Biotransformations of HBCDs by Rhodococcus Strain Stu-38 and Identication of Transformation Products

1, 2, 5, 6, 9, 10-Hexabromocyclododecanes (HBCDs) are new brominated ame retardants causing serious environmental pollution. Dozens of degradative bacteria have been found with capacity to transform HBCDs. In the present study, an aerobic functional bacterium Rhodococcus strain stu-38 was isolated from enriched culture of mangrove sediment using HBCDs as carbon source. This strain could stereoselectively transform HBCDs, the removal rate was α-(cid:0)γ-(cid:0)β-HBCD in the mineral salt medium, but was β-(cid:0)α- and γ-HBCD in the growth medium, and it selectively transformed γ- HBCD in the seawater. Transformation rate of strain stu-38 was lower than other functional strains, however, seven potential debrominated products of HBCDs were identied by using GC-MS. These debrominated products, included dibromocyclododecadiene, bromocyclododecadienol and bromocyclododecatriene were formed through reductive debromination, hydrolytic debromination and dehydrobromination. Overall, Rhodococcus sp. stu-38 diastereoisomer-specically transformed HBCDs to various debrominated products in the different cultural media, which highlighted the complicated stereoselective biotransformation of HBCDs.


Introduction
Hexabromocyclododecanes (HBCDs) are a new type of additive brominated ame retardants (BFRs) applied in extruded (XPS) and expanded (EPS) polystyrene foams. HBCDs are used to improve the ammability resistance or chemically bound to synthetic matrices such as plastics, textiles, electronic circuitry and other materials (Koch et al. 2015). HBCDs released from industry and waste products can in ltrate various ecosystems, leads to serious contamination (Cao et al. 2018). The rst detection of environmental HBCDs was in sh and sediment samples from the river Viskan in Sweden, 1995(Sellström et al. 1998). Since then, they have been found in various environmental media (Cao et al. 2018), even in the human body Roosens et al. 2009). Since HBCDs were recalcitrant to degradation and could lead to neurotoxicity (Mariussen and Fonnum et al. 2003), disruption of the in ammatory response in the immune cells (Yasmin and Whalen et al. 2018) and the bronchial epithelial cells (Koike et al. 2016), they had been listed in Annex A of the Stockholm Convention on Persistent Organic Pollutants (POPs) in May 2013.
Over three tons of HBCDs were released into the environment in Europe each year (Koch et al. 2015). The environmental distribution of HBCDs is mainly in the soil and sediment (Cao et al. 2018;Zhang et al. 2018b), which bene ts for biodegradation. Biotransformation of HBCDs has been observed in the sludge, soil and sediment. Half-lives of HBCDs in the environment varying from days to months (Davis et al. 2005(Davis et al. , 2006Gerecke et al. 2006;Stiborova et al. 2015). Both biotic and abiotic transformation contribute to removal of HBCDs while the biological activity can greatly enhance HBCDs transformation in the environment (Davis et al. 2006;Gerecke et al. 2006; Morris et al. 2004;Stiborova et al. 2015).
Present research investigated the HBCDs transformation by Rhodococcus sp. stu-38 in the different materials including mineral salt medium, seawater and the nutrient seawater. In three media, different diastereoisomer-speci c transformation trends of HBCDs by this strain were found yielding different debrominated products identi ed by using GC-MS. Therefore, the results obtained here provided the primary knowledge about the diastereoisomer-speci c transformation patterns of a functional strain in various environmental materials, and might contribute to the bioremediation of HBCDs contamination.
Seawater was obtained from the offshore area of Shantou, China. Seawater media was ltered through 0.45μm and sterilized. Seawater-LB medium was composed of seawater, 10 g/L peptone, 5 g/L yeast extract powder, pH 7.2-7.5. 100 μg HBCDs were dissolved in dichloromethane, injected into each 50 mL centrifugal tube. Dichloromethane was volatilized before addition of media and bacteria.

Strain isolation and identi cation
Mangrove sediment was collected from Zhanjiang, China. Sediment was mixed with 100 mL of MSM media using 30.0 mg/L HBCDs as the sole carbon source to enrich functional bacteria. The mixture was cultured at room temperature around 25 ℃, 150 r/min. After 30 days, 2 mL of mixture was transferred to a new ask with 100 mL MSM and 30.0 mg/L HBCDs. In the fth transfer, the culture was used for isolation on solid media contain HBCDs. Six strains were obtained and named 38-43 in order. The 16S rRNA gene of strains was ampli ed using the universal primers 27F and 1492R, and sequenced by BGI (Guangzhou, China). The 16S rRNA sequence was aligned using Nucleotide Blast on the NCBI web (https://blast.ncbi.nlm.nih.gov/Blast.cgi). The phylogenetic tree was generated using the neighbor-joining method with MEGA 5.0. Strain 38, renamed stu-38 was used for further research.

Batch experiments
Five groups were set up. In the freshwater group (FW), bacteria were cultured in the seawater-LB media for two days. Cells were harvested by centrifuge at 6000 rpm/min, resuspended with 1mL MSM. Then the cells were inoculated to a 50 mL tube containing 100 μg HBCDs and 5mL MSM. The nal concentration of HBCDs was 16.7 μg/mL. Controls (FW-C) were without addition of cells. In the other two groups, seawater and seawater-LB were used to replace MSM, designed as SW and SWLB which were containing cells of strain stu-38. SWLB-C and SW-C were controls without cells. All groups were cultured at room temperature (25-30℃), 150 rpm/min. Each group was triplicate. The residual HBCDs in the FW and FW-C were determined after two months. The residual HBCDs in the SWLB, SWLB-C, SW and SW-C were determined at one month.

Determination of HBCDs residues and identi cation of debrominated products
To analyze the residues and debrominated products of HBCDs, the cultures were frozen and dried, and extracted by 15 mL of n-hexane, dichloromethane and acetone in the proportion of 1:1:1. The extracts were dried and dissolved in the methanol, determined by Liquid Chromatograph Mass Spectrometer (LC-MS, Thermo TSQ-Endura, USA) equipped with a dual-mode discrete-dynode detector and a Hypersil GOLD 1.9 μ column (50×2.1 mm). Solvent A (methanol: acetonitrile = 80% :20%) and solvent B (10 mmol/L NH 4 Ac) were used as mobile phase. The mobile phase composition started with solvent A: B at 50%: 50%, then to 90%: 10% in 9 minutes, ended up with 100% solvent A in 13 minutes (Morris et al. 2004). The transformation of HBCD was followed by monitoring the HBCD molecular precursor ion (m/z = 640.6 amu) and its fragment ions (m/z = 79.1, 81.1 and 560.5 amu). Concentrations of HBCD were quanti ed based on the HBCD molecular ion (m/z = 81.1 amu) (Davis et al. 2005). The recoveries of HBCDs in the MSM, SW and SWLB were in the range of 111.1-122.5%, 64.0-70.7% and 140-146%, respectively. The transformation e ciency η was de ned as where C 0 was the concentration of control at time t in the FW or initial concentration of substrate in other groups, μg/L; C t was the concentration of HBCD at time t, μg/L.
After determination, the extracts were dried and dissolved in the mixture of n-hexane, dichloromethane and acetone, analyzed by Gas Chromatography/ Mass Spectrometer (GC-MS, QP-2010ULTRA model, Shimadzu, Japan) equipped with a HP-5ms Ultra Inert Capillary Column (30 m×0.25 mm×0.25 μm lm thickness). The procedures were 80 ℃ for 1 min, then increase to 250 ℃ in 13 min by 10 ℃/min, and end up with 250 ℃for 10 min. Full scan of molecules in the range of 30-700 m/z was performed in select ion monitoring (SIM) mode to detect possible brominated degradation products (Davis et al. 2005;Zhong et al. 2018). HBCDs dissolved in the acetone and placed in the room temperature for 48 hours to harvest debrominated products (Zhong et al. 2010) were used as the positive control for identi cation.

Isolation and identi cation of bacteria
Six of bacterial strains grew on MSM containing HBCDs were obtained. The colony of strain 38 turned slight red on solid SWLB medium after several days of culture (Fig. 1a). Its 16S rRNA gene sequence displayed 99% identity with Rhodococcus strains through Nucleotide Blast analysis on the NCBI website. The GenBank accession number of its 16S rRNA was MT815909. The isolated strain 38 was renamed Rhodococcus sp. stu-38. Strain 39 to 43 were identi ed as Marinobacter, Nitratireductor, Brucella, Sinomonas and Ochrobactrum in order, as displayed in the Fig. 1.

Transformation of HBCDs by Rhodococcus sp. stu-38
The transformation rates of HBCDs by isolates in the FW were determined. Strain Rhodococcus sp. stu-38 showed better removal ability than others. As was shown in Fig. 2. 37% α-HBCD and 24% γ-HBCD were removed in comparison with control in two months. And the lowest removal rate was on β-HBCD. Strain Marinobacter slightly transformed HBCDs. Nitratireductor, Brucella, Sinomonas, Ochrobactrum did not showed transformation ability. Therefore, strain Rhodococcus sp. stu-38 were used for further research.

Transformation of HBCDs by Rhodococcus strain stu-38
In the present research, six bacterial strains grew on MSM containing HBCDs were isolated. Strain Rhodococcus sp. stu-38 showed better transformation ability than the others. Stu-38 could not use HBCDs as the sole carbon source and might survive the oligotrophic HBCDs-containing media by living on CO 2 (Feisthauer et al. 2008;Ohhata et al. 2007;Yano et al. 2015). Some Rhodococcus strains had been found with dehalogenation capacity, for example, strains 1CP, JT-3 and EK2 could transform various organohalides (Khosrowabadi and Huyop et al. 2014;Roth et al. 2013;Zhang et al. 2018a). This is the rst study demonstrated the capacity of Rhodococcus strain to convert HBCDs.
The HBCDs removal ability of strain stu-38 was lower than functional strains from other research. For example, Bacillus sp. HBCD-sjtu was reported to consume 90% HBCDs at 321.0 μg/mL in four days (Shah et al. 2018;Shah et al. 2019). Sphingobium chinhatense IP26 could transform 78% and 63% of (-) β-and (+) β-HBCDs from initial concentration of 1.0 μg/mL in six days (Heeb et al. 2017). The low e cient transformation ability of stu-38 could be a heritage of converting natural organohalides in situ (Verma et al. 2014;Kaster et al. 2014;Yu et al. 2021). The sediment applied for enrichment culture was rarely contaminated HBCDs (unpublished data), which might not be a selective pressure for indigenous bacteria.

R. sp. stu-38 facilitated the abiotic transformation of HBCDs
Biotransformation half-lives of HBCDs are varying from a few days to over 100 days in the sludge, soil and sediment (Yu et al. 2021). Present study shew that abiotic transformation of HBCDs had a major contribution in the seawater media, which was corresponding to previous reports that abiotic loss was a large contribution in transformation of HBCDs in the aquatic sediment and active sludge (Davis et al. 2005(Davis et al. , 2006. The removal rates of SW and SWLB were higher than SW-C and SWLB-C, indicated that the presence of strain stu-38 could facilitate the removal of HBCDs, which was corresponding to other research . Chemical and physical factors can lead to abiotic transformation of HBCDs, for example, FeS, nanoscale zero-valent aluminum, sul dated nanoscale zerovalent iron, and ultraviolet light (Franke et al. 2017;Palau et al. 2017;Yu et al. 2015). Abiotic transformation of HBCDs could have been mediated by chemicals since the cultures were placed in the dark environment.
The augmentation of glucose increased bacterial diversity and improved the removal of HBCDs in the suspension of planted soil (Le et al. 2017). Biostimulation of carbon source could improve the removal of γ-HBCD in the sediment (Demirtepe and Imamoglu et al. 2019). Present research shew that the addition of nutrition enhanced the removal of HBCDs by stu-38 because bacteria might maintain high activity in the nutrient media. It suggested the augmentation of carbon source as a strategy to improve bioremediation in the contaminated sites. This was different from previous research using pure bacterial strain to transform 3-chlorobenzoate (Chobchuenchom et al. 1996).

The formation of the debrominated products
HBCDs could be chemically debrominated to tribromocyclododecadiene and dibromocyclododecadiene in the acetone (Fig. S2 and S6) (Zhong et al. 2010). At the same retention time, the intensity of tribromocyclododecadiene (m/z=401, 321, 241 and 159) and dibromocyclododecadiene (m/z=321, 241 and 159) were much smaller in the SW than in the acetone (Fig. S2). It was because of the high levels of chemicals from seawater and cells, which could result in the di culties for the identi cation of pentaand tetra-brominated products and the detection of α-and β-HBCD (Fig. S1, S7-S9). Moreover, the low transformation e ciency of strain stu-38 leaded to the debrominated products presented at low concentration and weak intensity ( Fig. S3-S5, S7-S9).
By using GC-MS, seven biodebrominated products were identi ed relying on mass spectra of GC-MS (Table 1; Fig. S3-S5, S6-S9). But the accurate identi cation and further determination were di cult because the limit of standards. Based on the debrominated products observed, the possible transformation pathways of HBCDs by strain stu-38 in the FW, SW and SWLB were proposed (Fig. 5).
The debromination pathways of HBCDs include HBr-elimination (dehydrobromination), HBrdihaloelimination and hydrolytic debromination (Ang et al. 2018;Yu et al. 2015). HBr-elimination of HBCDs yields lower brominated compounds with an HBr removed to form a carbon-carbon double bond (Kunze et al. 2017). Dihaloelimination involves electron transfer in which HBCDs serve as electron acceptor. Hydrolytic debromination yields lower brominated alkanol or alkenol. These debromination pathways were all observed in the transformation of HBCDs by R. sp. stu-38 (Fig. 5). As the toxicity of debrominated products were unknown, it was unclear if the toxicity of debrominated products resulted in the low e ciency of R. sp. stu-38 (Heeb et al. 2017;Lal et al. 2010;van Hylckama Vlieg et al. 2000).
Dibromocyclododecadiene, formed through HBr-dihaloelimination, were found in FW, SW and SWLB (Fig.  5). Full debromination via dihaloelimination and the cleavage of cyclododecatriene was not found in this research. The ring opening intermediate was observed in the study of strain GJY (Geng et al. 2019).
Strain HS9 ) and GJY (Geng et al. 2019) could convert HBCDs through both reductive and hydrolytic debromination yielding various intermediates. Stu-38 shew similar debromination patterns in the FW and SW, but the biotransformation of HBCDs in the SWLB was more complicated (Fig. 5, S5).

Conclusion
In summary, a new HBCDs-transforming bacteria Rhodococcus sp. stu-38 was identi ed in the present research. This strain selectively transformed HBCDs diastereoisomers in the mineral salt medium, seawater and nutrient seawater. Seven debrominated products were identi ed by using GC-MS. The formation of debrominated products were partially depended on the culture media. Together, this study demonstrated a functional Rhodococcus originated from mangrove sediment which could diastereoisomer-speci cally transform HBCDs depending on its living environment, which highlight the monitoring of various lower brominated products of contaminants during bioremediation.

Declarations
Authors' contributions Wenqi Luo, Xianbin Lin, Shanshan Meng, Lele Li performed the serial diluting culture with HBCDs supplied as the sole carbon source. Fei Yu isolated the bacteria, performed the rest experiment and completed the manuscript. Yuyang Li analyzed the genomic sequence of bacteria. Xueying Ye, Tao Peng, Hui Wang, Tongwang Huang, Zhong Hu provided the directions. All authors read and approved the nal manuscript.

Competing Interest
The authors declare that they have no known competing nancial interests or personal relationships that could have appeared to in uence the work reported in this paper. The HBCDs transforming rates of six isolated strains