Biodegradation of polychlorinated biphenyls (PCBs) by the novel identified cyanobacterium Anabaena PD-1

Polychlorinated biphenyls (PCBs), a class of hazardous pollutants, are difficult to dissipate in the natural environment. In this study, a cyanobacterial strain Anabaena PD-1 showed good resistance against PCB congeners. Compared to a control group, chlorophyll a content decreased 3.7% and 11.7% when Anabaena PD-1 was exposed to 2 and 5 mg/L PCBs for 7 d. This cyanobacterial strain was capable of decomposing PCB congeners which was conclusively proved by determination of chloride ion concentrations in chlorine-free medium. After 7 d, the chloride ion concentrations in PCB-treated groups (1, 2, 5 mg/L) were 3.55, 3.05, and 2.25 mg/L, respectively. The genetic information of strain PD-1 was obtained through 16S rRNA sequencing analysis. The GenBank accession number of 16S rRNA of Anabaena PD-1 was KF201693.1. Phylogenetic tree analysis clearly indicated that Anabaena PD-1 belonged to the genus Anabaena. The degradation half-life of Aroclor 1254 by Anabaena PD-1 was 11.36 d; the total degradation rate for Aroclor 1254 was 84.4% after 25 d. Less chlorinated PCB congeners were more likely to be degraded by Anabaena PD-1 in comparison with highly chlorinated congeners. Meta- and para-chlorines in trichlorodiphenyls and tetrachlorobiphenyls were more susceptible to dechlorination than ortho-chlorines during the PCB-degradation process by Anabaena PD-1. Furthermore, Anabaena PD-1 can decompose dioxin-like PCBs. The percent biodegradation of 12 dioxin-like PCBs by strain PD-1 ranged from 37.4% to 68.4% after 25 days. Results above demonstrate that Anabaena PD-1 is a PCB-degrader with great potential for the in situ bioremediation of PCB-contaminated paddy soils.


Introduction
Polychlorinated biphenyls (PCBs) are a class of anthropogenic chlorinated organic compounds listed as persistent organic pollutants (POP) by the Stockholm Convention on May 22, 2001. These ubiquitous environmental contaminants were dispersed widely in the global ecosystem because of their long half-lives and semi-volatility [1][2][3]. PCBs are characterized by high solid-phase extraction and gas chromatography. This study could provide a new knowledge for PCB-biodegradation and establish the application of dechlorination functional cyanobacteria for biological disposal of persistent organic wastes.

Chemicals
Aroclor 1254 was purchased from Sigma-Aldrich (St. Louis, MO, USA). The dioxin-like PCBs mixture standards including PCB77, PCB81, PCB105, PCB114, PCB118, PCB123, PCB126, PCB156, PCB157, PCB167, PCB169, and PCB189 were purchased from Aladdin Reagent (Shanghai, China). The purities of all PCB standards were >99%. The International Union of Pure and Applied Chemistry (IUPAC) numbers are used to identify the PCB congeners. Methanol and n-hexane purchased from Huadong Pharmaceutical (Hangzhou, China) were of GCor HPLC-grade. All the other reagents were of analytical grade.

Cyanobacterial strain and culture conditions
Anabaena PD-1 was isolated from PCB-contaminated paddy soils in Taizhou, Zhejiang, China (No specific permissions were required for the sampling locations and activities. The field studies did not involve endangered and protected species. The sampling site in the study is located at Latitude 28°32'N Longitude 121°27'E.). Anabaena PD-1 cells were grown at 25°C, 2000 lux, in BG11 liquid enrichment media [33]. One liter of the BG11 medium contained 0.04 g K 2 HPO 4 , 0.075 g MgSO 4 Á7H 2 O,0.036 g CaCl 2 Á2H 2 O, 6.0 mg citric acid, 6.0 mg ferric ammonium citrate, 1.0 mg Na 2 EDTA, 0.02 g Na 2 CO 3 , and 1.0 mL trace element solution A 5 . One liter of the trace element solution A 5  Culture temperature and illumination intensity were set to 25±2°C and 2000 lux, respectively. The light to dark ratio was 12h: 12h.
All the Anabaena PD-1cyanobacteria cells were cultured with the chlorine-free medium. The cultured cyanobacteria cells were divided into four groups (I, II, III and IV), and each group was treated with different concentrations of Aroclor 1254. There were another three Aroclor 1254 control groups (CII, CIII and CIV) without any cyanobacteria cells in the chlorine-free mediums. Groups and treatments were set up as shown in Table 1.

Determination of chlorophyll a content
Cyanobacterial cells in the exponential growth phase (OD 680 = 0.38) were exposed to 1, 2, and 5 mg/L Aroclor 1254 for 7 days. Chlorophyll a was determined using a modified hot-ethanol extraction method [34]. To determine chlorophyll a content, the cyanobacterial cultures were harvested and disrupted by ultrasonication for 30 min. After that, the chlorophyll a was extracted using 15 mL hot ethanol (70% v/v, 70°C) in a preheated water bath for 15 min and then in dark place for 6 h. The absorption of extracts was measured at 665 nm and 750 nm by an ultraviolet spectrometry (Shimazu Co. Japan). Four independent extraction and adsorption measurement experiments were performed for each group.

Dechlorination of PCBs by Anabaena PD-1
A chlorine-free medium was utilized to detect the PCB degradation ability of Anabaena PD-1. The chlorine-free medium was modified BG11 medium without addition of CaCl 2 Á2H 2 O and trace element solution. One liter of the chlorine-free BG11 medium contained 0.04 g K 2 HPO 4 , 0.075 g MgSO 4 Á7H 2 O, 0.040 g Ca(NO 3 ) 2 , 6.0 mg citric acid, 6.0 mg ferric ammonium citrate, 1.0 mg Na 2 EDTA, 0.02 g Na 2 CO 3 , and 1.0 mL trace element solution A 5  According to Table 1, Aroclor 1254 was added into 20 mL cultures with or without Anabaena PD-1 cells (OD 680 = 0.38). The concentrations of Aroclor 1254 were adjusted to 1, 2, and 5 mg/L by methanol, respectively. After 7-day exposure, the cyanobacterial cultures in control and PCB-treated groups were centrifuged for 10 min at 4000 rpm. A modified mercuric thiocyanate-ammonium ferric sulfate spectrophotometry method was used to test the concentration of chloride ion in the solution [35]. Five mL supernate was used to measure absorption at 460 nm by an ultraviolet spectrometry (Shimazu Co. Japan). Four independent experiments were performed for each group.
The 16S rRNA sequences were compared and aligned with sequences deposited in the Gen-Bank database using the BLAST program. The 16S rRNA of Anabaena PD-1 and other related cyanbacterial sequences retrieved from the NCBI database were aligned and analyzed using PAUP 4.0 beta, and these aligned sequences were used to construct a phylogenetic tree using the neighbor-joining [36] and Jukes-cantor distance correction matrix methods. The branching pattern was verified using 1000 bootstrap replicates.

PCBs extraction and gas chromatography (GC) analysis
To examine the PCB residues in the degradation system, the culture of Anabaena PD-1 exposed to Aroclor 1254 in the conical flask were directly extracted with n-hexane after ultrasonication for 30 min. The Aroclor 1254 residues in the clear n-hexane extracts were separated in a separatory funnel, and the extracts were dried over 5 g anhydrous sodium sulfate and 5 g silica gel. Then the extracts were nitrogen-dried and dissolved in 1 mL n-hexane for GC analysis.
A 1 μL aliquot of each extract was injected into an Agilent 6890 series gas chromatograph equipped with q capillary column (HP-5 fused-silica, 30 m × 0.32 mm × 0.25 μm) with an electron capture detector (Agilent Technologies, USA). According to a previous study [37], the amended parameters for GC detection were set as follows: the injection temperature was '+' and '-' mean the group was cultured with or without PD-1 cells.
300°C, and the detector temperature was 300°C; the column temperature was initially held at 90°C; ramped at 15°C/min to 180°C and then at 1°C/min to 220°C, held for 5 min; and then at 7°C/min to 290°C and held for 2 min. High-purity nitrogen as carrier gas was maintained at a constant flow rate of 1 mL/min. Both 2,4,5,6-tetrachloro-m-xylene (TMX) and PCB 209 were used as the internal standards. The recoveries of PCB congeners ranged from 93.3% to 103.5%. The detection limit of this method for all the PCB congeners in Aroclor 1254 was about 1 ng/L.

Biodegradation kinetics of Aroclor 1254 by Anabaena PD-1
Cells of Anabaena PD-1 during the exponential phase were applied for the experiment. Four hundred μL Aroclor 1254 (100 mg/L, dissolved in methanol) were added to the cell suspensions. Control cells were inactivated by autoclaving at 120°C for 30 min prior to the addition of Aroclor 1254. Cells were incubated in conical flasks for 0, 1, 2, 5, 10, 15, 20, and 25 d at 25±2°C in a phytotron with a light to dark ratio of 12h: 12h. For each experiment, the initial Aroclor 1254 concentration was adjusted to 2 mg/L. The biodegradation kinetics of Aroclor 1254 was fitted by the equation C t = C 0 e kt , where C 0 and C t are the initial and biodegraded concentrations of Aroclor 1254, respectively. The biodegradation half-life (T 1/2 ) of Aroclor 1254 was calculated using the equation T 1=2 ¼ ln2 k , where k is the biodegradation reaction coefficient. Percent biodegradation of PCB congeners were calculated according to the mass percentage of individual PCB congeners in the mixture.

Determination of dioxin-like PCBs
In order to determine the biodegradation effects of 12 kinds of dioxin-like PCBs by Anabaena PD-1 cells, cyanobacterial cells in the exponential phase were used. Four hundred μL dioxinlike PCBs (100 mg/L, dissolved in methanol) were added to 20 mL cyanobacterial cultures (OD 680 = 0.38). Total PCBs concentration was adjusted to 2 mg/L. Control cells were inactivated by autoclaving at 120°C for 30 min prior to the addition of dioxin-like PCBs. After incubation for 25 d, the cyanobacterial cultures was subjected to ultrasound and extracted for GC/ MS detection. The extracts were clean-up and applied to a SPE-NH 2 cartridge (Varian). The filtered extract was then concentrated under nitrogen and refilled with 100 μL of hexane. Four independent experiments were performed for each group.
Quantification of dioxin-like PCBs was accomplished by use of previously established method [38]. GC/MS analyses were performed on an Agilent Technologies 7890 gas-chromatograph coupled with a 5973 mass spectrometer using a DB5MS column (60 m × 0.25 mm ID × 0.25 μm film, Agilent Technologies, Palo Alto, CA, USA)). Details of the chemical analysis (S1 File), as well as recoveries can be found in the supporting information (S1 Table).

Statistical analysis
The statistical differences of the experimental data were determined using one-way ANOVA followed by two-sided Dunnett's t-test. Statistical tests were conducted using SPSS11.0, and the statistical significance values were defined as Ã P<0.05 and ÃÃ P<0.01. All data of repeated experiments were presented as mean±standard deviation (S.D.) in quadruplicate.

Effect of Aroclor 1254 on cyanobacterial growth
Chlorophyll a is an important index to estimate the biomass of phytoplankton including cyanobacteria [39]. In addition, chlorophyll is a sensitive parameter used to examine microalgae response to toxic organic contaminats. Thus, changes of chlorophyll a content could reflect cyanobacterium Anabaena PD-1 response to stress induced by PCBs. As shown in Fig 1, the chlorophyll a content in the group II increases to 0.82 mg/L. It is significantly higher than that in the group I (p<0.01), which indicates that cyanobacterium Anabaena PD-1 can stress under low dose exposure of Aroclor 1254. Nevertheless, when cyanobacterium Anabaena PD-1 is exposed to 5 mg/L Aroclor 1254 after 7 days, the chlorophyll a content is only 0.67 mg/L. It is significantly lower than that in the group I (p<0.05), which means higher dose exposure of Aroclor 1254 may affect the growth of cyanobacterium Anabaena PD-1. And there is no significant difference of the chlorophyll a content between in the group III and in the group I. It suggests that cyanobacterium Anabaena PD-1 can tolerate 2 mg/L Aroclor 1254.

Biodegradation efficiency of Aroclor 1254 by Anabaena PD-1
Changes of chloride ion concentration in cyanobacterial chlorine-free cultures exposed to 1, 2, and 5 mg/L Aroclor 1254 is presented in Fig 2. On incubation day 7, the Clconcentration of the PCB-treated groups increased significantly compared with the control group. When Anabaena PD-1 was exposed to 1 mg/L PCB, the maximum Clconcentration reached 3.55 mg/L. At PCB exposure concentrations of 2 and 5 mg/L, the Clconcentrations were 3.05 and 2.26 mg/L, respectively. Results above clearly proved that cyanobacterium Anabaena PD-1 could degrade PCB congeners via dechlorination.
16S rRNA sequence analysis and identification of Anabaena PD-1 As shown in Fig 2, it is clearly demonstrated that Anabaena PD-1 can degrade PCBs. The PCR products were linked to pUCm-T carrier, and sequencing was performed after purification and recycling. The detailed 16S rRNA sequencing results are in the supporting information (S2 File).
The 16S rRNA gene sequences of Anabaena PD-1 were aligned with the sequences deposited in the GenBank database using a BLAST search. The results of the phylogenetic analysis showed that strain PD-1 belongs to the genus Anabaena (Fig 3). The 16S rRNA gene sequence

Biodegradation of Aroclor1254
The biodegradation dynamics of Aroclor 1254 by Anabaena PD-1 is presented in Fig 4. After 25 d of incubation, the total concentration of Aroclor 1254 observably decreased. The total Aroclor 1254 removal efficiency of Anabaena PD-1 was approximately 85%. The degradation fit the first order reaction kinetics equation described as 1.4898e -0.0614kt (R 2 = 0.974). The calculated half-life for the biodegradation was 11.36 d. This strain degradation was detectable in mono-, di-, tri-, tetra-, penta-, and hexachlorobiphenyls for 25 d (Table 2), and almost all the tri-and pentachlorodiphenyl congeners of Aroclor 1254 were extensively degraded (greater than 80%). Within 6 d, Anabaena PD-1 degraded the major peaks to a certain degree with 2 mg/L Aroclor 1254.
The biodegradation of PCBs with different positions of chlorine substitution in the benzene ring by Anabaena PD-1 was examined. PCB congeners with different degradation percentages were used for this analysis. The results are shown in Table 2, which indicates that Anabaena PD-1 exhibits greater dechlorination activity toward PCB congeners with meta-and para-chlorine-substituted congeners than those with ortho-chlorine-substituted congeners.
To investigate the influence of PCBs with different chlorine substituent numbers on the degradation by Anabaena PD-1, the PCB congeners were separated into five groups (tri-, tetra-, penta-, hexa-, and hepta-PCBs) according to their chlorine substituent numbers ( Table 2). As can be seen from Table 2, Anabaena PD-1 exhibited greater degrading capability against tri-to pentaCB congeners than PCB congeners substituted with six or more chlorine atoms. After exposure in Aroclor 1254 for 25 d, the percent biodegradation of triCBs, tetraCBs, and pen-taCBs ranged from 61.8% to 100%, 66.8% to 100%, and 30.8% to 90.6%, respectively, whereas the maximum degradation percentage of hexaCB (PCB151) is 76.8%. These results clearly demonstrate that for the degradation of PCBs by cyanobacteria like Anabaena PD-1 the degradation effects for each individual PCB congeners depend on the chlorine substituent positions in biphenyl rings.  Degradation of dioxin-like PCBs by Anabaena PD-1 In addition, the cyanobacterial stain showed a slightly higher degradation rate of lower chlorinated dioxin-like PCBs than the highly chlorinated ones. However, the highest rate was exhibited by PCB169, which suggests the possible relationship between degradation rate and chlorine substituent position.

Discussion
The cyanobacterial strain Anabaena PD-1 originally isolated from PCB-contaminated paddy soils is the first cyanobacterium identified that is capable of PCB congeners degradation. As shown in Fig 1, the chlorophyll-a content in the group II is significantly higher than that in the control group (p<0.01). It demonstrates that under the exposure of 2 mg/L Aroclor 1254 Anabaena PD-1 can tolerate PCBs. The results of chloride ion content in chlorine-free medium (Fig 2) indicate that Anabaena PD-1 exhibits PCB dechlorination capability. Nevertheless, the chlorophyll-a content in the group IV is significantly lower than that in the control group (p<0.05). It means that when the exposure concentration of Aroclor 1254 is higher than 5 mg/L, PCBs may inhibit the growth of Anabaena PD-1. Thus, as shown in Fig 2, the chlorine concentration decreases with increasing PCBs concentration. The chlorine concentration in the group IV is significantly lower than that in the group II (p<0.05).
As PCB-degraders, many microorganisms have resistance against PCBs stress [40]. PCBs have been reported to impair the growth and survival of PCB-degrading microbes [41,42]. Different kinds of bacteria show various levels of tolerance to PCBs. The aerobic PCB-degrading bacteria B. xenovorans LB400 is highly resistant to the toxicity induced by PCBs and can thus survive when exposed to 500 mg/L Aroclor 1242 [43]. Pseudomonas sp. strain B4 shows tolerance to 2 mM 2-chlorobiphenyl [44]. Besides, plant growth-promoting rhizobacteria Pseudomonas fluorescens F113 [45], Pseudomonas putida Flav1-1 and Pseudomonas putida PML2 [46]  have tolerance to PCB stress as well. However, growth of microalgae Cholrella pyrenoidosa and Scenedesmus obliquus is inhibited by Aroclor 1254 (> 2 mg/L) [47]. In this study, the adverse effect of PCBs on Anabaena PD-1 was detected by measuring the contents of chlorophyll a (Fig 1). These results imply that Anabaena PD-1 can tolerate the toxicity of PCB congeners and even degrade PCBs. According to the results shown in Table 2, Anabaena PD-1 is confirmed as a new strain that can efficiently degrade both low-and high-chlorinated PCBs. It is clearly demonstrated that Anabaena PD-1 has stronger PCB commercial mixtures degradation capability than some PCB-degrading bacterial strains and fungi by comparing the degradation substrates and degradation time. According to Boyle et al. [48], Comamonas testosteroni, Rhodococcus rhodochrous, and a strain of Pseudomonas putida can metabolize Aroclor 1242 with total Aroclor 1242 losses of 13.8%, 19.1%, and 24.6%, respectively. Singer and co-workers conducted a study that used Arthrobacter sp. strain B1B and R. eutrophus H850 for Aroclor 1242 degradation [49]. The biodegradation percentages of the PCB commercial mixture by these strains is from 55% to 59%, lower than the decomposition percentages of Aroclor 1254 by Anabaena PD-1. In addition, Yadav et al. [50] found that 30.5% of Aroclor 1254 was degraded by the white rot fungus Phanerochaete chrysosporium. The results in Fig 4 show that strain PD-1 has a stronger capability for Aroclor 1254 degradation than P. chrysosporium. Thus, besides the aforementioned bacterial strains and fungi, the cyanobacterium Anabaena PD-1 may be a novel tool to the biodegradation of PCBs. The results in Table 2 also present the biodegradation percentages of different kinds of PCB congeners. These results show that the strain PD-1 can degrade an extremely wide range of congeners in Aroclor 1254. Strain PD-1 not only can degrade low-chlorinated products but also can decompose the highly chlorinated congeners, including hexa-and hepta-chlorobiphenyls. The degradation percentages of tri-, tetra-, and penta-chlorobiphenyls range from 99.3% to 72.1%. Strain PD-1 shows strong capability to degrade PCB congeners with five or fewer chlorines. The biodegradation effects of PCB congeners are depended on different PCBdegraders and the different experimental conditions. Ralstonia sp. SA-3 isolated from PCBcontaminated African system cannot degrade PCB91 and PCB99 in Aroclor 1242 [51]. However, the biodegradation percentages of PCB91 and PCB99 by Anabaena PD-1 are 84.5% and 75.2%, respectively. In addition, the 20-day biodegradation percentage of Aroclor 1254 by Ceriporia sp. ZLY-2010 is only 7.65% [52], while 84.4% of total Aroclor 1254 is transformed by Anabaena PD-1 after 25 d. Therefore, Anabaena PD-1 exhibit its advantage on different type of PCB congeners. On some tri-, tetra-, and pentachlorobiphenyls, such as 2,2',6-trichlorobiphenyl, 2,3,6-trichlorobiphenyl, 2,2',3,3'-tetrachlorobiphenyl, and 2,2',4,5,5'-pentachlorobiphenyl, Anabaena PD-1 shows biodegradation activities superior or equivalent to those of the previously reported strong PCB degraders, including Acinetobacter sp. P6 and Alcaligenes eutrophus H850 (Table 3). Moreover, compared with Acinetobacter sp. P6 and A. eutrophus H850, Anabaena PD-1 exhibits higher broad-spectrum degradation capability for Aroclor 1254. a: Data from this study; Anabaena PD-1 was exposed to 2 mg/L Aroclor 1254 for 6 days.
The degradation pattern of PCB congeners with different chlorines by Anabaena PD-1 shows that the degradation rate decreases as the number of chlorine atoms in the biphenyl ring increases, which is consistent with other studies [55].The chlorine number on the biphenyl ring can generally influence the PCB dechlorination activity of Anabaena PD-1. Meanwhile, the arrangement of chlorines on the phenyl ring can also affect dechlorination. Table 2 presents the biodegradation patterns for PCB congeners with different chlorine substituted positions. A previous study has shown that PCB congeners with two ortho chlorines (2, 2' or 2, 6) are rarely degraded by most bacterial strains that can degrade PCBs [56]. A high degree of chlorination has been proven to require more energy to break the stable carbon-chlorine bonds [57]. Similarly, our results show that trichlorodiphenyls and tetrachlorobiphenyls with fewer ortho-chlorines are more susceptible to degradation. For example, PCB19, which has three orthochlorines, has a degradation rate (61.8%) lower than that of trichlorodiphenyls with two or one ortho-chlorines such as PCB17 (100%), PCB18 (100%), and PCB28 (100%). The biodegradation rates of PCB41 and PCB60 are 95.7% and 100%, respectively, with no significant difference. Compared with PCB45, with total loss of 68.5% suggests that Anabaena PD-1 has prior meta-and para-chlorine dechlorination ability than ortho-chlorines. This dechlorination pattern is supported by Wiegel, who found that dechlorination activity generally occurred in meta-and para-chlorines of the biphenyl molecule [58]. In addition, the dechlorination of hexachlorobiphenyls and heptachlorobiphenyls by Anabaena PD-1 shows no significant pattern.
As a strong PCB-degrader, Anabaena PD-1 shows detoxification capability for dioxin-like PCBs as well. Dioxin-like PCBs are widespread in a global scale [59], and the residual dioxinlike PCBs in the environment is generally higher than that of the dioxins [60]. Given this high toxicity, most bacterial strains that can efficiently degrade many other PCB congeners cannot decompose dioxin-like PCBs effectively. In this study, the PCB-degrading algae Anabaena PD-1 not only can tolerate dioxin-like PCBs but can also degrade these highly hazardous pollutants. PD-1 can degrade all 12 kinds of dioxin-like PCBs. The biodegradation rate of PCB169 is 68% on incubation day 7. To a certain extent, Anabaena PD-1 can detoxicate dioxin-like PCBs and then reduce the risks of dioxin-like PCBs to the environment and ecology.

Conclusions
In this study, the novel cyanobacterium Anabaena PD-1 isolated from contaminated paddy soils was identified as a strain of the genus of Anabaena using morphology and 16S rRNA sequencing analysis. Anabaena PD-1 is capable of degrading PCBs. The biodegradation rate of Aroclor 1254 was approximately 85% for 25 d, and the biodegradation rate of the dioxin-like PCBs ranged from 40% to 68%. Considering the limited information on PCB degradation by cyanobacteria, this study provides new knowledge for the remediation of PCB-contaminated soils. The key enzymes in degradation, the degradation pathway, and the engineering application effect of Anabaena PD-1 require in-depth study.