Sphingobium fuliginis HC3: A Novel and Robust Isolated Biphenyl- and Polychlorinated Biphenyls-Degrading Bacterium without Dead-End Intermediates Accumulation

Biphenyl and polychlorinated biphenyls (PCBs) are typical environmental pollutants. However, these pollutants are hard to be totally mineralized by environmental microorganisms. One reason for this is the accumulation of dead-end intermediates during biphenyl and PCBs biodegradation, especially benzoate and chlorobenzoates (CBAs). Until now, only a few microorganisms have been reported to have the ability to completely mineralize biphenyl and PCBs. In this research, a novel bacterium HC3, which could degrade biphenyl and PCBs without dead-end intermediates accumulation, was isolated from PCBs-contaminated soil and identified as Sphingobium fuliginis. Benzoate and 3-chlorobenzoate (3-CBA) transformed from biphenyl and 3-chlorobiphenyl (3-CB) could be rapidly degraded by HC3. This strain has strong degradation ability of biphenyl, lower chlorinated (mono-, di- and tri-) PCBs as well as mono-CBAs, and the biphenyl/PCBs catabolic genes of HC3 are cloned on its plasmid. It could degrade 80.7% of 100 mg L −1 biphenyl within 24 h and its biphenyl degradation ability could be enhanced by adding readily available carbon sources such as tryptone and yeast extract. As far as we know, HC3 is the first reported that can degrade biphenyl and 3-CB without accumulation of benzoate and 3-CBA in the genus Sphingobium, which indicates the bacterium has the potential to totally mineralize biphenyl/PCBs and might be a good candidate for restoring biphenyl/PCBs-polluted environments.

Introduction contamination status of this area. A biphenyl-and PCBs-degrading bacterium, S. fuliginis HC3, was isolated and identified. Then, the effects of exogenous carbon sources on biphenyl degradation efficiency of HC3 were studied. After that, the variation trends of benzoate and 3-CBA during biphenyl and 3-CB degradation were investigated. Finally, the degradation ability of PCBs and CBAs of HC3 and the location of biphenyl/PCBs catabolic genes (termed bph) in HC3 were preliminarily studied.

Media
The composition of the mineral salts medium (MSM, g L -1 ) was as follows: KH 2 PO 4

Enrichment, isolation and identification of biphenyl-degrading bacteria
The soil samples (5 g) as well as MSM (100 mL) containing 200 mg L -1 biphenyl were placed in a flask and incubated on a shaker at 30°C and 180 rpm for 6 days. Then, 10 mL of enrichment culture was transferred into fresh media. After 6 generation times, the enrichment culture was diluted and spread on MSM agar plates and sprayed with biphenyl solution (dissolved in acetone, 10 g L -1 ) as a carbon source. The plates were incubated at 30°C, and colonies with yellow halos were selected and purified on LB agar plates. The yellow halos were caused by the formation of a kind of intermediate (2-hydroxy-6-oxo-6-phenylhex-2,4-dienoic acid, HOPDA) during biphenyl degradation.
A biphenyl-degrading bacterial strain HC3 isolated from the soil was identified using the Biolog GEN III system. In addition, the 16S rRNA gene of HC3 was amplified by PCR with a pair of forward and reverse primers: 27f (5 0 -AGA GTT TGA TCC TGG CTC AG-3 0 ) and 1492r (5 0 -TAC CTT GTT ACG ACT T-3 0 ) [34]. The 16S rRNA gene sequence and related sequences acquired from GenBank were aligned by BIOEDIT version 7.0 software. A phylogenetic tree was built by the neighbor-joining method [35] as implemented in MEGA version 4.0 software.
The sample of HC3 for scanning electron microscopy (SEM) was hardened by 2.5% glutaraldehyde solution for 12 h. Then the sample was dehydrated in a series of acetone solution (30%, 50%, 70%, 80%, 90%, 95% and 100%) for 30 min each. After that, the sample was suspended in isoamyl acetate for 15 min and dried by a critical point drier (Hitachi HCP-2) for 2 h. Finally, the sample was fixed and scanned under a SEM (Hitachi S-3000N).

Biphenyl degradation and cell growth experiments
To obtain highly active cells, HC3 was grown in 30 mL of LB medium with 10 mg L -1 biphenyl at 30°C and 180 rpm on a shaker for 24 h. Then, the cells were collected by centrifugation at 8000 rpm for 5 min and washed twice with phosphate buffers (0.05 M, pH 7.2). After that, the cells were resuspended in MSM and adjusted to an OD 600 of 1.0 for the following experiments.
The biodegradation of biphenyl by HC3 was conducted in 100 mL flasks at 30°C and 180 rpm. 2 mL of HC3 cells (OD 600 1.0) was separately inoculated into 18 mL of MSM with 10, 50, 100, 200, 500 and 1000 mg L -1 biphenyl. Residual biphenyl was extracted with the improved method of Hong et al. [14] at 24 h. The test samples were mixed with equal volumes of ethyl acetate, then moved to glass tubes, oscillated heavily on a vortex for 10 min and set statically for 30 min. The organic phases were dried using anhydrous sodium sulfate and measured by GC-MS.
To measure the growth of HC3 on biphenyl, 1 mL of HC3 cells (OD 600 1.0) was inoculated into 19 mL of MSM with 200 mg L -1 biphenyl. The flasks were then cultured at 30°C and 180 rpm. LB medium and the spread plate method were used to enumerate HC3 cell number every 12 h and residual biphenyl was extracted at the same time. Because biphenyl is not easy to utilize as a carbon source by microorganisms, readily available carbon sources were selected to study their influence on biphenyl degradation. In this study, 1 g L -1 glucose, sodium acetate, tryptone and yeast extract was added to MSM separately as carbon sources. 2 mL of HC3 cells (OD 600 1.0) was inoculated into 18 mL of adjusted MSM with 100 mg L -1 biphenyl. Both the culture conditions and the detection methods were as previously described.

Identification and analysis of benzoate and 3-CBA by GC-MS
For benzoate analysis, 100 μL of biphenyl solution (dissolved in acetone, 300 mg L -1 ) was added to sterile glass tubes. After the acetone evaporated, 2 mL of HC3 cells (OD 600 1.0) was added to the tubes. Then these tubes were sealed and cultured at 30°C and 180 rpm for 0, 40, 100 and 220 min. For 3-CBA analysis, 100 μL of 3-CB solution (dissolved in acetone, 200 mg L -1 ) was added to sterile glass tubes. Then 2 mL of HC3 cells (OD 600 1.0) was added to the tubes after the acetone evaporated. These tubes were sealed and cultured at 30°C and 180 rpm for 0, 0.5, 1, 1.5, 2, 2.5, 3, 4 and 4.5 h. At each timepoint, the samples were taken out and acidified to pH 2.0 using HCl (3 M). Biphenyl, 3-CB and their intermediates were extracted using an equal volume of ethyl acetate 2 times. Then, the organic phases were combined together and dried using anhydrous sodium sulfate and N 2 flow. After that, the extracts were dissolved in 100 μL of hexane and 100 μL of BSTFA-TMCS (99:1, v/v) at 60°C for 15 min. Finally, the samples were diluted with hexane to 1 mL and analyzed by GC-MS.

PCBs and CBAs degradation experiments
Before 3 mL of MSM and 2 mL of HC3 cells (OD 600 1.0) were added to sterile glass tubes, 10 PCB congeners (dissolved in hexane) were added to these tubes to a total concentration of 70 mg L -1 . In control groups, HC3 cells were inactivated twice by autoclaving at 121°C for 30 min. All of the tubes were sealed and cultured at 30°C and 180 rpm for 72 h. This treatment method was also adopted to CBAs degradation experiment (including 2-CBA, 3-CBA, 4-CBA and benzoate). PCBs were extracted with the method described by Tu et al. [36], while CBAs were extracted with the method described by Fave et al. [37].

Plasmid curing in strain HC3
The plasmid in HC3 was cured by growth in LB medium with sodium dodecyl sulfonate (SDS, 0. 005%, w/v). The wild type strain of HC3 (OD 600 1.0) was inoculated in this medium and cultured at 30°C and 180 rpm for 24 h. After 3 generation times the culture was spread on LB agar plates and the colonies were selected to extract plasmid DNA by the alkaline lysis method [38]. For restriction enzyme digestion of plasmid DNA, 17 μL extracted plasmid DNA was incubated with 1 μL EcoR I/Hind III and 2 μL 10 × buffer at 37°C for 2 h. Plasmid DNA preparations were electrophoresed in 1.0% agarose gels at 4.0 V min -1 in 1 × TAE buffer for 1 h. The gel was analyzed in a Bio-Rad universal hood II (Bio-Rad Laboratories, Segrate, Italy).
To analyze the stability of HC3, the wild type strain of HC3 was cultured in LB medium overnight and spread on LB agar plates. Then the colonies in these plates were inoculated on new LB agar plates and sprayed with biphenyl solution. Unsuccessful formation of yellow halos around the new colonies in 24 h meant these colonies lost their ability to use biphenyl. Biphenyl, benzoate, 3-CB and 3-CBA were used to study the degradation capacity of a plasmid-cured strain of HC3. The degradation system was identical with that mentioned in the above section (Identification and analysis of benzoate and 3-CBA by GC-MS) and all samples were cultured at 30°C and 180 rpm for 12 h and analyzed by GC-MS.

Analytical methods
Biphenyl was measured using an Agilent 7890A gas chromatography connected with an Agilent 5975C mass selective detector and a DB-5 capillary column (30 m × 0.25 mm, 0.25 μm). The initial column temperature was 80°C with a 1-min hold, a 25°C min -1 increase to 140°C, then an 8°C min -1 increase to 180°C and finally a 15°C min -1 increase to 280°C. The ion source, injector and detector temperatures were 230°C, 290°C and 280°C, respectively. Highly pure He was used as a carrier gas with a constant flow rate of 1.0 mL min -1 . The injecting volume was 2 μL without splitting. The electron impact mode was 70 eV, and the mass scan scope ranged from 50 to 550 amu. The derivatization reagents-treated samples were analyzed by the same method.
PCBs were analyzed by an Agilent 7890A gas chromatography connected with 63 Ni electron capture detector and a DB-5 capillary column (30 m × 0.25 mm, 0.25 μm). The GC programming was initial temperature 150°C, 1 min; ramped at 10°C min -1 to 175°C; isothermal for 3.5 min; then ramped at 1.5°C min -1 to a final temperature of 250°C; isothermal for 3 min. The injector and detector temperatures were 300°C and 325°C, respectively. The injecting volume was 1 μL without splitting. Highly pure N 2 was used as the carrier gas with a constant flow rate of 1.0 mL min -1 . Benzoate and CBAs were measured on an Agilent 1100 series HPLC system equipped with a Zorbax Eclipse XDB-C18 column (4.6 mm × 250 mm, 5 μm). The injecting volume was 20 μL. Methanol and 0.25% acetic acid (v/v) in a 58:42 ratio (v/v) were used as the mobile phase at a flow rate of 1.0 mL min -1 . Peaks were detected by UV absorption at 235 nm.

Data analysis
All of the statistical analysis was performed using SPSS version 16.0 software. One-way analysis of variance was used for statistical comparisons. The significance level was P < 0.05.

Isolation and identification of S. fuliginis HC3
A biphenyl-degrading bacterium, HC3, was isolated from the soil samples and could form round, yellow-colored colonies with a smooth surface on LB agar plates. The bacterium was rod-shaped (Fig 1), gram-negative, oxidase positive, catalase positive, and nitrate-reduction positive, but the bacterium could not grow at 50°C and could not use starch, D-fructose and α-lactose as sole sources of carbon and energy. The partial 16S rRNA gene sequence (1391 bp) of HC3 was 99% similar to Sphingobium fuliginis TKP and 96% similar to Sphingobium ummariense RL-3 and Sphingobium cloacae S-3 (Fig 2). Based on its 16S rRNA gene sequence, morphological features, and physiological and biochemical characteristics (S1 Table), HC3 was identified as S. fuliginis (accession number: KC747727).

Degradation characteristics of biphenyl
To study the biphenyl tolerance ability of HC3, a series of biphenyl concentrations from 10 to 1000 mg L -1 were adopted. The results are showed in Fig 3. With the increase of biphenyl concentration, the degradation percentage of biphenyl decreased, whereas the degradation amount of biphenyl presented the trend of first increasing and then decreasing. At an initial concentration of 500 mg L -1 , 225.5 mg L -1 biphenyl was degraded within 24 h. Meanwhile, when the initial concentration of biphenyl was 1000 mg L -1 , only 165.5 mg L -1 biphenyl was removed within a similar period of time. These results indicated that HC3 could tolerate at least 1000 mg L -1 biphenyl, but its biphenyl degradation ability could be inhibited at this concentration.
The degradation of biphenyl and cell growth indicated that HC3 could utilize biphenyl as the sole carbon source (Fig 4). Four readily available carbon sources, namely glucose, sodium acetate, tryptone and yeast extract were chosen to study their influence on biphenyl degradation efficiency of HC3. The data indicated that biphenyl degradation efficiency was obviously improved by yeast extract and tryptone. Over 97% of biphenyl was degraded by HC3 within  24 h when yeast extract or tryptone was added to the media (Fig 5). Whereas in the other three groups, more than 20% of biphenyl still remained in the media at 24 h.

Identification and analysis of metabolic intermediates: benzoate and 3-CBA
Biphenyl and its intermediates were extracted from the degradation system and analyzed by GC-MS after derivatization with BSTFA-TMCS. Biphenyl (molecular weight 154.2) was eluted at 6.583 min with a molecular-ion peak at m/z 154 (Fig 6a) and a new compound (molecule weight 194) was also eluted at 5.198 min with a molecular-ion peak at m/z 195 (Fig 6b). Both the retention time and the mass spectrogram of the new compound were identical with that of the derivative from benzoate (derivatization with BSTFA-TMCS), which indicated that the precursor of the new compound was benzoate. On the basis of the peak areas of biphenyl and benzoate, it was found that the content of benzoate initially increased and then decreased (Fig 6b) in pace with the degradation of biphenyl (Fig 6a), though the exact benzoate was not quantified.
When 3-CB was used as the degradation substrate of HC3, the production of 3-CBA was also found. GC-MS was used to identify and quantify 3-CB and 3-CBA in order to accurately analyze their variation trends. In the experimental group 3-CB was totally degraded in the initial 2 h (Fig 7). Meanwhile the concentration of 3-CBA increased from 0 mg L -1 to 2.94 mg L -1 in the initial 1 h and then decreased to 0 mg L -1 in the following 1 h. Therefore, biphenyl and 3-CB could be separately degraded into benzoate and 3-CBA and further transformed into other substances by HC3.

Degradation characteristics of PCBs and CBAs
It is well known that various PCBs can be cometabolized by biphenyl-degrading bacteria in aerobic condition, especially PCB congeners with less than four chlorine substituents. The results of PCBs degradation by HC3 are presented in Table 1. After 3 days of biodegradation, mono-, di-and tri-PCBs were significantly degraded in comparing with the control group, while the degradation percentage of tetra-and hex-PCBs were insignificant. Table 2 shows the results of CBAs and benzoate degradation by HC3. 3-CBA, 4-CBA and benzoate were nearly totally degraded after 3 days of incubation. However, the degradation of 2-CBA was not remarkable.

Plasmid curing
Curing of the plasmid in HC3 was implemented with SDS. Compared with the wild type strain of HC3, the plasmid-cured strain completely lost its power to degrade biphenyl, 3-CB, benzoate and 3-CBA, which indicated that the bph genes of HC3 were cloned on its plasmid but not on its chromosomes (Table 3). Agarose gel electrophoresis showed HC3 might contain one plasmid, but the plasmid had no restriction enzyme sites for EcoR I and Hind III (Fig 8). And 6% of the colonies grown overnight in LB medium lost their capacity to degrade biphenyl, which meant that biphenyl degradation characteristic of HC3 was unstable. Discussion S. fuliginis HC3 was isolated from PCBs-contaminated soil in an e-waste recycling region and is able to degrade biphenyl and PCBs. It has been observed that many species in the genus Sphingobium, such as Sphingobium sp. strain P2 [39], Sphingobium amiense sp. nov. strain YT T [40], Sphingobium fuliginis sp. nov. strain TKP T [41], Sphingobium fuliginis TIK-1 [42] and Sphingobium yanoikuyae B1 (formerly Sphingomonas yanoikuyae B1) [43], have the ability to degrade aromatic compounds, such as naphthalene, phenanthrene, nonylphenol, 4-tertbutylphenol, toluene and biphenyl. Among the many members of the genus Sphingobium, HC3 is the first discovered that is able to degrade biphenyl and PCBs without accumulation of benzoate and CBAs.
It has been reported that Mycobacterium sp. PYR-1 was able to degrade over 98% of 80 mg L -1 biphenyl within 72 h [44], whereas Dyella ginsengisoli LA-4 could degrade 95% of 100 mg  L -1 biphenyl within 72 h [16] and Achromobacter sp. BP3 could totally degrade 50 mg L -1 biphenyl within 28 h [14]. In this study, HC3 could degrade 80.7% of 100 mg L -1 biphenyl within 24 h (Fig 3). However, when the concentration of biphenyl was over 500 mg L -1 , the degradation capacity of HC3 could be inhibited, even though it could grow on 1000 mg L -1 biphenyl, which was similar to Dyella ginsengisoli LA-4 [45]. Because of the strong hydrophobicity, low biodegradability and high biotoxicity of recalcitrant organics, such as polycyclic aromatic hydrocarbons (PAHs), PCBs and biphenyl, the biodegradation efficiency of them can be greatly limited [5,19,46,47,48]. When biphenyl was used as the sole carbon source, the lag phase of HC3 lasted for 24 h (Fig 4). After the lag phase, the biphenyl degradation efficiency of HC3 began to increase in pace with the cell growth. This phenomenon has also been found by Chang et al. [49] and Adebusoye et al. [50]. Thus, biodegradation of these compounds could be possibly enhanced by stimulating the growth of microorganisms with readily available carbon sources such as pyruvate and glucose, which has been proven by Lee et al. [51] and Luo et al. [52]. In the current study, the biphenyl degradation efficiency of HC3 was obviously improved by tryptone and yeast extract (Fig 5). One possible reason is that tryptone and yeast extract contain abundant proteins, amino acids and vitamins, which serve as carbon sources, nitrogen sources, energy sources as well as growth factors for HC3.
Removal of biphenyl or PCBs does not mean these compounds can be totally mineralized to carbon dioxide and water. In fact, it is very easy to accumulate toxic intermediates during biphenyl and PCBs biodegradation because many biphenyl/PCBs-degrading bacteria do not contain a complete biphenyl catabolic pathway [22]. And the most easily accumulated dead-end intermediates are benzoate and CBAs, which have been reported to have negative effects on microorganisms growth and biphenyl/PCBs biodegradation [23,25,28]. Until now, only a few bacteria have been reported to have the ability to degrade biphenyl and benzoate or PCBs and CBAs simultaneously, including Dyella ginsengisoli LA-4 [16], Burkholderia xenovorans LB400  [53] and Rhodococcus sp. RHA1 [54]. LB400 and RHA1 have been proved to have relatively complete bph genes [55,56], which make these two strains have the potential to degrade a wide range of aromatic compounds. In our research, benzoate and 3-CBA transformed from biphenyl and 3-CB could be rapidly degraded by HC3 (Figs 6 and 7), which revealed that HC3 might contain a complete biphenyl catabolic pathway. Moreover, without the accumulation of benzoate or CBAs, the toxicity of these compounds on HC3 cells could be alleviated. This is one of the possible reasons why HC3 has high biphenyl degradation efficiency.
Although only a few bacteria have been reported to have the ability to degrade PCBs and CBAs simultaneously, many researchers have proved that bacterial recombinants could express the enzymes for the upper and lower PCBs degradation pathways through the assemblage of plasmid encoded genes and completely metabolize low chlorinated biphenyl [29,30]. However, plasmids are liable to be lost and bacteria will lose their ability to metabolize substrates when they are grown under non-selective conditions [30,31]. The bph genes of HC3 are cloned on its plasmid and thus its PCBs degradation capacity is unstable (Fig 8; Table 3). The loss rate of the ability to utilize special substrates per generation in non-selective media is 6% in HC3, 4 to 8% in Acinetobacter sp. strain P6 and Arthrobacter sp. strain M5 [57], 1% in Alcaligenes eutrophus JMP298 [31] and 8% in Pseudomonas cepacia CSV90 [58], which are possibly caused by the loss of plasmids in these strains. Since there are only a few bacteria like HC3 which is able to completely metabolize low chlorinated biphenyl, it is still necessary to screen this kinds of isolates from natural environments to enrich the bank of PCBs-degrading bacteria. In addition, wide-type degrading-bacteria can be used safely in soil remediation while the application of genetic engineering bacteria in natural environment is still restricted at present. By the way, it is well known that PCB degradation process is more efficient when performs by a mix population expressing both the upper and lower pathways [59,60]. Since a mix population usually contains different microorganisms, its metabolic pathways are more diverse and thus its substrate range is wider. In fact, the function of a mix population is also unstable. If the culture condition of a mix population changes, its original functions will change or lose [61,62]. So PCBs degradation capacity of both mix populations and pure cultures are still worth researching.
It has been reported that bph genes are not only present on plasmids but also present on bacterial chromosomes and transposons [63,64]. And the typical bph gene cluster is composed of bphR1bphA1A2(orf3)bphA3A4BCX0X1X2X3D, which is found in Burkholderia sp. strain LB400 [65][66][67] and Pseudomonas pseudoalcaligenes KF707 [68,69]. Furukawa et al. [1] compared the bph gene clusters of many bacteria and demonstrated that some are similar but some are very different on the basis of the structure of each gene and gene organization. These results indicated that certain bph gene clusters can transfer among soil bacteria and have evolved from a common ancestor. Thus, in order to determine the type of bph gene cluster in HC3, the differences of bph genes between HC3 and the ones that are already recorded in databank should be studied in subsequent research.
Actually, a major bottleneck in the PCB degradation pathway is the substrate range of the first enzyme (BphA) of the upper pathway. This enzyme is composed of a terminal dioxygenase containing a large subunit (encoded by bphA1) and a small subunit (encoded bphA2), and a ferredoxin (encoded by bphA3), and a ferredoxin reductase (encoded by bphA4) [68]. And it has been reported that the large subunit (BphA1) and the small subunit (BphA2) are responsible for the substrate specificity [63]. The amion acid sequences of BphA between Pseudomonas pseudoalcaligenes KF707 and Burkholderia cepacia LB400 are nearly identical (BphA1, 95.6%; BphA2, 99.5%; BphA3,100%; BphA4, 100%) [68,70], but the substrate ranges of these two strains are different [71,72]. In this research, the substrate range of HC3 is narrow compared to LB400, but HC3 has an excellent capacity to degrade biphenyl and low chlorinated biphenyl. Thus, it is significant to analyze BphA amion acid sequences of HC3. To obtain more information of BphA amion acid sequences will help researchers find out the relationship between BphA amion acid sequences and substrates specificity of PCBs-degrading bacteria.
In conclusion, an excellent biphenyl-degrading bacterial strain HC3, which could degrade 80.7% of 100 mg L -1 biphenyl within 24 h, was isolated from PCBs-contaminated soil. HC3 is also able to degrade benzoate, 3-CBA, 4-CBA, and PCBs with three or fewer chlorine atoms. The bph genes of HC3 are cloned on its plasmid. And the bacterium is the first reported that can degrade biphenyl and 3-CB without accumulation of benzoate and 3-CBA in the genus Sphingobium. All of these findings indicate that HC3 might contain relatively complete bph genes and has the potential to totally mineralize biphenyl and PCBs. Therefore, the bacterium might be a good candidate for restoring biphenyl/PCBs-contaminated environments.
Supporting Information S1