-hydroxylating oxygenases and chemical dehydration of cis- or trans-dihydrodiols (Kweon et al., 2007).

Figure 7

Proposed catabolic pathways of fluoranthene by aerobic bacteria. The compounds are fluoranthene (1), 7,8-dihydroxy fluoranthene (2), 7-methoxy-8-hydroxy -fluoranthene (3), (2Z, 4Z)-2-hydroxy-4-(2-oxoacenaphthylen-1(2H)-ylidene) but-2-enoic acid (4), 1-acenaphthenone-2-carboxylic acid (5), acenaphthylene-1(2H)-one (6),1H,3H-benzo[de] isochromen-1-one (7), acenaphthylen-1-ol (8), acenaphthylen-1,2-diol (9), naphthalene-1,8-dicarboxylic acid (10), 2-(hydroxymethy)-acenaphthylene-1-carboxylic acid (11), 2-formylacenaphthylene-1-carboxylic acid (12), 1,2-dihydroacenaphthylene-1,2-diol (13), 2,3-dihydroxy fluoranthene (14), 1,2-dihydroxy fluoranthene (15), (9E)-9-(carboxymethylene)-9H-fluorene-1-carboxylic acid (16), 9-fluorenone-1-carboxylic acid (17), 9-hydroxy-9H-fluorene-1-carboxylic acid (18), 9-fluorenone (19), 9-hydroxyfluorene (20), 1,2,3-benzenetricarboxylic acid (21), phthalic acid (24), and monohydroxyfluoranthene (22).

There are only limited studies that document extensive mineralization of PAHs that have more than four rings (Kanaly & Harayama, 2000). Most research is focused on the five-ring benzo[α]pyrene (BαP) because it is one of the most important environmental contaminants, and because its mutagenesis and carcinogenesis that cause ubiquitous xenobiotics are well recognized (Kalf & Crommentuijn, 1997; NTP, 2002). All reported BαP oxidization by bacteria has occurred when the bacteria were grown on other substrates, which could induce microorganisms to produce some enzymes that can degrade BαP. An early observation of BαP biodegradation was its oxidation to dihydrodiols by S. yanoikuyae B8/36 (formerly Beijerinckia sp. strain B8/36), a strain that can be induced with biphenyl, m-xylene, or salicylate. However, no ring cleavage products were detected (Gibson et al., 1975; Mahaffey et al., 1988; Gibson et al., 1999). Schneider et al. (1996) identified cis-7,8-dihydrodiol from growing cultures of the Mycobacterium sp. RJGII-135 strain where pyrene was used to maintain PAH degradation. It was suggested that the Mycobacterium sp. RJGII-135 strain was capable of transforming BαP to initial ring oxidation and ring cleavage products by enzymatically attacking at the C-4,5, C-7,8, and/or C-9,10 position (Schneider et al., 1996). When grown in a mixture of yeast extract, peptone, and soluble starch, M. vanbaalenii PYR-1 can biotransform 0.5 mg BαP L−1 to 24.7% aqueous and organic-extractable metabolites. This strain was also shown to slightly degrade BαP in a six-component PAH mixture. Recently, Moody used resting M. vanbaalenii PYR-1 cultures induced with phenanthrene to produce several dihydrodiols and one ring-cleavage product, 10-oxabenzo[def]chrysene-9-one, from benzo[α] pyrene (Moody et al., 2003, 2004).

The M. vanbaalenii strain PYR-1 initially oxidized BαP with dioxygenases and monooxygenases at C-4,5, C-9,10, and C-11,12 (Fig. 8). It was determined from the metabolites separated by reversed-phase HPLC and characterized by UV-visible, mass, NMR, and circular dichroism spectral analyses that the major intermediates of benzo[α]pyrene metabolism were benzo[α]pyrene cis-4,5-dihydrodiol, benzo[α]pyrene cis-11,12-dihydrodiol, benzo[α]pyrene trans-11,12-dihydrodiol, 10-oxabenzo-[def] chrysen-9-one, and hydroxymethoxy and dimethoxy derivatives of BαP. The reaction of dioxygenation in the K region (4,5 positions) to form the cis-dihydrodiol and ortho-cleavage to form chrysene 4,5-dicarboxylic acid are consistent with phenanthrene and pyrene degradation. The ortho-ring fission products 4-formylchrysene-5-carboxylic acid and 4,5-chrysene-dicarboxylic acid and a monocarboxylated chrysene product were formed when replacement culture experiments were conducted with benzo[α]pyrene cis-4,5-dihydrodiol. Chiral stationary-phase HPLC analysis of the dihydrodiols indicated that benzo[α]pyrene cis-4,5-dihydrodiol had 30% 4S,5R and 70% 4R,5S absolute stereochemistry. The initial dioxygenation of BαP at the C-9 and C-10 positions yields benzo[α]pyrene cis-9,10-dihydrodiol. Dehydration of the dihydrodiol to the dihydroxy intermediate with subsequent meta-cleavage and aromatic-ring closure may lead to the formation of 10-oxabenzo-[def] chrysen-9-one. The PYR-1 strain also oxidizes BαP to cis- and trans-11,12-dihydro-11,12-dihydroxybenzo[α]pyrene, which then form cis- and trans-dihydrodiols. Benzo[α]pyrene cis-11,12-dihydrodiol adopted an 11S,12R conformation with 100% optical purity. The enantiomeric composition of benzo[α]pyrene trans-11,12-dihydrodiol was an equal mixture of 11S,12S and 11R,12R molecules. The identification of the cis- and trans-dihydrodiols is an indication that dioxygenation and monooxygenation reactions occurred in naphthalene, anthracene, phenanthrene, and pyrene degradation. The formation of the benzo[α]pyrene trans-11,12-dihydrodiol was probably due to the oxidation by cytochrome P450 to form benzo[α]pyrene-11,12-epoxide and the successive hydrolysis of the latter by epoxide hydrolase (Moody et al., 2004).

Figure 8

Proposed catabolic pathways of benzo[α]pyrene by aerobic bacteria. The compounds are benzo[α]pyrene (1), benzo[α]pyrene-11,12-epoxide (2), trans-benzo[α]pyrene -11,12-dihydrodiol (3), cis-benzo[α]pyrene-11,12-dihydrodiol (4), 11,12-dihydroxy-benzo[α]pyrene (5), hydroxymethoxybenzo[α]pyrene (6), dimethoxybenzo[α]pyrene (7), cis-benzo[α]pyrene-4,5-dihydrodiol (8), 4,5-dihydroxy-benzo[α]pyrene (9), 4-formylchrysene-5-carboxylic acid (10), 4,5-chrysene-dicarboxylic acid (11), chrysene-4(5)-carboxylic acid (12), cis-benzo[α]pyrene-9,10-dihydrodiol (13), 9,10-dihydroxy- benzo[α]pyrene (14), cis-4-(8-Hydroxypyrene-7-yl)-2-oxobut-3-enoic acid (15), pyrene-8-hydroxy-7-aldehyde (16), pyrene-8-hydroxy-7-carboxylic acid (17), cis-benzo[α]pyrene-7,8-dihydrodiol (18), 7,8-dihydroxy-benzo[α]pyrene (19), cis-4-(7-hydroxypyrene-8-yl)-2-oxobut-3-enoic acid (20), pyrene-7-hydroxy-8-aldehyde (20), and pyrene-7-hydroxy-8-carboxylic acid (22).

Additional experiments with [14C]7-benzo[α]pyrene demonstrated that 3.8% mineralization of BαP will take place by the S. yanoikuyae JAR02 strain grown in the presence of both succinate (2.0 mM) and salicylate (2.5 mM) over 7 days. Benzo[α]pyrene-cis-4,5-dihydrodiol and benzo[α]pyrene-cis-7,8-dihydrodiol were observed using HPLC/MS. Further characterization of the radio-labeled metabolite using HPLC/MS and HPLC/MS/MS identified radio-labeled pyrene-8-hydroxy-7-carboxylic acid and unlabeled pyrene-7-hydroxy-8-carboxylic acid as novel ring-cleavage metabolites. The similar chirality of these metabolites suggested that the enzymatic attack position will be at C-7,8 and C-9,10 (Rentz et al., 2008).

Fungi catabolism of polycyclic aromatic hydrocarbons

Several studies have shown that diverse fungi are capable of PAH mineralization. These fungi can be classed into two groups: ligninolytic and nonligninolytic fungi (Cerniglia, 1997). At least two mechanisms are involved in PAH biodegradation: one uses the cytochrome P-450 system and the other uses the soluble extracellular enzymes of lignin catabolism, including lignin peroxidase (LiP), manganese peroxidase (MnP) and laccases (Tortella & Diez, 2005). These enzymes are nonspecific and oxidize a wide variety of organic compounds. These enzymes lack selectivity because lignin contains a variety of aromatic structures formed in plant cell walls by oxygen-radical coupling reactions of 4-hydroxy cinnamyl, 3-methoxy-4-hydroxy cinnamyl and 3,5-dimethoxo-4-hydroxy cinnamyl, resulting in a variety of intermonomer linkages (Hammel, 1992). In nonligninolytic fungi, cytochrome P450 monooxygenase enzymes catalyze the oxidation of PAHs to arene oxides, which are initial products of PAH metabolism.

The extracellular enzymes of lignin catabolism in ligninolytic fungi

Because of their advantage of being able to diffuse to the immobile PAHs, the fungal extracellular enzymes of lignin catabolism appear to be more likely than bacterial intracellular enzymes to make the initial attack on PAHs in soil. White-rot fungi are ubiquitous in nature in their ability to degrade and mineralize recalcitrant plant polymer lignin (Otjen & Blanchette, 1986; Martínez et al., 2005). Several reports have shown that white-rot fungi are capable of degrading PAHs and have potential in the bioremediation of PAH-polluted soils and sediments. Large quantities of mycelia of several species of white-rot fungi are used to increase the extent of PAH bioremediation in soil. The best-characterized strain –Phanerochaete chrysosporium– can oxidize pyrene, anthracene, fluorine and benzo[α]pyrene to the corresponding quinines by lignin peroxidase and manganese peroxidase (Bogan et al., 1996a, b). In contrast to P. chrysosporium, the metabolism of PAHs with Phanerochaete laevis is faster and more extensive (Bogan & Lamar, 1996).

Lignin peroxidase and manganese peroxidase can catalyze the oxidation of the recalcitrant nonphenolic lignin to form a high redox potential oxo-ferryl intermediate during the reaction of the heme cofactor with hydrogen peroxide (H2O2) (Jensen et al., 1996; Gold et al., 2000; Pérez et al., 2002) (Fig. 9a). Two features in the molecular structure confer ligninolytic peroxidases with distinctive catalytic properties: (1) a heme group, conferring high redox potential to the oxo-ferryl complex and (2) the existence of specific-binding sites for oxidation of their characteristic substrates, including nonphenolic aromatics in the cases of LiP, manganous iron in the case of MnP (Reddy & D'souza, 1994; Reddy et al., 2003; Martínez, 2005).The heme group is sandwiched between an N-terminal and a C-terminal helix and rests at the bottom of the space formed by the surfaces of both structures. The iron coordination and the residues involved in the active site are conserved among most peroxidases. The iron is pentacoordinated to the four pyrrole nitrogens of the heme and to the nitrogen in the imidazole group of the proximal histidine. Another distal histidine, assisted by an asparagine residue, takes part in the transfer of the oxidizing equivalents from H2O2 to the heme. It is suggested that the length of the Fe–imidazolic nitrogen (Fe–Nɛ2) bond is related to the redox potential of the different peroxidases (Piontek et al., 1993; Poulos et al., 1993; Sundaramoorthy et al., 1997). Laccases catalyze a one-electron oxidation concomitantly with the four-electron reduction of molecular oxygen to water using a range of phenolic compounds as hydrogen donors (Fig. 9a). The catalysis is carried out by the presence of different copper centers, which were arranged in a trinuclear cluster with one type-1 (T1), one type-2 (T2) and two type-3 (T3) copper ions. The presence of the four cupric ions, each co-ordinated to a single polypeptide chain, is an absolute requirement for optimal activity (Messerschmidt et al., 1997; Solomon et al., 2001; Baldrian et al., 2006).

Figure 9

Catalysis reaction of polycyclic aromatic hydrocarbons by the fungal lignin peroxidase, manganese peroxidase, and laccase. (a) The mechanism of reaction is described in equation. AH2 is the reducing substrate. AH is the reducing substrate after one electron oxidation. (b) Oxidation of polycyclic aromatic hydrocarbons by ligninolytic fungi. Lignin peroxidase, manganese peroxiase, and laccase are capable of oxidizing PAHs to the corresponding quinines.

Various researches have revealed that the mechanism of oxidation of PAHs by fungi ligninolytic enzymes is similar to the degradation of nonphenolic lignin. A correlation has been found between the ionization potential (IP) of PAHs and the specific activity of MnP and LiP. Aromatic substrates with a lower IP were oxidized by the two ligninolytic enzymes and the lower the IP, the faster the oxidation rate. A threshold value of IP was found for each enzyme: LiP oxidizes PAHs with an IP ≤7.55 eV (Vazquez-Duhalt et al., 1994), while MnP oxidizes PAHs with an IP up to 8.2 eV. For example, MnP isolated from Irpex lacteus was able to efficiently degrade three- and four-ringed PAHs with an IP higher than 7.8 eV (Bogan & Lamar, 1995; Bogan et al., 1996a, b). Despite the possibility of direct electron transfer from the aromatic substrate to the heme environment, long-range electron transfer has also been detected in LiP from P. chrysosporium (Schoemaker, 1994). Thus, extracellular LiPs can directly oxidize condensed PAHs to produce unstable aryl cation radicals, whereas MnPs catalyze the oxidation of PAHs in a reaction that requires Mn2+, oxygen, and unsaturated lipids (Baborová et al., 2006; Eibes et al., 2006). A novel manganese–lignin peroxidase (MnLiP) hybrid enzyme produced by Bjerkandera adusta can efficiently oxidize Mn(II) to Mn(III) and can also carry out Mn(II)-independent activity on aromatic substrates (Wang et al., 2003). The role of laccase in the oxidation of nonphenolic lignin structures is similar to LiP. All laccases can be divided into three groups according to their redox potential (E°) at the T1 copper site (Cambria et al., 2008). It has been observed that the different activities of various laccases toward substrates can mainly be attributed to differences in their redox potential. PAHs show E° higher than that of laccase (500–800 mV), and so the PAH degradation by laccase needs to be aided by small mediators (Johannes et al., 1996; Majcherczyk et al., 1998). The most common mediators used are 2,2′-azino-bis-(3-ethylbenzothiazoline-6-sulfonic acid) (ABTS) (Bourbonnais et al., 1995) and –N–OH-type mediators such as 1-hydroxybenzotriazole (HBT) and violuric acid (VA). The E° of ABTS, HBT, and VA are 1.09, 1.12, and 0.91 V, respectively (Xu et al., 2000, 2001). A greater ΔE° could create a more favorable transitional energy state for the interaction between molecular orbitals of the mediator and the enzyme, resulting in faster electron transfer. Laccase efficiently oxidized [14C]phenanthrene in the presence of HBT and Tween 80. 73% of initially added phenanthrene was degraded within 182 h to yield phenanthrene-9,10-quinone and 2,2′-diphenic acid as the major products (Fig. 9b). The addition of the redox mediator ABTS to the reaction mixture increased oxidation of phenanthrene by c. 40% (Han et al., 2004). However, the use of these synthetic mediators implies high added costs and the possible generation of toxic oxidized species. Some of the natural compounds, such as ρ-coumaric acid, strongly promoted the removal of PAHs by laccase. This compound was a better laccase mediator than ABTS and was similar to HBT, attaining 95% removal of anthracene and benzo[α]pyrene and around 50% removal of pyrene within 24 h (Cañas, 2007).

The cytochrome P450 system

The biological functions of cytochrome P450 monooxygenases, such as detoxification of xenobiotics and steroidogenesis, are based on its ability to catalyze the insertion of oxygen into a wide variety of compounds. It was well known that mammalian P450 plays key roles not only in the detoxification of PAHs but also in the activation of PAHs (Shimada et al., 2006). The pathways leading to the formation of PAH diol-epoxides have been extensively studied in rat liver P450s and epoxide hydrolase using BαP as a prototype (Fig. 10a). Traditionally, a K-region oxide, B[α]P-4,5-oxide, was thought to be an oxidation product of BαP by P450s. The metabolite was found to be readily hydrolyzed to an inactive B[α]P-4,5-diol by epoxide hydrolase (Conney et al., 1982). Subsequent examination of several BαP metabolites revealed that CYP1A1 and CYP1B1 enzymes can activate PAHs to toxic and carcinogenic metabolites in collaboration with epoxide hydrolase by the formation of B[α]P-7,8-diol, which were again oxidized to the highly reactive bay region oxides, B[α]P-7,8-diol-9,10-epoxide (Levin et al., 1976; Pelkonen & Nebert, 1982; Shimada & Fujii-Kuriyama, 2004).

Figure 10

Similar metabolic pathways of PAHs by mammals and fungi P450s monooxygenase. (a) Proposed metabolic pathways for activation of benzo[α]pyrene by mammalian P450s. The compounds are benzo[α]pyrene (1), (+)benzo[α]pyrene-7,8-epoxide (2), (−)benzo[α]pyrene-7,8-diol (3), (−)benzo[α]pyrene-7,8-diol-9,10-epoxide-1 (4), (+)benzo[α]pyrene-7,8-diol-9,10-epoxide-2 (5), (−)benzo[α]pyrene-7,8-epoxide (6), (+)benzo[α]pyrene-7,8-diol (7), (−)benzo[α]pyrene-7,8-diol-9,10-epoxide-2 (8), and (+)benzo[α]pyrene-7,8-diol-9,10-epoxide-1 (9). (b) Proposed pathway for phenanthrene degradation by fungi. A metabolic pathway of fungi for PAHs degradation is similar to that of mammals, which is considered to be the formation of diol-epoxides by P450s and epoxide hydrolase. Another metabolic pathway of fungi for PAHs degradation involving: hydroxylation by a monooxygenase, conjugation with sulfate ion (methyl, glucoside, glucuronide, and xyloside), followed by a further hydroxylation to hydroxylarylsulfates (methyl, glucoside, glucuronide, and xyloside) compounds.

Fungi metabolize PAH compounds to metabolites similar to those formed by mammalian enzymes. The predominant pathway in the initial oxidation of PAHs is a cytochrome P450 monooxygenase/epoxide hydrolase that catalyzes the reaction that forms trans-dihydrodiols (Mueller et al., 1995; Cerniglia et al., 1997) (Fig. 10b). These metabolic steps are found in nonligninolytic fungi, such as Cunninghamella elegans, and ligninolytic fungi, such as Pleurotus ostreatus. For example, C. elegans metabolized fluoranthene into fluoranthene trans-2,3-dihydrodiol, 8- and 9-hydroxyfluoranthene trans-2,3-dihydrodiol (Tortella & Diez, 2005), whereas P. ostreatus metabolized pyrene into pyrene trans-4,5-dihydrodiol and anthracene into anthracene trans-1,2-dihydrodiol and 9,10-anthraquinone (Bazalel et al., 1996a, b). The cytochrome P450 can also oxidize numerous PAHs to phenols that are subsequently conjugated with sulfate, glucoronic acid, or glucose (Cerniglia et al., 1982, 1986; Casillas et al., 1996; Pothuluri et al., 1996). Cunninghamella elegans metabolizes benzo[e]pyrene to 10-hydroxy-3-benzo[e]pyrenylsulfate (Cerniglia et al., 1989; Pothuluri et al., 1996) and the fungus A. niger SK9317 metabolizes pyrene to 1-pyrenylsulfate and 1-hydroxy-8-pyrenylsulfate (Wunder et al., 1994; Capotorti et al., 2005).

The fungus cytochrome P450 monooxygenases are a superfamily of heme-thiolate proteins that are associated with membranes in the endoplasmic reticulum (Degtyarenko, 1995; Yadav et al., 2003). Whole-genome sequencing of the white-rot basidiomycete P. chrysosporium has revealed the presence of the largest P450 monooxygenases in fungi comprising c. 150 P450 genes, which could be classified into 12 families and 23 subfamilies and under 11 fungal P450 clans. The amino acid sequence of these P450 monooxygenase proteins is extremely diverse, with levels of identity as low as 16% in some cases, but their structural fold has remained the same throughout evolution (Rabinovich et al., 2004; Doddapaneni et al., 2005a, b). Under two distinct nutrient conditions, nutrient-limited (ligninolytic) and nutrient-sufficient (nonligninolytic), P450 monooxygenases proved to be a functional divergence in transcriptional profiling. P450 genes in this basidiomycetous fungus showed varying degrees of structural similarities to the P450 genes from different ascomycetous fungi such as Aspergillus and Fusarium, suggesting that P. chrysosporium has acquired these P450 families as a part of a vertical descent from a common ancestor. For instance, the CYP63 family of proteins of P. chrysosporium, which consists of seven members (pc-1 through pc-7), shows a structural resemblance to the CYP52 family of ascomycetous yeasts and has, therefore, been classified under the CYP52 clan (Doddapaneni & Yadav, 2004; Doddapaneni et al., 2005a, b; Yadav et al., 2006). All seven genes of the CYP63 family of proteins showed differential transcriptional induction with alkanes and their derivatives (Anzenbacher & Anzenbacherova, 2001). Moreover, their substrate specificity seems to extend beyond alkanes/substituted alkanes as they also showed induction in the presence of various aromatics, including polycyclic aromatic compounds like PAHs. For example, the gene of pc-1 was induced by mono-aromatics (m-hydroxy benzoic acid, nitrophenol, and phenoxy acetic acid), polycyclic aromatic compounds (naphthalene, phenanthrene, pyrene, 3-methylcholanthrene, p-p′-biphenol and polychlorinated biphenyl), and alkyl-substituted aromatics (dodecylbenzene sulfonate, 1-phenyl dodecane and limonene). On the other hand, pc-2 showed no induction in response to mono-aromatic compounds, but could be induced by polycyclic aromatic compounds. pc-4 showed induction specifically in response to polycyclic aromatic compounds and alkyl-substituted aromatics. pc-6 showed high levels of induction in response to some polycyclic aromatic compounds (1-naphthol and 1,2-benzanthracene) and several alkyl-substituted aromatic compounds (Yadav et al., 2006).

Under optimized conditions, the ligninolytic enzymes LiP, MnP, and laccase are capable of oxidizing PAHs with an IP <7.6. However, for PAHs with a higher IP, a cytochrome P450 monoxygenase activity is likely required (Bazalel et al., 1996a, b). For example, the amount of CO2 released from BαP oxidation decreased by 66% in 10 days of culture in the presence of the inhibitor of cytochrome P450, 1-aminobenzotriazole, suggesting that such an enzyme is involved in BαP degradation by Fusarium solani (Rafin et al., 2000). In some fungi that degrade PAHs, extracellular laccase and peroxidase activities were not induced by the presence of PAHs. Moreover, the percentages of degradation did not seem to correlate with the ligninolytic enzyme activities (Gramss et al., 1999; Verdin et al., 2004). In certain fungi, it seems that a wide variety of PAH can be degradated by cytochrome P450 monoxygenase.

Although cytochrome P450-type monooxygenases might be involved in the oxidation of PAHs by several fungi, they oxidize PAHs via a pathway similar to the mammalian enzyme systems to form potent carcinogens – epoxides and dihydrodiols (Mueller et al., 1995; Cerniglia, 1997). On the contrary, peroxidase-mediated extracellular oxidation of PAHs in cultures of white rot fungi results initially in quinines (Bumpus, 1989; Vyas et al., 1994). Because quinones are less toxic than their respective PAH and dihydroxylated metabolites generated by cytochrome P450 monooxygenases, enzymatic oxidation of PAHs by ligninolytic oxidoreductases could be a more useful strategy in detoxification and bioremediation processes.

Degradation of environmental PAH mixtures

Several bacteria have the ability to degrade low- and high-MW PAHs. For example, the Mycobacterium sp. AP1 strain, for which pyrene and fluoranthene degradation pathways have been proposed, can also utilize phenanthrene as its sole source of carbon and energy (Vila et al., 2001; López, 2006). The kinetics of the disappearance of individual PAHs was in agreement with differences in their respective aqueous solubilities (Volkering et al., 1992; Harms & Bosma, 1997). Mycobacterium sp. AP1 showed a significant depletion of naphthalene, phenanthrene, and anthracene when compared with uninoculated controls after 30 days. The decrease in the concentrations of fluoranthene, pyrene and benzo[α]pyrene with respect to the controls was not statistically significant, although typical ring-cleavage products from those compounds could be detected (Vila et al., 2001; López et al., 2007).

It was suggested that the simultaneous bacteria-mediated biodegradation of PAHs dissolving from a PAH mixture was strongly influenced by their different bioavailabilities. Bioavailability is considered to be a dynamic process that is determined by the rate of substrate mass transfer to microbial cells relative to their intrinsic catabolic activity. At high mass transfer rates, the overall biodegradation rate is controlled by the metabolic activity of the bacteria (Koch et al., 1990; Bosma et al., 1997). The observed disappearance rates and extents of the PAH compounds are in agreement with their aqueous solubilities. When the growth-linked compounds phenanthrene, fluoranthene, and pyrene are all added together to a Mycobacterium sp. AP1 strain culture, they will reduce simultaneously and their reduction rates will reach maximum values during the first 168 h. However, the biodegradation rates of fluoranthene and pyrene are significantly lower than that of phenanthrene (López et al., 2007).

Degradation of PAHs in contaminated soil will be also influenced by the diffusional release of PAHs from the nonaqueous phase liquids (NAPLs) phase. The specific surface area of the NAPLs may negatively influence biodegradation because of kinetic restrictions to the dissolution process (Ortega-Calvo & Alexander, 1994). For bacteria growing on NAPL-dissolved PAHs in aqueous systems, production of bioemulsifiers would generate a larger surface area of the substrates and subsequently a higher substrate flux (Reddy et al., 1982). sphingomonads and Mycobacteria growing on NAPL-dissolved PAHs can adhere strongly to the NAPL phase. Consequently, it appears that attachment to and biofilm formation on PAH sources is a widespread mechanism among bacteria to overcome mass-transfer limitations when growing on weakly soluble and strongly sorbed PAHs (Johnsena et al., 2005).

Environmentally derived bacterial isolates often degrade only a narrow range of PAHs, and patterns of simultaneous degradation of PAH mixtures are complex. The observed pattern of utilization of PAHs is not only the result of bioavailability but also of metabolic interactions related to competition and cometabolism effects (Bouchez et al., 1995). Cometabolism is an important feature of the degradation of PAHs. Cometabolism of one kind of PAH could have a synergistic effect increasing the degradation of another. Therefore, cometabolism can widen the range of PAH attacked by a defined strain, especially for the degradation of high-MW PAHs (van Herwijnen et al., 2003a, b). During the incubation, a number of acid metabolites carried out by high-MW PAH-degrading Mycobacteria accumulated in the medium. Most of these metabolites were dicarboxylic aromatic acids formed as a result of the utilization of growth substrates (phenanthrene, pyrene, or fluoranthene) by multibranched pathways including meta and ortho-ring-cleavage reactions: phthalic acid, naphthalene-1,8-dicarboxylic acid, phenanthrene-4,5-dicarboxylic acid, diphenic acid, Z-9-carboxymethylenefluorene-1-carboxylic acid and 6,6′-dihydroxy-2,2′-biphenyl dicarboxylic acid. These partially oxidized compounds produced by one microbial strain could be degraded by others (López et al., 2007).

A number of species of white-rot fungi can cometabolize PAHs with bacteria by improving the compound bioavailability. Fungi cannot utilize any high-MW PAH as their sole carbon and energy source, but can partially degrade these if cultured in a nutrient broth (Sutherland, 1992). Some fungi can oxidize PAHs rapidly to a higher aqueous solubility with their extracellular ligninolytic enzyme systems, resulting in higher rates of mineralization of these metabolites by bacteria. A study carried out with anthracene has confirmed that all detected oxidation products of this compound breakdown by white-rot fungi can be mineralized by indigenous bacteria (e.g. activated sludge) more rapidly than anthracene itself is mineralized (Meulenberg et al., 1997). Addition of indigenous microbial communities to the white-rot fungus Bjerkandera sp. strain BOS55 resulted in an initially rapid increase in the level of CO2 recovery from [14C]BαP and the level of water-soluble label decreased to 16% of initial levels. This demonstrated that some polar metabolites of fungal benzo[α]pyrene oxidation were readily degraded by indigenous microorganisms (Kotterman et al., 1998). Some fungi can use cytochrome P450 to transform PAHs and activate PAHs for further degradation by bacteria. When the nonligninolytic fungus Penicillium janthinelum VUO 10 201 and either its bacterial consortium VUN 10 009 or S. maltophilia VUN 10 010 were combined in one culture, 25% of the BαP was mineralized to CO2 by these fungal–bacterial cocultures over 49 days, accompanied by the transient accumulation and disappearance of intermediates. Inoculation of fungal–bacterial cocultures into PAH-contaminated soil resulted in significantly improved degradation of some high-MW PAHs, such as chrysene, benzo-[α]anthracene and dibenzo[α,h]anthracene (Boonchan et al., 2000). Thus, it was suggested that PAH degradation in nature is a consequence of sequential breakdown by fungi and bacteria, with the fungi performing the initial oxidation step (Meulenberg et al., 1997; Sack et al., 1997; Kotterman et al., 1998).

With mixtures of two individually degradable PAHs, metabolic competition took place whether cometabolism occurred or not (Bouchez et al., 1995). Competition at the active site of enzymes, in particular the initial oxygenase, has been observed most commonly. For example, the presence of low-MW PAHs, such as phenanthrene and fluorene, can inhibit the degradation of fluoranthene and pyrene (Bouchez et al., 2000; Dean-Ross et al., 2002; Lotfabat & Gray, 2002). A second possible cause of inhibition is the dead-end products that result from cometabolic oxidation of nongrowth substrates, such as o-xylene and p-xylene, in a Pseudomonas stutzeri strain (Barbieri et al., 1993; Stringfellow & Aitken, 1995).

The moisture content, pH value, aeration condition, and availability of nutrients were determined to be the key factors in the bioremediation of contaminated soils (Roling-Wilfred et al., 2002; Kim et al., 2005). At early stages (0–45 days) of the bioremediation process, the addition of KNO3 and K2HPO4 caused significant increases in both the heterotrophic and the PAH-degrading microbial populations, while soil stimulated only by aeration and optimum humidity exhibited smaller increases in the sizes of both populations and a marked delay in the increase in microbial PAH degraders. The proportion of the PAH-degrading microbial population compared with the heterotrophic population not treated with nutrients increased noticeably at 21 and 45 days, and reached 100% at 135 days, while lower proportions were observed for the populations treated with nutrients. Denaturing gradient gel electrophoresis analysis and principal component analysis confirmed that there was a remarkable shift in the composition of the bacterial community due to both the biodegradation process and the addition of nutrients. At early stages of biodegradation, the Alphaproteobacteria group (genera Sphingomonas and Azospirillum) was the dominant group in all treatments. At later stages, the Gammaproteobacteria group (genus Xanthomonas), the Alphaproteobacteria group (genus Sphingomonas), and the Cytophaga–Flexibacter–Bacteroides group (Bacteroidetes) were the dominant groups that did not receive nutrients, while the Gammaproteobacteria group (genus Xathomonas), the Betaproteobacteria group (genera Alcaligenes and Achromobacter), and the Alphaproteobacteria group (genus Sphingomonas) were the dominant groups that received nutrients (Vinas et al., 2005). Monitoring the microbial community by terminal restriction fragment length polymorphism (TRFLP) analysis suggests that the aeration change stimulated by soil excavation would cause the change in the microbial community composition. The fragments show several ribotypes considerably increasing in aerobic organisms with a corresponding decrease in anaerobic organisms such as Clostridia sp. (Grant et al., 2007).

Conclusion

During the past decade, a variety of microorganisms have been isolated and characterized for their ability to degrade different PAHs. Furthermore, many metabolic enzymes for the degradation of different PAHs have been isolated from microorganisms and several novel pathways have been elucidated based on the identification of initial ring oxidation and ring cleavage products. The genes responsible for PAH catabolic pathways are always localized as gene clusters, and some gene clusters have been cloned and sequenced. An analysis of the gene organization and transcription patterns in the clusters of different species and an analysis of the homology of related loci led to the conclusion that genetic exchange and rearrangements played a major part in the evolution of the regions that play a role in PAH degradation. Genes that catalyze PAH degradation are believed to have adaptively evolved in nature by various genetic events. In bacteria, some aromatic ring hydroxylating dioxygenases share high homologies in structure and exhibit very relaxed substrate specificities, whereas some dioxygenases with small differences in their amino acid sequences display major differences in their enzymatic properties, such as substrate range and regiospecificity. In fungi, more than one isoenzyme of extracellular enzymes and P450 monooxygenases was found. These genes offer an insight into the evolution of the catalytic versatility of fungus in the detoxification and biodegradation of a broad range of xenobiotic chemicals.

In the natural environment, PAHs tend to interact with non-aqueous phases and soil organic matter and, as a consequence, become potentially unavailable for microbial degradation. Microorganisms have evolved different strategies to adapt to the PAH compounds present in the contaminated situation. Many microorganisms produce biosurfactants to increase the bioavailability of the poorly available substrates. Many microorganisms show chemotaxis towards PAH compounds, which could lead to improved degradation. Environmentally derived bacterial isolates often degrade only a narrow range of PAHs, and degradation of a PAH mixture becomes visible as a cooperative process involving a consortium of strains with complementary capacities. The ubiquitous coexistence of bacteria and fungi in soil and their catabolic cooperation suggests that fungal–bacterial interactions may be of importance for PAH degradation in the natural environment.

Combing genetic engineering tools such as gene conversion, gene duplication and transposition, we can produce novel strains with desirable properties for bioremediation applications according to the knowledge of PAH degradation by microorganisms. For example, the rates of PAH mineralization can be enhanced when NAH plasmids from different bacteria are hybridized using colony hybridization. Overexpression of the dioxygenase fusion system from the Mycobacterium sp. strain PYR-1 in other bacteria might improve its ability to degrade high-MW PAHs. Other technology designed according to the principles of microbial PAH-degradation in soil may also be useful for PAH bioremediation. For example, sequential treatment by bacterial consortia and fungi (e.g. white-rot fungi) or defined fungal–bacterial cocultures are proposed to improve the degradation of PAHs. Salicylate, acting as an inducer of genetic operons, can be used to enhance the degradation of PAH compounds. Addition of small amounts of biosurfactants to the PAH-containing soil may increase the compound's dispersion and bioavailability, whereas it appears to be a good strategy to improve emulsification of NAPL-dissolved PAH in aqueous systems with turbulent mixing.

Acknowledgements

This research was supported by 863 Program (2006AA06Z358; 2006AA10Z117; 2008AA10Z401); Shanghai Subject Chief Scientist (06XD14017); Shanghai Key laboratory and basic research project (06DZ19103-07dz22011); National and Shanghai Natural Science Foundation (06ZR14073-30670179); and Shanghai municipal committee of agriculture (No. 2006-2-1).

Footnotes

  • Editor: David Gutnick

References

View Abstract