Identification of an anaerobic bacterial consortium that degrades roxarsone

Abstract The degradation of roxarsone, an extensively used organoarsenic feed additive, occurs quickly under anaerobic conditions with microorganisms playing an important role in its degradation. Here, an anaerobic bacterial consortium that effectively degraded roxarsone was isolated, and its degradation efficiency and community changes along a roxarsone concentration gradient under anaerobic conditions were assessed. We used batch experiments to determine the roxarsone degradation rates, as well as the bacterial community structure and diversity, at initial roxarsone concentrations of 50, 100, 200, and 400 mg/kg. The results showed that roxarsone was degraded completely within 28, 28, 36, and 44 hr at concentrations of 50, 100, 200, and 400 mg/kg, respectively. The anaerobic bacterial consortium displayed considerable potential to degrade roxarsone, as the degradation rate increased with increasing roxarsone concentrations. Roxarsone promoted microbial growth, and in turn, the microorganisms degraded the organoarsenic compound, with the functional bacterial community varying between different roxarsone concentrations. Lysinibacillus, Alkaliphilus, and Proteiniclasticum were the main genera composing the roxarsone‐degrading bacterial community.

When the poultry litter is used as an organic fertilizer, arsenic pollution is likely to occur (Walrod, Burriss, Blue, Beck, & Atwood, 2016); for example, in 212 samples of animal manure-based compost collected in China, 13.7% exceeded arsenic limits (Yang et al., 2017). ROX is easily biodegraded to HAPA, but HAPA persists for long periods of time in the environment, increasing the risk of arsenic contamination (Shi, Wang, Yuan, & Hu, 2014;Zhang, Wang, Yuan, & Hu, 2014). ROX can be transferred from the diet of chickens to rice plants, and soil attributes govern the phytoavailability of ROX metabolites to rice plants (Yao et al., 2019(Yao et al., , 2016. Seasonal stratified analyses by poultry type strongly suggest that the historical use of arsenic-based poultry drugs contributes to arsenic exposure in the population of the United States (Nigra, Nachman, Love, Grau-Perez, & Navas-Acien, 2017). Similarly, in Chinese province of Guangdong, both ROX and inorganic arsenic are detected at elevated levels in the chicken tissues from live poultry markets, particularly liver, and the overall health risk from dietary exposure to inorganic arsenic associated with chicken consumption is rather high (Hu, Zhang, Cheng, & Tao, 2017). Therefore, litter applied to soil as organic fertilizer can release ROX and arsenic into soil and groundwater, affecting both the environment and human health (Chen, 2007;Fisher, Yonkos, & Staver, 2015;Jackson et al., 2003;Morrison, 1969;Oyewumi & Schreiber, 2017;Wang, Chen, Sun, Gao, & Yu, 2006a;Wang & Liao, 2005).
Because of the potential food safety and environmental risks, in 1999, 2011, and 2013, the European Union, Canada, and the United States, respectively, announced that they were scrapping or discontinuing ROX use. However, it continues to be used in many other countries (Fu, He, Gong, Blaney, & Zhou, 2016). In 2003, China used 1,200 tons of ROX between the livestock and poultry industry (Zhu, 2013), and together, these animals produced 3.19 billion tons of manure which could contribute to the release of arsenic species into the environment (Wang, Ma, et al., 2006b). The China pharmacopoeia bureau issued a strong suggestion in June 2017 discouraging the use of ROX, but most livestock and poultry breeding companies continue to use the chemical.
Degradation is faster in anaerobic conditions than in aerobic conditions, with ROX being completely degraded after 48 hr of dark anaerobic incubation, while only 79.9% and 94.5% were degraded after 288 hr of dark aerobic and light aerobic incubation, respectively (Liu, Zhang, Li, Wen, et al., 2017a). Light and microbial action are the main factors responsible for ROX degradation, which is also controlled by environmental factors such as moisture, temperature, and the organic content of the vadose zone (Fu, Blaney, & Zhou, 2017;Katherine et al., 2017;Sun, 2012). Some ROX-related bacteria have been reported (Stolz et al., 2007); however, the ROX-degrading bacterial community of anaerobic conditions and its degradation ability at different ROX concentrations have not been completely described.
This paper isolated a stable anaerobic bacterial consortium that effectively degraded ROX and evaluated its response to the initial ROX concentration through high-throughput sequencing.
Meanwhile, the ROX degradation rates of the stable bacterial consortium at different ROX concentrations under anaerobic conditions were determined. Ultimately, the relationship between ROX and the bacteria was revealed.

| Poultry litter samples
The poultry litter used for experiments was taken from a chicken farm (36°07′N, 115°49′E) in Yanggu County, northern Shandong plain, China. The ROX concentration of the poultry diet (34 mg/ kg) was within the recommended concentration range (50 mg/kg).
There were approximately 15,000 broiler chickens on the farm, and the poultry litter was perennially stored in the open air and regularly collected to use as fertilizer. The collected litter for experiments was stored at 4°C until further analysis. The collected litter was determined to have a concentration of 11.3 mg/kg of HAPA and no ROX, likely because of the rapid degradation rate of ROX.

| Enrichment of microbial communities degrading ROX
The basal medium consisted of (per L) 4.2 g of NaHCO 3 , 0.095 g of MgCl 2 , 0.5 g of yeast extract, 10 mM lactate (Stolz et al., 2007), 10 ml of trace elements, and a vitamin mix. The pH was adjusted to 7.3. To prepare the slurry used for experiments, 5 g of chicken litter was suspended in 100 ml of sterile basal medium.
The mixed solution was dispensed into a 50 ml headspace bottle containing 40 ml of medium, 1 ml of litter slurry, and 200 mg/kg ROX, aerated with oxygen-free N 2 , sealed, and incubated at 37°C in the dark. All experiments were performed in triplicate. The solution changed from yellow to colorless after 24 hr, and 1 ml of the solution was then transferred into a new 50-ml headspace bottle containing 40 ml of medium and 200 mg/kg ROX. This solution was then incubated for 24 hr under the same culture conditions. The above steps were repeated 10 times to attain a stable microbial community that effectively degraded ROX, which was then used as the inoculum in the next experiment.

| ROX degradation along a concentration gradient under anaerobic conditions
To evaluate the ability of the microbial community to degrade ROX along a concentration gradient, a 150-ml headspace bottle was filled with 150 ml of basal mineral medium and 1 ml of a mixed bacteria solution, which was obtained from the above experiment.
Four ROX concentrations (50, 100, 200, and 400 mg/kg) were examined in batch experiments. In addition, a blank control was prepared without the addition of ROX, which was compared with the samples containing ROX to determine its influence on the diversity and abundance of the microbial community. Each test was conducted in triplicate. The medium and headspace were aerated with oxygen-free N 2 , sealed, and incubated at 37°C in the dark.
Aliquots (1 ml) were taken every 4 hr and filtered through a 0.22µm millipore membrane for ROX analysis. Aliquots (1 ml) were taken every 2 hr to assess cell growth by measuring the optical density at 600 nm.
At the end of the anaerobic incubation period, the microorganisms from different ROX concentrations were harvested by filtering the solution through a 0.22-µm millipore membrane. Then, the DNA was extracted for high-throughput sequencing.

| Kinetic model
First-order kinetic models are often used to describe the degradation of organic compounds (Gao, Zhang, Chen, Zheng, & Liu, 2010). Therefore, the biotransformation of ROX was characterized by a first-order kinetic model.

| Amplification and sequencing of bacterial 16S rRNA genes
High-throughput sequencing was conducted at Shanghai Personal Biotechnology Co., Ltd. Total bacterial genomic DNA samples were extracted using Fast DNA SPIN extraction kits (MP Biomedicals), following the manufacturer's instructions, and stored at −20°C prior to further analysis. The quantity and quality of extracted DNA were measured using a NanoDrop ND-1000 spectrophotometer (Thermo Fisher Scientific) and agarose gel electrophoresis, respectively. PCR amplification of the bacterial 16S rRNA genes V4-V5 region was performed using the forward primer 515F (5′-GTGCCAGCMGCCGCGGTAA-3′) and the reverse primer 907R (5′-CCGTCAATTCMTTTRAGTTT-3′).
Sample-specific 7-bp barcodes were incorporated into the primers for multiplex sequencing. PCR amplicons were purified with Agencourt AMPure Beads (Beckman Coulter) and quantified using the PicoGreen dsDNA Assay Kit (Invitrogen). After the individual quantification step, amplicons were pooled in equal amounts, and pair-end 2 × 300 bp sequencing was performed using the Illlumina MiSeq platform with the MiSeq Reagent Kit v3.

| High-throughput sequence analysis
The Quantitative Insights Into Microbial Ecology (QIIME, v1.8.0) pipeline was employed to process the sequencing data, as previously described (Caporaso et al., 2010). Briefly, raw sequencing reads with exact matches to the barcodes were assigned to respective samples and identified as valid sequences. Low-quality sequences were excluded. Paired-end reads were assembled using FLASH (Magoc & Salzberg, 2011). After chimera detection, the remaining high-quality sequences were clustered into operational taxonomic units (OTUs) at 97% sequence identity by UCLUST (Edgar, 2010). The most abundant sequence of each cluster was picked to be the representative sequence. OTU taxonomic classification was conducted by BLAST searching the representative sequences set against the Greengenes Database (DeSantis et al., 2006). An OTU table was further generated to record the abundance of each OTU in each sample and the taxonomy of these OTUs.
The taxonomic compositions and abundances were visualized using MEGAN (Huson, Mitra, Ruscheweyh, Weber, & Schuster, 2011) and F I G U R E 1 ROX degradation and the first-order fitting curves of different ROX concentrations under anaerobic conditions GraPhlAn (Altschul et al., 1997). A Venn diagram was generated to visualize the shared and unique OTUs among samples using R package "VennDiagram," based on the occurrence of OTUs across samples/ groups regardless of their relative abundance (Zaura, Keijser, Huse, & Crielaard, 2009). Partial least squares-path modeling (PLS-PM) was used to explore the relationships between bacterial communities and ROX concentration by using the R package (Sanchez, 2013).  Table 1. At concentrations of 50, 100, 200, and 400 mg/kg, ROX was degraded completely within 28, 28, 36, and 44 hr, respectively, and the half-lives were 8.9, 8.8, 7.3, and 11.55 hr, respectively. The marginally longer half-life at 400 mg/L when compared to the other concentrations indicates that, under anaerobic conditions, ROX was degraded more rapidly at higher initial concentrations. There were no significant differences between the degradation rates of different concentrations of ROX over time (p > .05; Figure 1), when compared with a one-way analysis of variance (ANOVA) using SPSS Statistics for Windows (Fu et al., 2017). This indicates that increasing the ROX concentration had minimal effect on ROX degradation and further illustrates that anaerobic microbes possess great potential to degrade ROX.

| ROX degradation at different concentrations
The microorganisms in all media grew rapidly, and a lag period was hardly observed. Cell numbers increased exponentially, reaching the stationary phase by 20 hr (Figure 2). The ROX-free medium had the lowest microbial biomass, with microorganisms growing faster as ROX concentrations increased. The cell density at 400 mg/kg ROX was nearly threefold higher than that at 0 mg/kg. These results revealed a relationship between ROX and the microorganisms. The presence of ROX did not inhibit microbial growth; instead, it promoted microbial growth, and in turn, the microorganisms degraded ROX.

| Alpha diversity of the bacterial community
Analysis by high-throughput sequencing resulted in 60,282 highquality, nonplastid, and partial sequences were obtained for the five samples after quality control. The alpha diversity indices are shown in

| Bacterial community composition and relative abundances
There were seven phyla identified in the five samples, with Firmicutes being the major phylum in each sample, with relative abundances of 96.70%-99.85% (Figure 5a). The bacterial community composition and relative abundance of each sample at the genus level (Figure 5b)

| D ISCUSS I ON
The results of ROX biodegradation at different concentrations illustrated that anaerobic microbes possess great potential for this process. Similar results were obtained in a study of ROX degradation at different ROX concentrations (0, 50, 100, and 200 mg/kg) in soil, and it was found that the degradation rate increased with increasing concentrations, further highlighting the considerable potential of soil microbes to degrade ROX (Liu, Zhang, Li, & Fei, 2017b). The presence of ROX promoted microbial growth, and in turn, the microorganisms degraded ROX. These results are consistent with previous observations in which the addition of ROX did not inhibit, but rather stimulated, the growth of the anaerobe Shewanellaoneidensis MR-1 (Chen, Ke, Liang, Liu, & Wang, 2016). A previous study also showed that an isolated aerobic bacterial consortium degraded ROX and that the growth rate of the aerobic bacterial consortium was 1.4-fold higher in the presence of ROX (Guzmán-Fierro et al., 2015). ROX has a hazardous effect on the native microbial community diversity and metabolic activity of soil and significantly affects overall microbial activity in soil (Jiang, Li, Wang, Li, & Wang, 2013;Mangalgiri, Adak, & Blaney, 2015). An analysis of fluorescein diacetate hydrolysis in soil found that ROX does not exert acute toxicity on soil microbes; however, fluorescein diacetate hydrolysis was inhibited gradually with the release of inorganic arsenic (Liang, Ke, Chen, Liu, & Chen, 2014). Measuring the half-maximal inhibitory and half-maximal effective concentrations of the As(V)-and As(III)-bearing photodegradates of ROX exhibited 10-fold higher toxicity than ROX itself, which was attributed to the improved membrane permeability of the inorganic arsenicals (Zhang, Xu, Han, Sun, & Yang, 2015). The toxicity for eukaryotic or prokaryotic cells is primarily caused by ROX biodegradation by microorganisms (Mafla et al., 2015).
Therefore, in the present study, the inhibition of some soil bacteria might be from the arsenic compounds released following ROX degradation. The increasing ROX concentrations enriched the bacteria that metabolized ROX and promoted the growth of these same bacteria. higher concentrations led to arsenic accumulation and a dramatic decrease in the efficiency of its adsorption (Mohamed & Farag, 2015). Arsenic stress causes negative cell responses, such as a decrease in surface area and shrinking. The Lysinibacillus strain B1-CDA showed potential to bioremediate arsenic compounds from contaminated water, forming a long chainlike structure following arsenic exposure (Rahman et al., 2014). The carboxyl groups of the glutamic acid residues in peptidoglycan are the major sites of metal deposition (Zouboulis, Lazaridis, Karapantsios, & Matis, 2010).
Alkaliphilus oremlandii sp. nov. strain OhILAs is capable of transforming ROX (Fisher et al., 2008). In previous studies, Alkaliphilus was shown to be closely correlated with the presence of arsenic and ROX (Liu, Zhang, Li, & Fei, 2017b), which aligns with our own results showing Alkaliphilus to be the predominant genus of the bacterial community that degraded ROX. However, to our knowledge, we are the first to report the relationship between Proteiniclasticum and arsenic, although further studies are needed to reveal the function of Proteiniclasticum in arsenic transformation. Second, the presence of ROX promoted the growth of functional microorganisms, and microorganisms grow faster as ROX concentrations increased from 0 mg/kg to 400 mg/kg, resulting in a nearly 3-fold increase in cell density from samples with no ROX to 400 mg/kg ROX. Moreover, a stable anaerobic bacterial consortium that effectively degraded ROX was identified, and it is the first to report the relationship between Proteiniclasticum and arsenic. Oliver from CSIRO in Australia for their valuable suggestions for improving the manuscript.

CO N FLI C T O F I NTE R E S T S
None declared.

E TH I C S S TATEM ENT
None required.

DATA AVA I L A B I L I T Y S TAT E M E N T
All data generated or analyzed during this study are included in this