The metabolism and morphology mutation response of probiotic Bacillus coagulans for lead stress

Science of the Total Environment 693 (2019) 133490

Contents lists available at ScienceDirect

Science of the Total Environment journal homepage: www.elsevier.com/locate/scitotenv

The metabolism and morphology mutation response of probiotic Bacillus coagulans for lead stress Si-Cheng Xing a,b,c,1, Jian-Dui Mi a,b,c,1, Jing-Yuan Chen a,b,c, Lei Xiao a,b,c, Yin-Bao Wu a,b,c, Juan Boo Liang d, Lian-Hui Zhang e,⁎, Xin-Di Liao a,b,c,⁎⁎ a

College of Animal Science, South China Agricultural University, Guangzhou 510642, Guangdong, China Guangdong Provincial Key Lab of Agro-Animal Genomics and Molecular Breeding, Key Laboratory of Chicken Genetics, Breeding and Reproduction, Ministry Agriculture, Guangzhou, 510642, Guangdong, China c National-Local Joint Engineering Research Center for Livestock Breeding, Guangzhou, 510642, Guangdong, China d Laboratory of Sustainable Animal Production and Biodiversity, Institute of Tropical Agriculture and Food Security, Universiti Putra Malaysia, Serdang 43400, Malaysia e Guangdong Province Key Laboratory of Microbial Signals and Disease Control, State Key Laboratory for Conservation and Utilization of Subtropical AgroBioresources, College of Agriculture, South China Agricultural University, Guangzhou 510642, China b

H I G H L I G H T S

G R A P H I C A L

A B S T R A C T

• Genes involved in flagellar formation pathways in the lead-adapted strains upregulated. • B. coagulans R11 mutants tended to form flagellar and chemotaxis systems to avoid lead ions rather than export it. • Lead-adapted strains shown higher lead intracellular accumulation ability and reducing ability than that of wild type. • The flagellar motion system mutants of B. coagulans R11 formation needs exposed in lead for long time. • B. coagulans R11 are potential genetic engineering candidates for synthesizing glutathione and selenocompounds.

a r t i c l e

i n f o

Article history: Received 16 January 2019 Received in revised form 10 May 2019 Accepted 18 July 2019 Available online 19 July 2019 Editor: Xinbin Feng Keywords: Bacillus coagulans Lead Morphology Genomics

a b s t r a c t Lead is among the most common toxic heavy metals and its contamination is of great public concern. Bacillus coagulans is the probiotic which can be considered as the lead absorption sorbent to apply in the lead contaminant water directly or indirectly. A better understanding of the lead resistance and tolerance mechanisms of B. coagulans would help further its development and utilization. Wild-type Bacillus coagulans strain R11 isolated from a lead mine, was acclimated to lead-containing culture media over 85 passages, producing two leadadapted strains, and the two strains shown higher lead intracellular accumulation ability (38.56-fold and 19.36-fold) and reducing ability (6.94-fold and 7.44-fold) than that of wild type. Whole genome sequencing, genome resequencing, and comparative transcriptomics identified lead resistance and tolerance process significantly involved in these genes which regulated glutathione and sulfur metabolism, flagellar formation and metal ion transport pathways in the lead-adapted strains, elucidating the relationships among the mechanisms regulating lead deposition, deoxidation, and motility and the evolved tolerance to lead. In addition, the B. coagulans mutants tended to form flagellar and chemotaxis systems to avoid lead ions rather than export it,

⁎ Corresponding author. ⁎⁎ Correspondence to: X.-D. Liao, College of Animal Science, South China Agricultural University, Guangzhou, 510642, Guangdong, China. E-mail addresses: [email protected] (L.-H. Zhang), [email protected] (X.-D. Liao). 1 These authors contributed equally to the paper as first authors.

https://doi.org/10.1016/j.scitotenv.2019.07.296 0048-9697/© 2019 Elsevier B.V. All rights reserved.

2

S.-C. Xing et al. / Science of the Total Environment 693 (2019) 133490

Metabolism Evolution

suggesting a new resistance strategy. Based on the present results, the optimum lead concentration in environment should be considered when employed B. coagulans as the lead sorbent, due to the bacteria growth ability decreased in high lead concentration and physiology morphology changed could reduce the lead removal effectiveness. The identified deoxidization and compound secretion genes and pathways in B. coagulans R11 also are potential genetic engineering candidates for synthesizing glutathione, cysteine, methionine, and selenocompounds. © 2019 Elsevier B.V. All rights reserved.

1. Introduction Lead is one of the most common toxic heavy metals on the earth, and lead contamination is of great public concern, in particular, in developing countries, as lead is utilized in various industrial and agricultural products, including batteries, dyes, glass, plastics and petroleum (Florea et al., 2005; Zoghi et al., 2014). Lead ion can easily dissolve in the water while it was used by above mentioned agents, thus lead ions could enter human and animal bodies through food chain by the contaminant water. Once inside the body, the metal is hard to get rid of because it is nondegradable and can easily deposit in organs and bones. Consequently, the accumulated lead causes intoxication and damages to the nervous, haematopoietic, renal, endocrine, and skeletal systems of exposed subjects (Fu and Wang, 2011; El-Sayed, 2013; Lei et al., 2014), especially infants and children. Lead contamination and intoxication often occur in developing countries due to poor understanding of lead dangers and lack of preventive and protective measures (Mohiuddin et al., 2011; Hezbullah et al., 2016; Wu et al., 2018). These facts stress the importance to implement preventive measures to reduce lead contamination in the environment, and to develop lead intoxication remedies. Over the last few decades, remediation of heavy metals, including lead, in the environment and in organisms has attracted increased interest in fundamental and application research (Yang et al., 2016; Statham et al., 2016; Yang et al., 2017; Tian et al., 2012; Zhai et al., 2013). Remediation methods include chemical, physical and biological techniques. Among them, biological remediation is regarded as the safest and most cost-effective method because other methods are associated with high costs or potential side effect on environment. Microorganisms, which are abundant in our ecosystem (Trevors, 1999), are one of the tools used in biological remediation. Various studies have elucidated the lead-binding effects that occur during the utilization of microorganisms for remediation (Jin et al., 2017; Naik et al., 2012; Jahromi et al., 2017; Waite et al., 2016). At the same time, metal ion binding mechanisms have also been characterized by numerous studies. In general, microbes remove metals from surrounding environment through the following processes: 1) metal cations bind to cell wall surfaces; 2) metal ions are actively translocated into the cells by metal binding proteins; 3) metal precipitation reacts with extracellular polymers or anions produced by microbes; and 4) metal volatilization occurs by enzyme-mediated biotransformation (Ahemad and Kibret, 2013; Giller et al., 2009; Hughes and Poole, 1989). Metal removal capacity of microorganisms appears to be associated with their metal resistance or tolerance abilities because the high heavy metal concentration in contaminant sites can threaten the survival of microorganisms (Oladipo et al., 2018; Piotrowska-Seget et al., 2005; Hassen et al., 1998). Hence, the microorganisms with high levels of heavy metal tolerance could be potent candidates for pollution remediation applications. In recent years, a number of lead-resistant bacterial strains have been isolated to assess their lead removal capacity and lead resistance and tolerance mechanisms for further utilization or research (Yin et al., 2016; Miao et al., 2018; Li et al., 2017; Ren et al., 2015; Qu et al., 2018). Most of these studies focused mainly on the lead deposition ability of microorganisms, but lacked specific and systemic characterization of resistance and tolerance mechanisms.

B. coagulans is a certified probiotic, and has been used in fermentation industry and as a feed additive (Orrù et al., 2014; Salvetti et al., 2016). The previous studies not only focused on its L-lactate yield production properties and its ability to enhance livestock growth, but also some studies have employed B. coagulans as a heavy metal (including lead) sorbent in waste water (Lei et al., 2014; Quintelas et al., 2008). Due to the characters of B. coagulans, i.e., its safety, easy storage and tolerance of unfavourable environments, B. coagulans could be considered the strain with most potential for heavy metal remediation for future applications. Thus, the lead resistance and tolerance of B. coagulans could make it an optimal candidate as a lead remediation tool. To better employ B. coagulans as a lead sorbent in the environment even in animal and human bodies, it is necessary to gain a better understanding of its lead resistance and tolerance mechanisms, as these linkage mechanisms will promote the development of products and the further utilization of genetic resources. Consequently, we isolated one lead-tolerant B. coagulans from a lead mine as the model bacterial strain, labelled B. coagulans R11 (GenBank association number was BanKIt2035011 Bacillus MF539755). In this study, the wild-type strain was acclimated to different lead concentrations over 85 passages, and two lead-adapted strains were produced. Here, we assessed the total removal, extracellular accumulation and intracellular accumulation capacity of lead of the wild-type and acclimated strains. We also tested the reducing ability and glutathione yield of the three bacterial strains. In prior experiments, we found that the colonies of a high lead-adapted strain were significantly different from those of the wild type. Although the lead accumulation and apparent resistance phenomena of B. has been demonstrated, the connection between its lead accumulation capacity and resistance and tolerance abilities remains elusive. Hence, we employed comparative omics in wild-type and mutation strains to identify the possible mechanisms of lead resistance and tolerance in B. coagulans R11. To confirm the conference genome veracity, we sequenced the whole genome of the B. coagulans R11 wild type. Next, the wild-type and acclimated strains were incubated in pure culture medium and lead-containing culture medium to better stimulate and compare the regulation of key genes and associated pathways. The DNA and RNA of the tested bacteria were extracted for genome resequencing and transcriptomics. The results of the present study provide potential lead resistance and tolerance mechanisms for further studies. Moreover, this study identified promising genes or pathways that could be potential genetic engineering candidates for the production of glutathione, cysteine, methionine, and selenocompounds. The genetic engineering of bacteria to be employed in lead remediation could also use the functional genes identified in this study to produce bacteria that tolerate high concentrations of lead in the environment. 2. Results 2.1. Lead resistance acclimatization and growth curve To obtain derivatives with enhanced level of lead tolerance, B. coagulans R11 was inoculated in the media supplemented with 100 mg/L and 1000 mg/L lead, respectively, and subcultured every 24 h for about 85 passages till the bacterial cell density (OD600)

S.-C. Xing et al. / Science of the Total Environment 693 (2019) 133490

3

Fig. 1. Growth curve of wild type and acclimatization bacteria.

remained constant. Then, a lead-adapted bacterial colony from each treatment was selected randomly from the lead-containing (100 mg/L and 1000 mg/L) agar plate, designated as strains Pb100 and Pb1000, respectively, and maintained in enrichment medium for further tests. The estimated cumulative cell division was approximately 1.21 × 109 CFU/mL for the WT strain, 2.95 × 108 CFU/mL for the Pb100 strain, and 2.70 × 108 CFU/mL for the Pb1000 strain. The growth curve analysis showed that the OD600 of strainsPb100 and Pb1000 were both lower than that of the WT strain in the stationary phase, but the growth rate of strain Pb100 was the highest in the log phase among all three bacterial strains (Fig. 1). In addition, the stationary phase of strain Pb100 began at 16 h, whereas for WT strain R11 and strain Pb1000, the stationary phase began at 22 h (Fig. 1). After 44 h, the OD600 readings of all three strains were decreased sharply (Fig. 1), likely due to cell death and lysis.

intracellular accumulation capacities among the WT and lead-adapted strains were significantly different at all the tested lead concentrations with WT strain being the lowest. Significantly, the intracellular accumulation capacities of both lead-adapted bacterial strains were 100-fold higher than that of the WT strain when the initial lead concentration was 20 mg/L, and the trend of total lead removal capacity was in accordance with the trend of the intracellular lead accumulation capacity when the lead concentration was 60 mg/L, 80 mg/L, and 100 mg/L, respectively. However, the intracellular capacities of bacterial strains were not as high as their extracellular accumulation capacities, except for in the strain Pb100 when the lead concentration was 100 mg/L (7.06 mg/g intracellular accumulation capacity vs 3.74 mg/g extracellular accumulation capacity). Taken together, the above results suggest that lead adaptation significantly enhanced the bacterial lead tolerance, which is mainly due to the much increased intracellular lead accumulation capability of the adapted strains.

2.2. The lead-adapted strains showing increased lead accumulation capacity To determine the lead accumulation capacity associated with lead resistance and tolerance, the WT and adapted strains were incubated in medium with 20 mg/L, 40 mg/L, 60 mg/L, 80 mg/L, and 100 mg/L of lead, respectively, to quantify the total lead removal and extracellular and intracellular lead accumulation capacities. The total lead removal capacity of Pb100 was significantly higher than that of WT strain when lead concentration exceeding 60 mg/L, for example, 9.93 mg and 7.00 mg of lead were removed from culture medium by per g of strain Pb100 and WT cells, respectively, at a lead initial concentration of 60 mg/L, but there was no obvious difference between the lead removal capacity of strains Pb100 and Pb1000. Similarly, there was no obvious difference between the lead removal capacity of the WT and strain Pb1000 (Fig. 2A). In contrast, assay on extracellular lead accumulation capacities showed no obvious differences among the tested strains, except when the initial lead concentration was 20 mg/L or 100 mg/L (Fig. 2B). However, as shown in Fig. 2C, the

2.3. B. coagulans R11 and its derivatives are varied in reducing ability and glutathione yield Excessive accumulation of heavy metals in cells is known to cause oxidative damages. We speculated that the lead-adapted strains might have evolved deoxidization systems to prevent oxidative stresses resulted from increased lead accumulation. Glutathione is one of the crucial compounds for deoxidization and detoxification in cells, participating in the transformation of heavy metal ions into non-toxic forms (Wang et al., 2018). To compare the detoxification capacity of the WT and acclimated bacterial strains, the glutathione yield and reducing ability were tested. As expected, the reducing abilities of the lead-adapted bacterial strains were significantly increased by approximately 7-fold, compared to that of the WT strain (Fig. 3A). However, the reducing abilities of the Pb100 and Pb1000 strains were almost same, with values of approximately 1.11 μg.cys/mL and 1.19 μg.cys/ mL, respectively (results based on the representation by cysteine).

Fig. 2. Lead accumulation capacity of wild type and acclimatization bacteria. Fig. 2A was total lead removal capacity of wild type and acclimatization bacteria, Fig. 2B was extracellular lead binding capacity of native and acclimatization bacteria, Fig. 2C was intracellular lead accumulation capacity of wild type and acclimatization bacteria, “*” means significantly different (Pb0.05), “**” means extremely significant difference (Pb0.01).

4

S.-C. Xing et al. / Science of the Total Environment 693 (2019) 133490

Fig. 3. Reducing ability and glutathione yield of native and acclimatization bacteria. Fig. 3A was reducing ability of wild type and acclimatization bacteria, Fig. 3B and C were intracellular and extracellular glutathione yield of wild type and acclimatization bacteria respectively. “*” means significantly different (Pb0.05), “**” means extremely significant difference (Pb0.01).

Addition of lead in the culture medium significantly increased the intracellular glutathione yields of WT and Pb1000 strains by about 3.3-fold and 10-fold, respectively (Fig. 3B), which is in prefect accordance with their relative reducing ability (Fig. 3A). In contrast, however, addition of lead in culture medium did not cause any obvious changes in the intracellular glutathione yield of strain Pb100. Therefore, we further determined the extracellular yields of the three bacterial strains. The results showed that the extracellular glutathione level of WT and strain Pb1000 remained unchanged with or without lead in culture medium, but the extracellular glutathione yield of strainPb100 was decreased by about 18.0% in the lead-containing broth compared to that in the

absence of lead (Fig. 3C). Taken together, these results suggest that the increased reducing ability of strain Pb1000 is correlated well with the increased production of glutathione, but other factor(s) other than glutathione is responsible for the enhanced reducing ability of strain Pb100. 2.4. B. coagulans R11 and its derivatives may form lead crystals to reduce lead biotoxicity Previous studies have reported that some bacterial strains can reduce heavy metal biotoxicity via sulfur metabolism. For instance,

Fig. 4. X-ray polycrystalline diffractometer analysis. Fig. 4A, C, and E were control groups for wild type, Pb100 and Pb1000 respectively, Fig. 4B, D, and E were lead treatment groups for wild type, Pb100 and Pb1000 respectively.

S.-C. Xing et al. / Science of the Total Environment 693 (2019) 133490

5

Fig. 5. Colony morphology modification analysis. Fig. 5A was wild type, 5B was acclimatization bacteria Pb100 and 5C was acclimatization bacteria Pb1000.

sulfate-reducing bacteria reduce sulfate to dissociate HS− to form metal sulfide (Niu et al., 2018; Li et al., 2018). To investigate whether the lead resistance mechanisms in B. coagulans strains involve formation of leadsulfide (PbS), X-ray polycrystalline diffractometer (XRD) was used for assessment of PbS crystals. However, the results did not clearly show the formation of PbS crystals (Fig. 4, green line refers to the peak of standard PbS) after incubation of bacterial strains in the lead-containing medium, because of the peaks was not all same as the standard PbS. In a comparison of the XRD curves between the control group (Fig. 4A, C, E) and lead treatment group (Fig. 4B, D, F), there were no significant spiculate peaks in the control groups, but significant spiculate peaks did form in the treatment groups, especially for the Pb1000 strain, and these curve peaks were all somewhat similar to the standard curve peaks of PbS. These results could mean that when cultured in the lead-containing medium, bacterial cells might form certain lead compound crystals through a detoxification mechanism similar to formation of PbS crystals. The chemical nature and properties of these putative lead crystal compounds require further characterizations. 2.5. The lead-adapted trains show changed colony morphology It was clearly observed that the bacterial colonies could change in response to lead acclimatization in our study. The colonies of the WT strain were smooth, stereoscopic and mellow (Fig. 5A), but the colonies of strain Pb1000 were flat and rough, and the colony dimension was increased compared with WT (Fig. 5C). In contrast, compared with the WT strain (Fig. 5A), the morphology of most the Pb100 colonies on the agar did not seem to be changed except a few (Fig. 5B). This suggests that the colony morphology change is a lead dosage dependent process. 2.6. Lead stress caused substantial genome variations To elucidate the potential lead resistance mechanisms, we sequenced the whole genome of the WT strain as the reference genome

for further resequencing and transcriptomics analyses. The whole genome sequence of B. coagulans R11 has been uploaded to NCBI database (Accession number: CP026649). The genome annotation was conferenced by KEGG, COG and GO database. By comparison of the reference genome with the genomes of the acclimated strains, we found 11 common single nucleotide polymorphisms (SNPs) and 97 insertion/deletions (Indels) in strainsPb100 and Pb1000 strains. In addition, two SNPs were synonymous mutations. The SNP and Indel information was neatened and is presented in Tables 1A, 1B, and 1C. The amount of SNPs for the Pb100 strain was higher than that for the Pb1000 strain. The total numbers of SNPs for strains Pb100 and Pb1000 were 58 and 42, respectively. Among them, there were 44 and 33 SNPs in CD regions for Pb100 and Pb1000, respectively (Supplementary Table S1). In Table 1A, the SNPs were found in the genes R11. WTGM000628 and R11.WTGM003998, which were not annotated in any of the common databases. The SNPs in R11.WTGM003884 and R11. WTGM002061 were synonymous mutations, and the affected genes encoding amino acids valine and threonine. One SNP occurred in the gene (ID: R11.WTGM003068) encoding a transcriptional regulator, which functions as a transmembrane anti-sigma factor to regulate bacterial adaptation and survival in changing environment. The total numbers of Indels for strains Pb100 and Pb1000 were 173 and 172, respectively, and among them, 97 were common Indels (data not show). Most of these Indels were insertions with only 31 and 34 deletions being found for strains Pb100 and Pb1000, respectively. We listed in Table 1B the 4 Indels that occurred in the genes encoding transcriptional regulators and sensors, including R11.WTGM000675 and R11. WTGM004046, encoding the ARAC family and GNTR family transcriptional regulators, and R11.WTGM003198 and R11.WTGM004137, encoding a histidine sensor kinase and a sensor protein LYTS, respectively. In addition, there was a mass of common Indels that were not annotated by any database; these mutations may indicate unknown mechanisms involved in lead tolerance of B. coagulans R11derivatives.

Table 1A SNP mutation summary. Gene ID R11.WTGM000628 R11.WTGM001151 R11.WTGM001655 R11.WTGM002808 R11.WTGM002999 R11.WTGM002441 R11.WTGM003068 R11.WTGM003884 R11.WTGM002061 R11.WTGM003998 R11.WTGM004027

Gene tag

WT

SNP

Amino acid mutate

Annotation

hyuA yhaA cheC sftA – rsmE NA – – mbl ywlC

GTG GCC AAT GCA GCG CGC CTC GTC ACC CCA GCG

GCG GTC TAT ACA ACG CAC CCC GTT ACT TCA GTG

V563 to A A1028 to V N58 to Y A28 to T A10 to T R686 to H L404 to P V51 to V T207 to T P406 to S A221 to V

Hydantoinase/oxoprolinase Catalyzes the cleavage of p-aminobenzoyl-glutamate to p-aminobenzoate and glutamate, subunit A Chemotaxis protein CheC – inhibitor of MCP methylation Cell division protein FtsK Spore germination protein GerKB Ribosomal RNA small subunit methyltransferase E Transmembrane anti-sigma factor Hypothetical protein Acetylglutamate kinase NA tRNA threonylcarbamoyladenosine biosynthesis protein

6

S.-C. Xing et al. / Science of the Total Environment 693 (2019) 133490

Table 1B Indel mutation summary. Gene ID R11.WTGM000675 R11.WTGM003198 R11.WTGM004046 R11.WTGM004137

Gene tag

Indel type

Indel sequence

Annotation

– ykoH – –

Insertion Insertion Insertion Insertion

C G C G

Transcriptional regulator, AraC family Integral membrane sensor signal transduction histidine kinase Transcriptional regulator, GntR family Sensor protein LytS

For those mutations affecting physiological functions, the odds ratio (OR) was used to calculate the gene enrichment with the COG database, and an OR N 2 was the standard value to adjust the enrichment. The enriched genes were listed in Table 1C, which showed that Intracellular trafficking, secretion and vesicular transport, defense mechanisms and extracellular structure were the three common categories enriched in mutations in the lead-acclimated bacterial strains. These results suggest that the lead stress caused various bacterial responses including cell wall modifications, intracellular transport and secretion systems. In addition, mutations in bacteria cell growth, development and motility categories were found only in strain Pb1000, which may indicate that high lead stress could stimulate bacteria to evolve more physiological features to adapt to changed environmental conditions.

2.7.1. Lead stress stimulated upregulated genes in WT and lead-adapted strains The lead stress response of the lead-adapted and WT strains were investigated by incubation of bacterial cells in medium with or without lead for 24 h. For lead treatment, WT and strain Pb100 were both incubated in MRS broth supplemented with 100 mg/L and strain Pb1000 was cultured in MRS broth with 1000 mg/L lead. Three bacterial strains were also inoculated in pure MRS broth and cultured under the same conditions as a control. Three bacteria strains all differently replied the lead stress by upregulated genes in nucleic acid protection and lead resistance perspective. As the results, the upregulated genes which involved in the bacteriaself nucleic acid protection, one group of the up-regulated functional genes is related to DNA protection, as oxidative stress induced by heavy metals can damage nucleic acids in cells (Yepiskoposyan et al., 2006; Mikowska and Świergosz-Kowalewska, 2018; Qing et al., 2017). Interestingly, there were three pathways involved in DNA protection were upregulated in WT strain (Mismatch repair, Homologous recombination, Nucleotide excision repair) but only one DNA protection pathway was upregulated in strain Pb100 (Base excision repair). These results may suggest either that DNA damage occurred more easily in WT than in strain Pb100 or that the DNA damage was more serious in WT than in strain Pb100 or both. The upregulated pathways for the WT strain were nucleotide excision repair, mismatch repair, and base excision repair, while only the base excision repair pathway was upregulated in the strain Pb100. To our surprise, there were three pathways related to DNA protection were upregulated in strain Pb1000 (Mismatch repair, Base excision repair, Homologous recombination), which were highly same as those upregulated pathways in the WT strain except the nucleotide excision repair pathway and base excision

2.7. Transcriptomics analysis of the WT and lead-adapted strains Genome resequencing of the lead-adapted strains unveiled substantial genomic variations by comparision of the genome sequence of WT strain B. coagulans R11. To understand how these variations could affect the expression of functional genes and to elucidate the bacterial lead resistance or tolerance mechanisms, transcriptomics analysis was conducted to identify the upregulated and downregulated genes in the WT and lead-adapted bacterial strains. For the convenience of analysis, we summarized the upregulated genes which involved in the bacteriaself nucleic acid protection process in Table 2, and lead resistance involving upregulated genes data in Fig. 6, lead resistance genes upregulated level listed in (Supplementary Table S2), in addition, lead resistance genes downregulated level was listed in (Supplementary Table S3).

Table 1C Mutation genes enrich physiology function. “Y” means “yes” and “N” means “no”. Annotation

Chromatin structure and dynamics Energy production and conversion Cell cycle control, cell division, chromosome partitioning Amino acid transport and metabolism Nucleotide transport and metabolism Carbohydrate transport and metabolism Coenzyme transport and metabolism Lipid transport and metabolism Translation, ribosomal structure and biogenesis Transcription Replication, recombination and repair Cell wall/membrane/envelope biogenesis Cell motility Posttranslational modification, protein turnover, chaperones Inorganic ion transport and metabolism Secondary metabolites biosynthesis, transport and catabolism General function prediction only Function unknown Signal transduction mechanisms Intracellular trafficking, secretion, and vesicular transport Defense mechanisms Extracellular structures Mobilome: prophages, transposons Cytoskeleton

Pb100 OR

P

0 0.7404 1.4268 1.2151 0.8416 1.101 0.7796 1.1306 0.7287 0.5418 1.0037 1.1216 1.3994 1.1083 0.5243 0.5044 1.3655 0.9802 1.1216 2.8814 2.0744 9.3116 0 0

0.3172 0.0021 0.003 0.0019 0.1487 0.1474 0.011 0.1532 0.0002 0 0.03 0.1818 0.005 0.2968 0 0 0 0.1842 0.1818 0 0 0 0 0.3172

Pb1000 Enrichment N N N N N N N N N N N N N N N N N N N Y Y Y N N

OR

P

0 0.5285 2.4235 1.3398 0.4479 1.6265 0.5562 0.908 1.4912 0.8077 1.5023 0.5868 2.376 0 1.2009 0.5508 0.5087 0.2513 0.9009 4.992 2.274 10.1679 0 0

0.3172 0 0 0 0 0 0 0.306 0 0.0092 0 0 0 0 0.0263 0.0002 0 0 0.2679 0 0 0 0 0.3172

Enrichment N N Y N N N N N N N N N Y N N N N N N Y Y Y N N

S.-C. Xing et al. / Science of the Total Environment 693 (2019) 133490

repair. Particularly, the corresponding DNA protection genes were highly upregulated in strain Pb1000 under the lead treatment. This result may suggest that the oxidative stress caused by addition of 1000 mg/L lead was too high for bacterial cells, even for the leadadapted strain. The lead upregulated DNA protection genes common in three bacterial strains were listed in Table 2. In the common pathway of base excision repair, the exodeoxyribonuclease gene R11. WTGM002003 was upregulated by 1.2-, 1.76- and 3.5-fold for WT, Pb100 and Pb1000 strains, and the probable A/G-specific adenine glycosylase yfhQ gene R11.WTGM003285 was upregulated by 1.59-, 1.85- and 3.81-fold for WT, Pb100 and Pb1000 strains, respectively, under the lead treatment conditions. The DNA polymerase I gene R11. WTGM002719 was upregulated in both the WT and Pb1000 strains after lead treatment. The mismatch repair pathways was significantly upregulated in the WT and Pb1000 strains, and the common upregulated genes were R11.WTGM002526 (ATP-dependent RecD-like DNA helicase), gene R11.WTGM000101 (recombination protein RECR), R11. WTGM002719 (DNA polymerase I), R11.WTGM000099 (DNA polymerase III subunit gamma/tau), gene R11.WTGM001670 (DNA polymerase III PolC-type), R11.WTGM002452 (uncharacterized protein YQEN), R11. WTGM002667 (endonuclease MUTS2) and R11.WTGM001732 (DNA mismatch repair protein MUTS). The fact was that mismatch repair and base excision repair genes were upregulated in both WT and Pb1000 suggests that mismatch repair and base excision repair might be the major processes used to repair the DNA damage caused by lead stress for B. coagulans R11. As expected, multiple putative pathways related to lead resistance were significantly upregulated when bacteria was cultured in the medium containing lead, based on the KEGG database analysis (Fig. 6). Based on the observed lead stress responses of the WT and leadadapted bacterial strains, it is crucial to further understand the lead resistance and tolerance mechanisms of B. coagulans R11. The transcriptomics results showed that there were nine major pathways involved in lead resistance and tolerance. Three of the nine pathways were related to deoxidation, including glutathione metabolism, cysteine and methionine metabolism, and selenocompound metabolism. Among them, the genes encoding glutathione metabolism were commonly up-regulated in three bacterial strains when treated with lead. Specifically, three genes in the glutathione metabolism pathway were upregulated in the WT and lead-adapted strains. Gene R11.WTGM002030, encoding a 6phosphogluconate dehydrogenase, was upregulated by 1.23-, 1.68and 2.17-fold in the WT, Pb100 and Pb1000 strains, respectively, in the medium containing lead compared with that in the same medium without lead. Gene R11.WTGM002293 that codes a glucose-6phosphate 1-dehydrogenase was upregulated by 1.36-, 1.71- and 1.80-fold in the WT, Pb100 and Pb1000 strains, respectively, in the lead medium. Gene R11.WTGM003592, encoding a glutathione peroxidase homologue BSAA, was upregulated by 1.30-, 1.61- and 2.58-fold in the WT, Pb100 and Pb1000 strains, respectively, in the lead medium. Compared to WT and Pb100 strains, additional three genes in the pathway of glutathione metabolism were specifically upregulated in strain Pb1000 (Fig. 6). Interestingly, the genes for cysteine and methionine metabolism, and selenocompound metabolism were significantly upregulated only in WT and Pb1000 strains under the lead treatment conditions. There were three common genes in the pathway of cysteine and methionine metabolism for the WT and Pb1000 strains, i.e., R11.WTGM001688, R11.WTGM002655 and R11.WTGM002859 (aspartate-semialdehyde dehydrogenase, aspartokinase and S-adenosylmethionine synthase, respectively) and four common genes in the selenocompound metabolism pathway were upregulated in the WT and Pb1000 strains in the lead medium. Gene R11.WTGM003035 (probable cysteine desulfurase) and gene R11.WTGM000525 (methionine synthase) were upregulated 1.21- and 1.20-fold, respectively, in the WT strain in the lead medium. R11.WTGM000402 (cystathionine beta-lyase PATB) and R11.WTGM003703 (5-methyl-tetrahydropteroyltrigluamate-

7

homocysteine methyltransferase) were upregulated 2.85- and 1.22fold respectively, in the WT strain in the lead medium. The XRD results showed that some crystallization of lead compounds similar to PbS crystals formed in the WT and lead-adapted strains in the lead medium; therefore, we identified some genes related to the sulfur metabolism pathway that were significantly upregulated (Fig. 6). Generally, the upregulated genes in the WT and Pb1000 strains were more abundant than those in the Pb100 strain. There were three common upregulated sulfur metabolism genes in the WT and Pb1000 strains, but only one of these genes was upregulated in the Pb100 strain in the lead medium. Gene R11.WTGM000912 (NAD(P)H-dependent FAD/FMN reductase), gene R11.WTGM001969 (sulfite reductase [NADPH] alpha-component) and gene R11.WTGM001970 (sulfite reductase [NADPH] beta-component) were upregulated 1.82-, 1.53- and 1.36fold, respectively, in the WT strain and upregulated 1.53-, 2.29 - and 1.69- fold, respectively, in the Pb1000 strain in the lead medium. The gene R11.WTGM000912 was upregulated 1.65-fold in strain Pb100 by lead treatment. R11.WTGM003428, cystathionine gamma-lyase, which is involved in catalysing the decomposition of sulfur amino acids to hydrothion, was upregulated by 1.82-fold in strain Pb1000 in the presence of lead (Fig. 6). Notably, these upregulated genes in the sulfur metabolism pathway were also linked to the cysteine and methionine metabolism pathway; thus, the metabolism of sulfur amino acids in B. coagulans R11 was not only to secrete deoxidization compounds but may also be associated with modification of lead ions through crystallization. The lead stress also stimulated transport activities, as five genes in the ABC transporter pathway were all regulated in the WT, Pb100 and Pb1000 strains in the lead medium. R11.WTGM002793, sulfate/thiosulfate import ATP-binding protein CYSA, was highly upregulated by approximately 3.76-fold in strain Pb1000, 1.5-fold in WT strain and 1.94fold in strain Pb100. CYSA is involved in the transport of sulfate/thiosulfate from outside to inside the cell and provides sulfur for sulfocompound synthesis (Wickett et al., 2011; Sirko et al., 1995; Stevanato et al., 2018). There were two upregulated metal ion-related transporter genes, which could be the putative lead ion transporters; R11.WTGM002403 (probable metal transport system membrane protein TP_0036) was upregulated by 1.37-, 1.53- and 2.04-fold, and R11. WTGM002404 (zinc uptake system ATP-binding protein ZURA) was upregulated by 1.55-, 1.55- and 1.58-fold in the WT, Pb100 and Pb1000 strains, respectively, in the lead medium. The upregulated genes in the ABC transporter pathway were also much more abundant in the WT and Pb1000 strains than in the Pb100 strain (Fig. 6). There were various transporter genes belonging to the ABC protein family that were upregulated by lead treatment in the WT strain and especially in the Pb1000 strain. For example, R11.WTGM004155 (multidrug resistance ABC transporter ATP-binding/permease protein BMRA) and multidrug resistance protein YKKD were upregulated 1.53- and 2.04-fold, respectively, in strain Pb1000 in the lead medium. Furthermore, teichoic acid related transporter genes (R11.WTGM000807 and R11.WTGM001086) were also upregulated (1.29- and 1.82-fold) in Pb1000 (Fig. 6). Teichoic acid is the main compound of the positive bacteria cell wall. In the previous morphology modification analysis of this study, we found that the bacteria colonies of the Pb1000 strain were significantly different from those of the WT strain. Comparative transcriptomics showed that the genes involved in flagellar formation, motility, signal sensors and even the secretion systems of proteins or some peptides were significantly upregulated by lead treatment. As expected, the upregulated genes were more abundant in the Pb100 and Pb1000 strains in the lead medium by than in the WT strain, and most of the flagellar assembly genes were upregulated in strain Pb1000 (Fig. 6). There were seven genes belonging to the flagellar assembly pathway that were upregulated in strains Pb100 and Pb1000 by lead treatment, and these seven genes were all highly upregulated in strain Pb1000. Specifically, R11.WTGM001649, R11.WTGM001648, R11.WTGM001630, R11. WTGM001631, R11.WTGM001645, R11.WTGM001644 and R11.

8

S.-C. Xing et al. / Science of the Total Environment 693 (2019) 133490

R11.WTGM002030 6−phosphogluconate dehydrogenase

3

R11.WTGM002293 Glucose−6−phosphate 1−dehydrogenase R11.WTGM003592 Glutathione peroxidase homolog BsaA

2

R11.WTGM001312 Putative gamma−glutamyltransferase YwrD R11.WTGM002722 Isocitrate dehydrogenase [NADP]

1

R11.WTGM003003 Probable cytosol aminopeptidase R11.WTGM000912 NAD(P)H−dependent FAD/FMN reductase R11.WTGM001969 Sulfite reductase [NADPH]alpha−component

0

R11.WTGM001970 Sulfite reductase [NADPH]beta−component R11.WTGM000074 Cysteine synthase

−1

R11.WTGM000361 Phosphoadenosine phosphosulfate reductase R11.WTGM003470 Thiosulfate dehydrogenase [quinone] large subunit R11.WTGM000122 Serine acetyltransferase R11.WTGM003186 NAD(P)H−dependent FAD/FMN reductase R11.WTGM003428 Cystathionine gamma−lyase R11.WTGM000976 Homoserine O−succinyltransferase R11.WTGM002655 Aspartokinase R11.WTGM002859 S−adenosylmethionine synthase R11.WTGM001688 Aspartate−semialdehyde dehydrogenase R11.WTGM000524 Bifunctional homocysteine S−methyltransferase/5,10−methylenetetrahydrofolate reductase R11.WTGM000525 Methionine synthase R11.WTGM000857 Homoserine dehydrogenase R11.WTGM001048 Homoserine dehydrogenase R11.WTGM000402 Cystathionine beta−lyase PatB R11.WTGM001588 Probable L−serine dehydratase, beta chain R11.WTGM001589 Probable L−serine dehydratase, alpha chain R11.WTGM001690 Aspartokinase 1 R11.WTGM002505 Adenosylhomocysteine nucleosidase R11.WTGM003429 O−acetylserine dependent cystathionine beta−synthase R11.WTGM003430 S−ribosylhomocysteine lyase R11.WTGM001999 L−lactate dehydrogenase 2 R11.WTGM003431 Uncharacterized methyltransferase R11.WTGM002721 Malate dehydrogenase R11.WTGM002795 Molybdate−binding periplasmic protein R11.WTGM002793 Sulfate/thiosulfate import ATP−binding protein CysA R11.WTGM000335 Probable siderophore−binding lipoprotein YfiY R11.WTGM002403 Probable metal transport system membrane protein TP_0036 R11.WTGM002404 Zinc uptake system ATP−binding protein ZurA R11.WTGM002794 Molybdenum transport system permease protein ModB R11.WTGM000333 Iron(3+)−hydroxamate import system permease protein FhuG R11.WTGM000334 Iron(3+)−hydroxamate import system permease protein FhuB R11.WTGM001228 Oligopeptide transport ATP−binding protein OppD R11.WTGM001229 Oligopeptide transport ATP−binding protein OppF R11.WTGM003271 Putative multidrug export ATP−binding/permease protein R11.WTGM003893 Uncharacterized ABC transporter substrate−binding lipoprotein YvrC R11.WTGM004139 Probable D−methionine transport system permease protein MetI R11.WTGM004140 Methionine import ATP−binding protein MetN R11.WTGM000176 Energy−coupling factor transporter ATP−binding protein EcfA1 R11.WTGM000177 Energy−coupling factor transporter ATP−binding protein EcfA2 R11.WTGM000178 Energy−coupling factor transporter transmembrane protein EcfT R11.WTGM000292 Glycine betaine/carnitine/choline transport ATP−binding protein OpuCA R11.WTGM000293 Glycine betaine/carnitine/choline−binding protein OpuCC R11.WTGM000330 Multidrug resistance protein YkkD R11.WTGM000807 Teichoic acids export ATP−binding protein TagH R11.WTGM001043 Cell division ATP−binding protein FtsE R11.WTGM001044 Cell division protein FtsX R11.WTGM001054 Uncharacterized ABC transporter permease protein YclN R11.WTGM001055 Uncharacterized ABC transporter permease protein YclO R11.WTGM001056 Uncharacterized ABC transporter ATP−binding protein YclP R11.WTGM001057 Uncharacterized ABC transporter solute−binding protein YclQ R11.WTGM001086 Teichoic acid translocation permease protein TagG R11.WTGM001859 Bacitracin export permease protein BceB R11.WTGM002880 Zinc−binding lipoprotein AdcA R11.WTGM003177 L−cystine−binding protein TcyA R11.WTGM003178 L−cystine transport system permease protein TcyB R11.WTGM003179 L−cystine import ATP−binding protein TcyC R11.WTGM003894 Ferric enterobactin transport ATP−binding protein FepC R11.WTGM003895 Hemin transport system permease protein HmuU R11.WTGM001628 Flagellar M−ring protein R11.WTGM001649 Flagellar biosynthesis protein FlhA R11.WTGM001648 Flagellar biosynthesis protein FlhB R11.WTGM001630 Probable flagellar assembly protein FliH R11.WTGM001631 Flagellum−specific ATP synthase R11.WTGM001645 Flagellar biosynthesis protein FlhQ R11.WTGM001644 Flagellar biosynthesis protein FlhP R11.WTGM001632 Flagellar FliJ protein R11.WTGM001625 Flagellar basal body rod protein FlgB R11.WTGM001007 Flagellar hook−associated protein 3 R11.WTGM001012 Flagellin R11.WTGM001635 FlaA locus uncharacterized protein YlxG R11.WTGM001637 Flagellar basal−body rod protein FlgG R11.WTGM001640 Flagellar motor switch protein FliM R11.WTGM001641 Flagellar motor switch phosphatase FliY R11.WTGM001643 Flagellar biosynthetic protein FliZ R11.WTGM003415 Motility protein B R11.WTGM003416 Motility protein A R11.WTGM003995 Flagellar hook−basal body complex protein FlhP R11.WTGM003996 Flagellar hook−basal body complex protein FlhO R11.WTGM001772 Methyl−accepting chemotaxis protein McpA R11.WTGM002817 Hybrid signal transduction histidine kinase K R11.WTGM001653 Chemotaxis protein CheA R11.WTGM002204 Chemotaxis protein methyltransferase R11.WTGM002079 Chemotaxis protein CheY R11.WTGM001656 Chemoreceptor glutamine deamidase R11.WTGM001642 Chemotaxis protein CheY R11.WTGM001652 Chemotaxis response regulator protein−glutamate methylesterase R11.WTGM002696 Methyl−accepting chemotaxis protein TlpB R11.WTGM003389 Methyl−accepting chemotaxis protein 4 R11.WTGM001654 Chemotaxis protein CheW R11.WTGM001655 CheY−P phosphatase CheC R11.WTGM001629 Flagellar motor switch protein FliG R11.WTGM004024 Methyl−accepting chemotaxis protein TlpB R11.WTGM002990 Kinase−associated lipoprotein B R11.WTGM003115 RNA polymerase sigma−54 factor R11.WTGM001060 Alkaline phosphatase synthesis transcriptional regulatory protein PhoP R11.WTGM001061 Alkaline phosphatase synthesis sensor protein PhoR R11.WTGM001373 Sporulation kinase C R11.WTGM001382 Sporulation kinase E R11.WTGM001446 Heme A synthase R11.WTGM001777 Putative carboxypeptidase YodJ R11.WTGM001861 Sensor protein BceS R11.WTGM001862 Sensory transduction protein BceR R11.WTGM002245 Sensor histidine kinase ResE R11.WTGM002246 Transcriptional regulatory protein R11.WTGM002310 Stage 0 sporulation protein A R11.WTGM002573 Sporulation initiation phosphotransferase B R11.WTGM002734 Probable NAD−dependent malic enzyme 4 R11.WTGM004191 Sensor histidine kinase YycG R11.WTGM004192 Transcriptional regulatory protein YycF R11.WTGM003106 Probable protein−export membrane protein SecG R11.WTGM004223 Membrane protein insertase YidC 2 R11.WTGM001035 Protein translocase subunit SecA R11.WTGM002552 Protein translocase subunit SecDF R11.WTGM000129 Protein translocase subunit SecE R11.WTGM000167 Protein translocase subunit SecY R11.WTGM002558 UPF0092 membrane protein YrbF R11.WTGM001808 Membrane protein insertase YidC 1 R11.WTGM001599 Signal recognition particle receptor FtsY R11.WTGM001602 Signal recognition particle protein R11.WTGM001855 Thioredoxin reducase R11.WTGM000022 Methionine−tRNA ligase R11.WTGM001098 Thioredoxin reducase R11.WTGM002037 Selenide, water dikinase R11.WTGM003703 5−methyltetrahydropteroyltrigluamate−homocysteine methyltransferase R11.WTGM003035 Probable cysteine desulfurase

−2

Pathway ABC transporters Bacterial chemotaxis Bacterial secretion system Cysteine and methionine metabolism Flagellar assembly Glutathione metabolism pathway Selenocompound metabolism Sulfur metabolism pathway Two−component system

Pb1000−1000

Pb1000−C

Pb100−100

Pb100−C

WT−100

WT−C

S.-C. Xing et al. / Science of the Total Environment 693 (2019) 133490

9

Table 2 Pathway of lead stress response of acclimatization and WT strains. Pathway Nucleotide excision repair

Mismatch repair

Base excision repair

Homologous recombination

Gene tag

Annotation

Log2 fold changes

uvrC uvrC yjcD uvrB polA mutS2 yqeN yjcD mutS polC dnaX exoA yfhQ polA yqeN polC polA yrrC recR dnaX

UvrABC system protein C UvrABC system protein C Putative ATP-dependent DNA helicase YjcD UvrABC system protein B DNA polymerase I Endonuclease MutS2 Uncharacterized protein YqeN Putative ATP-dependent DNA helicase YjcD DNA mismatch repair protein MutS DNA polymerase III PolC-type DNA polymerase III subunit gamma/tau Exodeoxyribonuclease Probable A/G-specific adenine glycosylase YfhQ DNA polymerase I Uncharacterized protein YqeN DNA polymerase III PolC-type DNA polymerase I ATP-dependent RecD-like DNA helicase Recombination protein RecR DNA polymerase III subunit gamma/tau

0.67 0.37 0.47 0.37 0.25 0.28 0.68 0.47 0.25 0.29 0.43 0.75 0.34 0.25 0.68 0.29 0.25 0.47 0.36 0.43

WTGM001632 were upregulated 4.24-, 3.87-, 2.69-, 4.03-, 6.60-, 3.78and 5.04-fold, respectively, in strain Pb1000 in the lead medium. The fold changes of these seven genes were lower in Pb100 than in Pb1000 in the lead medium. In contrast, only one gene, R11. WTGM001628 (flagellar M-ring protein), was upregulated in the WT strain in the lead medium. The two-component system pathway and bacterial chemotaxis pathway are related to signal acceptance and responses in bacteria (Falke et al., 1997; Bi and Sourjik, 2018; Kuroda et al., 2003). Among these three strains, the upregulated genes in the two pathways were the most abundant in Pb1000 in the lead medium, and only two genes (R11.WTGM001772, R11.WTGM002990) were upregulated in the WT, Pb100, and Pb1000 strains in these two pathways. In addition, six chemotaxis pathway genes, i.e., cheY, cheB, cheA, cheW, cheC and cheD (R11.WTGM001642, R11.WTGM001652, R11.WTGM001653, R11. WTGM001654, R11.WTGM001655 and R11.WTGM001656), were highly upregulated by 3.74-, 3.57-, 3.72-, 3.65-, 2.92- and 4.52-fold in leadtreated strainPb1000. Multiple methyl-accepting chemotaxis genes were also upregulated in Pb1000 (Fig. 6), but only one methylaccepting chemotaxis protein gene mcpA (R11.WTGM001772), was upregulated in all three bacterial strains in the lead medium. Methylaccepting chemotaxis protein and chemotaxis protein are two crucial protein families for flagellar motility, which helps bacteria sense extracellular signals and promotes the motility of bacteria in the environment in order to avoid or approach these signal compounds (Williams and Stewart, 2002; Zatakia et al., 2017; Batra et al., 2016; Ehinger, 1977; Wu et al., 2004). The bacterial secretion system pathway was not significantly upregulated in the Pb100 strain under the lead treatment, but multiple genes were upregulated in the Pb1000 strain under the lead treatment, and three genes were upregulated in the WT strain under the lead treatment. Surprisingly, no common genes were upregulated in both WT and strain Pb1000, but the results showed that the major secretion pathways for B. coagulans R11 were Sec (Sec-SPR cooperation pathway) and Tat (two-arginine translocation pathway). In addition, after the lead treatment, the upregulated genes in the WT and Pb1000 strains all belonged to the Sec-SPR pathway. These results suggest that, in response to lead stress, production of the Sec signal-peptide containing

Gene ID R11.WTGM002657 R11.WTGM000305 R11.WTGM003509 R11.WTGM001080 R11.WTGM002719 R11.WTGM002667 R11.WTGM002452 R11.WTGM003509 R11.WTGM001732 R11.WTGM001670 R11.WTGM000099 R11.WTGM002003 R11.WTGM003285 R11.WTGM002719 R11.WTGM002452 R11.WTGM001670 R11.WTGM002719 R11.WTGM002526 R11.WTGM000101 R11.WTGM000099

secretion protein was increased, and these functional proteins might be secreted to extracellular environment to enhance lead tolerance. 2.8. Lead stress stimulated downregulated genes in WT and lead-adapted strains Notably, multiple genes in the flagellar assembly and bacterial chemotaxis pathways were downregulated in the WT strain under the lead treatment. In particular, the genes R11.WTGM001631, R11. WTGM001644, R11.WTGM001648, R11.WTGM001649 and R11. WTGM001656 (Fig. 6) that were highly upregulated in strain Pb1000, were downregulated in the WT strain under the lead treatment. This finding suggests that the lead stress might cause a mutation (s) resulting in a contrast gene expression pattern in response to lead stress for better adaptation to the stressed environment. Multiple downregulated genes were also found in the ABC transporter pathway in the three bacterial strains under the lead treatment. Among them, five genes were downregulated in all the bacterial strains (R11. WTGM003967, R11.WTGM003969, R11.WTGM001067, R11. WTGM001116 and R11.WTGM001068), but these genes were more highly downregulated (approximately 2.8- times) in the Pb1000 strain than in the WT and Pb100 strains in the lead medium. These five downregulated genes were yqgH, yqgI, msmX, mdxE and mdxF. Among them yqgH and yqgI were probable ABC transport permeases, which are involved in substance import through cell membrane. msmX, mdxE and mdxF are the genes related to maltodextrin transport, belonging to sugar transporter family. These five genes were all downregulated in the strains treated with lead, suggesting that lead stress could inhibit the adsorption of some substances from the extracellular environment, such as maltodextrin. Multiple ABC transporter genes were more highly downregulated in strain Pb1000 than in the WT and Pb100 strains under the lead treatment. Most of the downregulated genes were involved in sugar transport, phosphate transport and sulfur amino acid transport systems. In the two-component system, only R11.WTGM002141, glutamine synthetase, was downregulated in all three bacterial strains in the lead medium, and this gene was more highly downregulated in strain Pb1000 (approximately 2.49-fold) than in the other two strains. It is worth

Fig. 6. Transcriptomic and genomic information of genes involved in lead resistance pathways. Gene readcount was normalized using variance stabilizing transformation in DESeq2 and software R to fit in the range of −2 to 3 across all genes. Warm colors indicate higher normalized counts (as shown in Color Key), and the left color ranges indicate the different lead resistance pathways (annotation shown in Pathway). WT-C: wild type cultured without lead, WT-100: wild type cultured with 100 mg/L lead, Pb100-C: Pb100 cultured without lead, Pb100-100: Pb100 cultured with 100 mg/L lead, Pb1000-C: Pb100 cultured without lead, Pb1000-1000: Pb100 cultured with 1000 mg/L lead. Each square on the heat map represents a biological replicate.

10

S.-C. Xing et al. / Science of the Total Environment 693 (2019) 133490

noting that glutamine synthetase is the enzyme for biosynthesis of glutamic acid, which is one of the amino acids used to synthetize glutathione (Ehinger, 1977; Wu et al., 2004); hence, this result could suggest that glutathione was highly synthetized under lead stress, and the quantity demand for the glutamic acid, “raw material”, was increased. To summarize the comparative transcriptomics results, DNA protection functional genes were upregulated in the WT and lead-adapted strains with lead stress, and these DNA protection pathways were much abundant and highly upregulated in the WT strain and the strain treated by the high lead concentration, Pb1000, this result suggests that the intracellular deposition of lead caused oxidative damage to nucleic acids, which stimulated the nucleic acid repair response. Eight lead resistance and tolerance pathways were significantly upregulated with lead stress, glutathione and sulfur amino acid synthesis was the major deoxidation pathway, and the form modification of lead ions by crystallization could be completed via the sulfur metabolism pathway. Lead stress also stimulated the mutation of genes involved in the formation of the motility system important to avoid lead in the environment, and chemotaxis proteins and methyl-accepting chemotaxis protein were upregulated to federatively enhance the flagellar motility. All involved upregulated genes and their pathways are shown in Fig. 6. The color of the heat map indicates the expression level of each pair of genes in the control (incubation with pure culture medium) and lead treatment (incubation with lead-containing medium) groups. Clearly, the expression levels of most of these lead resistance and tolerance genes were significantly increased or decreased in the Pb1000 strain compared to those in the WT and Pb100 strains, indicating that these genes and pathways were related to the response to lead stress, as Pb1000 was acclimatized to the medium with the high lead concentration. We summarize the complete lead resistance and tolerance mechanisms of the WT, Pb100 and Pb1000 strains observed in the present study in Fig. 7. The genes involved in these regulatory functions are labelled in Fig. 6. 3. Discussion In this work, we analysed the lead stress responses of WT and leadadapted bacterial strains of B. coagulans R11 by comparative genome sequencing, genome resequencing, and transcriptomics. Acclimatization of laboratory bacterial strains to various compounds has been used for response assessments in numerous studies (Yoneda et al., 2016; Dragosits and Mattanovich, 2013). In addition, we summarize the complete lead stress responses of the WT, Pb100 and Pb1000 strains observed in the present study in Fig. 7. 3.1. Lead adaptive strains and wild type It is well known that high concentration heavy metal could inhibit the growth of bacteria, even the heavy metal tolerance bacteria (Yoneda et al., 2016), but in this study, after 85 passages in the lead stress incubation, the wild type and lead acclimatization bacteria were tested the growth curve in pure MRS culture medium. Thus we hypothesized that sacrifices to the proliferation ratio could increase the energy and nutrition supplementation for the synthesis of deoxidation compounds and provide substrates for the synthesis of lead resistance and tolerance proteins in B. coagulans R11. The transcriptomics analysis supported this hypothesis, as many genes related to energy metabolism in the lead resistance and tolerance pathway were increased (Fig. 6). However, the mechanisms between the proliferation ratio and lead tolerance in lead-adapted bacteria need further study. Fig. 7. Lead resistance and tolerance mechanisms of Bacillus coagulans R11 and its acclimatization bacteria. Fig. 7A was the response pathway of lead stimulation for WT, 7B was the response pathway of lead stimulation for 100Pb adaption bacteria strain, 7C was the response pathway of lead stimulation for 1000Pb adaption bacteria strain.

S.-C. Xing et al. / Science of the Total Environment 693 (2019) 133490

3.2. SNP and Indel sites in lead adaptive strains To summarize these mutation sites in 100Pb and 1000Pb, we found no transcriptional regulator, sensor, lead resistance or lead tolerance mutation sites that were upregulated or downregulated in 100Pb, but R11.WTGM001655, which contained a SNP site, and R11.WTGM004046 and R11.WTGM003198, which contained Indel sites, were upregulated in 1000Pb under the lead treatment. In addition, R11.WTGM003068, which contained a SNP site, and R11.WTGM004137, which contained an Indel site, were downregulated in 1000Pb under the lead treatment. These results indicate that these mutations were related to the lead stress response, especially under exposure to the high lead concentration. In addition, these mutations occurred in transcriptional regulators and sensors, suggesting that they enabled the bacteria to overcome survival stress and led to the selection of these mutations, but these mutations do not largely occur in transcriptional regulators, promoter regions and sensors. 3.3. Gene response for the lead stimulation The lead accumulation capacity is a crucial parameter in the assessment of the potential utilization value of sorbents. In our study, the total lead removal and extracellular accumulation capacities of the WT and lead-adapted bacterial strains were did not exhibit much diversity in the different initial lead concentration groups. The extracellular accumulation capacity of the WT strain was the highest when the initial lead concentration was 20 mg/L, but the extracellular accumulation capacity of the Pb1000 strain exceeded that of the WT strain when the initial lead concentration was 100 mg/L. In the transcriptomics results, we observed that teichoic acid-related transporter genes (R11. WTGM000807 and R11.WTGM001086) were upregulated in Pb1000; previous studies have elucidated that the function of teichoic acid includes binding cations and reinforcing the cell wall (Pal et al., 1990; Beveridge et al., 1982; Thomas and Rice, 2014). Therefore, we concluded that teichoic acid was highly synthesized and secreted for the extracellular binding of lead ions and the reinforcement of the cell wall in order to decrease and prevent increases in the intracellular concentration of lead ions in Pb1000. The capacities of the lead-adapted bacterial strains to accumulate intracellular lead were both obviously higher than that of the WT strain. The intracellular lead (heavy metal) accumulation capacity was increased for the lead-adapted bacterial strains due to their lead (heavy metal) tolerance ability (Brunetti et al., 2012; Ryan et al., 2009; Romaidi and Ueki, 2016), as these bacterial cells can tolerate a larger number of lead ions and have the ability to prevent or cure toxin damage. Considering the XRD results, lead ions could be crystallized into sulfur-containing crystals. In addition, hydrothion formation crucial regulation (Huang et al., 2010; Mard et al., 2015; Sheibani et al., 2017; Modis et al., 2017), cystathionine gamma-lyase, cystathionine betalyase PATB and O-acetylserine dependent cystathionine beta-synthase genes (R11.WTGM003428, R11.WTGM000402, and R11.WTGM003429) were significantly upregulated in Pb1000 to better adapt to the high lead concentration by the reaction of hydrothion and lead ions (Fig. 6). In our study, we also found that the sulfate/thiosulfate import ATP-binding protein CYSA (R11.WTGM002793) was upregulated in the WT and lead-adapted strains (Fig. 6); this gene regulates extracellular sulfur absorption, indicating that more sulfur or sulfur compounds are imported into the cells by specific transporter proteins to better deposit lead by sulfur or its compounds. Thus, we assume that B. coagulans R11, especially the Pb1000 strain, secreted some substance that produces the hydrothion, which reacts with lead ions in order to deposit the lead, but the deposition is similar to that of galena, so the complete reaction process was very similar to the metabolic process of sulfate-reducing bacteria (Pruden et al., 2007; Kiran et al., 2017). In our results, we observed that putative lead ion transporters, R11.WTGM002403 (probable metal transport system membrane protein TP_0036) and R11.WTGM002404 (zinc uptake system ATP-binding protein ZURA), were upregulated in

11

the WT and the lead-adapted strains. As the previous study reported, these two systems could regulate the uptake of ions, and the transporter ZURA could inhibit the uptake of zinc when the intracellular zinc concentration is high (Patzer and Hantke, 1998). In addition, many studies have indicated that divalent metal ions could replace zinc to bind proteins or enzymes (Charlet et al., 2012; Neupane and Pecoraro, 2011; Morales et al., 2011), so we assumed that part of the inhibition of the absorption of lead ions could be achieved by ZURA and that lead could be exported by the probable metal transport system membrane protein. The lead intracellular capacities of WT and Pb1000 both stopped increasing (data not shown) when the initial lead concentration increased from 80 to 100 mg/L; this result might support the aforementioned assumption. However, for Pb100, the intracellular accumulation capacity remained constant when the initial lead concentration was 60 mg/L and 80 mg/L, but it increased when the initial lead concentration was 100 mg/L. Our present results could not explain this difference. Heavy metals induce oxidative damage in cells (Rehman et al., 2017; Javed et al., 2016; Xia et al., 2016), so decreasing oxidative damage in cells is important for normal physiological activities. Not only the redox system but also the substrate import genes were upregulated in the two lead-adapted strains. As above mentioned, the sulfate/thiosulfate import ATP-binding protein CYSA (R11.WTGM002793) was upregulated in all three strains. We found that the reducing ability was higher in the lead-adapted bacterial strains than in the WT strain. However, as the intracellular glutathione yield was not high for strain Pb100 under the lead treatment, and the extracellular yield was higher with no lead addition, we assumed that glutathione is not the unique reductant enabling strain Pb100 to adapt to lead stress and that glutathione could be secreted from a cell to the external environment. In addition, the extracellular glutathione could be utilized by strain Pb100 under the lead treatment, thus decreasing the extracellular glutathione yield. Furthermore, based on the transcriptomics results, genes R11.WTGM002030 (6-phosphogluconate dehydrogenase), R11.WTGM002293 (glucose-6phosphate 1-dehydrogenase) and R11.WTGM003592 (glutathione peroxidase homologue BSAA) were upregulated in the WT and leadadapted strains. Previous studies have mentioned that 6phosphogluconate dehydrogenase and glucose-6-phosphate 1dehydrogenase could catalyse glucose and glucose-6-phosphate to produce NADPH, and NADPH could provide the “H” for the reduction of glutathione (Pan et al., 2018). In addition, the glutathione peroxidase homologue BSAA could also catalyse the oxidation of glutathione, which cleans the oxide molecules in cells. These three upregulated genes enhance the metabolism of glutathione in the cell in response to lead stimulation. However, the degree of gene richness in the reduction pathway was not as high in the Pb100 strain as that in the WT and Pb1000 strains (Fig. 6), so it could be concluded that not much oxidative damage occurred in Pb100. The upregulated genes that belong to the cysteine and methionine metabolism pathway were only significantly upregulated in the WT and Pb1000 strains; R11.WTGM001688, R11. WTGM002655 and R11.WTGM002859 (aspartate-semialdehyde dehydrogenase, aspartokinase and S-adenosylmethionine synthase, respectively) were the three upregulated genes, and all are related the composition and metabolism of glutathione (Xu et al., 2014; Jeon et al., 2018; Lee et al., 2006). We did not find upregulated genes in nuclide acid repair, nucleotide excision repair, DNA mismatch repair or homologous recombination pathways (Table 2 and Fig. 6) in the Pb100 strain. These results may indicate that the oxidative damage was not as serious in strain Pb100 as that in WT and Pb1000 strains, and therefore it was unnecessary for strain Pb100 to increase the expression level of the redox system genes. 3.4. Morphological structure response for the lead acclimatization After lead acclimatization, the morphology of lead-adapted bacterial strains were changed significantly, especially strain Pb1000 (Fig. 5C). In addition, the transcriptomics results indicate that flagella were formed

12

S.-C. Xing et al. / Science of the Total Environment 693 (2019) 133490

in response to the lead stress. Several previous studies have mentioned that the protein that may belong to flagellar regulation was impacted by treatments with heavy metals, especially in pathogens, but this phenomenon has not been clearly clarified or completely studied (Luo et al., 2016). In our study, genes and pathways involved in flagellar formation and motility were obviously upregulated in the lead-adapted bacterial strains, but most of the genes were downregulated in the WT strain. It is well known that flagella are the motion structures for bacteria that help bacteria approach or avoid compounds in the environment (Silverman and Simon, 1974; Fujii et al., 2017; Wang et al., 2017). The flagellar formation in the lead-adapted bacteria suggests that the leadadapted bacterial strains could move in the culture medium to avoid high lead concentration areas or approach nutrient-enriched areas to uptake essential compounds. Moreover, in cooperation with the function of the flagella, the chemotaxis and two-component system pathways were also significantly upregulated in the lead-treated bacterial strains (Fig. 6), especially in Pb1000. These two pathways could enable bacteria to sense modifications in or the concentration of compounds of the extracellular environment and associated bacterial responses (Falke et al., 1997; Bi and Sourjik, 2018; Kuroda et al., 2003). Hence, the two pathways assist the motility of bacteria. However, we also observed that many genes involving flagellar formation and bacterial chemotaxis were upregulated in the Pb1000 strain but downregulated in the WT strain. We assume that the regulation of flagellar formation and bacterial chemotaxis genes did not occur in the WT strain after exposure to lead stress because these genes need time to be upregulated or to mutate, but survival under lead stress for long periods could promote the above described mutations and expression of survival-related genes to support resistance and tolerance to lead stress, as supported by the results found in our study. To summarize the omics results of this study, after acclimation to lead stress, mutations were observed in the two lead-adapted strains, and higher lead concentrations could induce mutations in genes related to multiple physiological functions in the cell. In our study, genes with SNPs and Indels were identified, and some environmental stress resistance transcriptional regulators or sensors were upregulated in Pb1000, which was acclimated to a high concentration of lead (Tables 1A, 1B, and 1C); this result indicates that increasing lead stress could stimulate mutations in B. coagulans R11, but these mutations require sufficient time. Because of the toxicology characteristics of heavy metals, oxidative damage could induce nucleic acid damage (Zheng et al., 2014; Evans and Cooke, 2004; Guentchev et al., 2002). The oxidative damage was serious in the WT and Pb1000 strains but not in the Pb100 strain, as evidenced by the number of damage-related genes that were upregulated in the WT and Pb1000 strains. This result indicates that the lead stress had a large effect on the WT strain because the WT resistance and tolerance genes could not be expressed immediately, and although the lead stress resistance and tolerance genes and pathways were expressed in Pb1000, the lead concentration in the medium of this strain was too high for the expressed genes to completely decrease the oxidative damage. Hence, the expression of lead resistance and tolerance genes required a long acclimation time, but the extent of the resistance and tolerance abilities was also limited. The major lead tolerance mechanism for WT was via methionine, cysteine and glutathione synthesis to reduce oxides, since these pathways were upregulated. In addition, based on the application value of the methionine, cysteine and glutathione, the involving genes in above substance synthesis could be the candidate for the further isolation or cultivation of Bacillus coagulans which acts as such producer. Our study shows that the lead resistance and tolerance mechanisms of Pb100 were simple; the major pathways were flagellar formation, the export of lead ions, and the inhibition of lead import. Based on the oxidative damage results, there was no obvious DNA oxidative damage in Pb100; therefore, we conclude that the flagellar motility system and ion transporters application could be the first choice in the lead stress resistance and tolerance mechanism of B. coagulans R11. In addition, the expression of the two

lead resistance and tolerance systems were sufficient for B. coagulans R11 to greatly avoid oxidative damage under low lead concentrations. In addition to the two abovementioned major lead resistance and tolerance mechanisms, the redox system was significantly expressed in Pb1000 because the bacteria could not completely manage the high level of lead stress with the flagellar system and ion transporters; thus, oxidative damage occurred. The most downregulated genes in the bacterial strains had no direct effect on lead resistance and tolerance, but we assume that the downregulation of genes such as R11. WTGM002141 (glutamine synthetase) could increase the synthesis of crucial compounds. It was difficult to determine the roles of the downregulated flagellar and bacterial chemotaxis genes in the WT strain with our results. We hypothesize that the downregulation of these genes could indirectly increase the synthesis of some compounds or save energy for other metabolism pathways. These results need to be elucidated in further studies. In conclusion, we demonstrated the lead resistance and tolerance mechanisms of B. coagulans R11 by comparative omics in two leadadapted strains and a WT strain. Many mutations were observed in both of the lead-adapted strains. The major lead resistance and tolerance mechanisms of these strains were accomplished via increased mobility by the flagellar system and the export of lead ions by the transporter system in the cell wall, suggesting the discovery of a novel resistance mechanism of probiotics in this study. However, the expression of these two systems requires a long acclimation period. The expression level of the redox system depended on the level of oxidative damage. Furthermore, the downregulated genes observed in the bacteria could contribute to the synthesis of compounds in the lead resistance and tolerance pathways. In addition, the increase in the lead resistance and tolerance abilities could also enhance the intracellular accumulation of lead. 4. Materials and methods 4.1. Lead resistance acclimatization Bacillus coagulans R11 was isolated from a lead mine in Yunnan Province, China, by our laboratory. The lead minimum inhibition concentration (MIC) of this bacterial strain was 2500 mg/L. To obtain strains adapted to different lead concentrations, the wild-type (WT) strain of B. coagulans R11 was inoculated into De Man Rogosa, Sharpe (MRS) broth (HuanKai Microbiology, Guangzhou, China) supplemented with 100 mg/L and 1000 mg/L lead ions, respectively, and cultured at 37 °C for 24 h in anaerobic jars containing Anaero Pack (Mitsubishi Gas Chemical Company, Inc. Tokyo, Japan) for each subculture. Once the culture optical density at 600 nm (OD600) of each acclimatization group increased and remained constant (approximately 85 passages), the bacterial strain was collected and frozen with 40% glycerol in −80 °C for later study. The pure adaptive bacterial strain prepared for further experiments in this study was picked from the frozen stock by culturing in MRS agar that contained 100 mg/L or 1000 mg/L of lead. The growth curves of the WT and lead-acclimated bacterial strains were tested in accordance with the methods of previous studies (Yoneda et al., 2016). Briefly, the WT, 100 mg/L acclimated (Pb100) and the 1000 mg/L acclimated (Pb1000) strains were inoculated at a ratio of 2% in MRS broth with the culture conditions described above and cultured for 48 h. The culture OD600 of each group was tested every 2 h, except for the period from eighth hour to sixteenth hour. 4.2. Lead accumulation capacity of growing bacterial strains To determine the lead accumulation capacities of the WT and acclimated strains, bacterial strains were inoculated in MRS broth containing different lead concentrations, i.e., 20 mg/L, 40 mg/L, 60 mg/L 80 mg/L and 100 mg/L. Based on the results of the growth curves, the incubation time was 24 h, and the culture conditions were those mentioned above.

S.-C. Xing et al. / Science of the Total Environment 693 (2019) 133490

After incubation for 24 h, bacterial pellets and culture supernatants were collected separately by centrifugation. Bacterial pellets were used to assess extracellular binding capacity and intracellular accumulation capacity according to the test methods described by Huang (Huang et al., 2013). The culture supernatants were used for total lead accumulation capacity test. The lead concentration was determined by inductively coupled plasma mass spectrometry (7700 series ICP-MS, Agilent Technologies, Australia). The Pb2+ accumulation capacity of B. coagulans R11 and derivatives was calculated using the following equation: Q¼

C1−C2 N

where Q is the Pb2+ binding capacity of the bacteria (mg/g), C1 and C2 are the initial and final Pb2+ concentrations, respectively (mg/L), and N is the concentration of the bacterial pellets (g/L). 4.3. Detoxification capacity test The bacterial reducing ability was performed following the method described by Lin (Lin and Yen, 1999). In order to calculate the reducing ability of the test bacteria strains, cysteine reducing ability was as the representation, briefly, the reducing ability standard curve of cysteine was prepared by the method which was also referenced the method same as bacteria strains testing. For quantification analysis of glutathione production, the bacterial strains were incubated in pure MRS broth and lead-containing MRS broth under the same culture conditions described above, except the WT and Pb100 strains were incubated in MRS broth containing 100 mg/L lead, and Pb1000 was cultured with the lead concentration at 1000 mg/L. The bacterial pellets were collected by centrifugation for analysis of intracellular glutathione level, and the supernatant was collected for quantification of glutathione in extracellular portion. The glutathione content was assessed using the Reduced Glutathione (GSH) Assay Kit (Solarbio, Beijing, China) following the manufacturer's protocol. 4.4. X-ray polycrystalline diffractometer assessment WT and acclimated bacterial strains were cultured in pure MRS broth and lead-containing MRS broth with the abovementioned experiment procedure to characterize the formation of lead sulfide (PbS). The bacterial pellets were harvested and washed with deionized water to remove any remaining broth. Before X-ray polycrystalline diffractometer analysis (XRD, Bruker D8 ADVANCE Germany), The pellets were dried at 37 °C for 3 days, and then crushed and tested using a X-ray polycrystalline diffractometer (XRD, Bruker D8 ADVANCE Germany). 4.5. DNA and RNA extraction Bacterial genomic DNA samples were extracted for whole genome sequencing and genome resequencing. The WT, Pb100 and Pb1000 strains were harvested from the MRS broth after culturing for 24 h. Genomic DNA samples were prepared using the Quick-DNA Fungal/Bacterial Miniprep Kit (Zymo) following the manufacturer's procedure. For RNA purification, three bacterial strains were inoculated in pure MRS broth and lead-containing MRS broth, respectively. Briefly, for the control group, the three bacterial strains were cultured in pure MRS broth. For the treatment group, WT and Pb100 were cultured in the MRS broth that contained 100 mg/L lead, and the Pb1000 strain was cultured in the 1000 mg/L lead MRS broth. After 24 h incubation, RNA samples from these three strains were extracted using the ZR Fungal/Bacterial RNA MiniPrep kit (Zymo) following the manufacturer's instructions. The isolated DNA and RNA were stored at −80 °C and then sent to a commercial company (Novogene Co., Ltd) for omics testing. The nucleic acid sequencing platform was Illumina Hiseq 4000 for transcriptomics,

13

Illumina Hiseq PE150 for resequencing. The nucleic acid pretreatment was conferenced the methods which described by Yoneda before sequencing by Illumina Hiseq 4000 and Illumina Hiseq PE150 (Yoneda et al., 2016). 4.6. Illumina library preparation Before sequencing, total RNA quality was assessed by (1) agarose gel electrophoresis to assess RNA degradation and potential contamination; (2) Nanodrop was used to assess the purity of RNA (OD 260/280 ratio); (3) Qubit was used to accurately quantify the RNA concentration; (4) Agilent 2100 was used to accurately assess the integrity of RNA. After the sample was tested, rRNA was removed using the ribo-zero kit, which enriched the mRNA. The mRNA molecules were broken into short pieces by adding fragmentation buffer, and used as the template to synthesize one chain cDNA with six bases random primers (random hexamers). Double strand cDNA molecules were then synthesized by adding corresponding buffer solution, dNTPs (dTTP was replaced to dUTP), DNA polymerase I and RNase H. AMPure XP beads (Beckman Coulter Inc., USA) were used in purification of double-stranded cDNA, using the USER enzyme to degrade the second chain of cDNA which contained U. The purified double-stranded cDNA was telomere maintenance at first, after dA –tailing, and then the fragment size selection used AMPure XP beads. Finally, to obtain the final library, PCR amplification was performed and PCR products purification was carried by AMPure XP beads. When library building was completed, Qubit 2.0 was used for preliminary quantification. The library was diluted to 1 ng/μL, then Agilent 2100 was used to assess the insert fragment length of library (insert size), and Q-PCR was employed to accurately quantity the effective concentration of library (library effective concentration N 2 nM), in order to ensure the quality of library (Yoneda et al., 2016). 4.7. Genome sequencing and mutation analysis Raw Illumina data was assessed by the probability model to get the value of Phred, and the data was mapped to the genome of wild type B. coagulans R11 (Genbank accession number CP026649) by using Bowtie 2. The mapped files were converted to sorted BAM files with BAM index files using SAM tools. For the mutation analysis, BWA and SAM tools were used to map the conference sequence (B. coagulans R11), the SNPs and Indels were assessed and counted by SAM tools, in addition, the structural variation was assessed by BreakDancer (Yoneda et al., 2016). 4.8. Genome sequencing data analysis The gene expression level of each sample was analysed by HTSeq software, and the model used was union, the number of genes in different expression levels and the expression level of individual genes were statistically analysed. In general, the value of FPKM is 0.1 or 1 as the threshold for determining whether genes are expressed. The software DESeq was used for normalization of the read counts from analysis of genes expression levels. Supplementary data to this article can be found online at https://doi. org/10.1016/j.scitotenv.2019.07.296. Consent for publication All authors agree to publication. Funding This work was supported by the Earmarked Fund for Modern Agroindustry Technology Research System (CARS-41), Special Fund for Agro-scientific Research in the Public Interest (201303091) and

14

S.-C. Xing et al. / Science of the Total Environment 693 (2019) 133490

Graduate Student Overseas Study Program form South China Agricultural University (2018LHPY001).

Declaration of Competing Interest None declared. Acknowledgements We thank professor Xi-quan Zhang for discussion and scientific guidance in transcriptomics results, Test Center of South China Agricultural University for sharing equipment important for carrying out experiments and Guangzhou Sagene Biotechnology Company for the painting of mechanisms elucidation figure. References Ahemad, M., Kibret, M., 2013. Recent trends in microbial biosorption of heavy metals: a review. Biochem. Mol. Biol. 1, 19–26. Batra, M., Sharma, R., Malik, A., Dhindwal, S., Kumar, P., Tomar, S., 2016. Crystal structure of pentapeptide-independent chemotaxis receptor methyltransferase (CheR) reveals idiosyncratic structural determinants for receptor recognition. J. Struct. Biol. 196, 364–374. Beveridge, T.J., Forsberg, C.W., Doyle, R.J., 1982. Major sites of metal binding in Bacillus licheniformis walls. J. Bacteriol. 150, 1438–1448. Bi, S., Sourjik, V., 2018. Stimulus sensing and signal processing in bacterial chemotaxis. Curr. Opin. Microbiol. 45, 22–29. Brunetti, G., Farrag, K., Soler-Rovira, P., Ferrara, M., Nigro, F., Senesi, N., 2012. The effect of compost and Bacillus licheniformis on the phytoextraction of Cr, Cu, Pb and Zn by three brassicaceae species from contaminated soils in the Apulia region, southern Italy. Geoderma 170, 322–330. Charlet, L., Chapron, Y., Faller, P., Kirsch, R., Stone, A.T., Baveye, P.C., 2012. Neurodegenerative diseases and exposure to the environmental metals Mn, Pb, and Hg. Coord. Chem. Rev. 256, 2147–2163. Dragosits, M., Mattanovich, D., 2013. Adaptive laboratory evolution – principles and applications for biotechnology. Microb. Cell Factories 12, 64. Ehinger, B., 1977. Glial and neuronal uptake of GABA, glutamic acid, glutamine and glutathione in the rabbit retina. Exp. Eye Res. 25, 221–234. El-Sayed, M.T., 2013. Removal of lead(II) by Saccharomyces cerevisiae AUMC 3875. Ann. Microbiol. 63, 1459–1470. Evans, M.D., Cooke, M.S., 2004. Factors contributing to the outcome of oxidative damage to nucleic acids. BioEssays 26, 533–542. Falke, J.J., Bass, R.B., Butler, S.L., Chervitz, S.A., Danielson, M.A., 1997. The two-component signaling pathway of bacterial chemotaxis: a molecular view of signal transduction by receptors, kinases, and adaptation enzymes. Annu. Rev. Cell Dev. Biol. 13, 457–512. Florea, R.M., Stoica, A.I., Baiulescu, G.E., Capotă, P., 2005. Water pollution in gold mining industry: a case study in Roşia Montană district, Romania. Environ. Geol. 48, 1132–1136. Fu, F., Wang, Q., 2011. Removal of heavy metal ions from wastewaters: a review. J. Environ. Manag. 92, 407–418. Fujii, T., Kato, T., Hiraoka, K.D., Miyata, T., Minamino, T., Chevance, F.F.V., Hughes, K.T., Namba, K., 2017. Identical folds used for distinct mechanical functions of the bacterial flagellar rod and hook. Nat. Commun. 8, 14276. Giller, K.E., Witter, E., McGrath, S.P., 2009. Heavy metals and soil microbes. Soil Biol. Biochem. 41, 2031–2037. Guentchev, M., Siedlak, S.L., Jarius, C., Tagliavini, F., Castellani, R.J., Perry, G., Smith, M.A., Budka, H., 2002. Oxidative damage to nucleic acids in human prion disease. Neurobiol. Dis. 9, 275–281. Hassen, A., Saidi, N., Cherif, M., Boudabous, A., 1998. Resistance of environmental bacteria to heavy metals. Bioresour. Technol. 64, 7–15. Hezbullah, M., Sultana, S., Chakraborty, S.R., Patwary, M.I., 2016. Heavy metal contamination of food in a developing country like Bangladesh: an emerging threat to food safety. J. Toxicol. Environ. Health Sci. 8, 1–5. Huang, S., Chua, J.H., Yew, W.S., Sivaraman, J., Moore, P.K., Tan, C.-H., Deng, L.-W., 2010. Site-directed mutagenesis on human cystathionine-γ-lyase reveals insights into the modulation of H2S production. J. Mol. Biol. 396, 708–718. Huang, F., Dang, Z., Guo, C.L., Lu, G.N., Gu, R.R., Liu, H.J., et al., 2013. Biosorption of Cd(ii) by live and dead cells of Bacillus cereus rc-1 isolated from cadmium-contaminated soil. Colloids Surf. B: Biointerfaces 107 (4B), 11–18. Hughes, M.N., Poole, R.K., 1989. Metals and Microorganisms. Chapman and Hall, London, UK. Jahromi, M.F., Liang, J.B., Ebrahimi, R., Soleimani, A.F., Rezaeizadeh, A., Abdullah, N., et al., 2017. Protective potential of lactobacillus species in lead toxicity model in broiler chickens. Animal 11, 755–761. Javed, M., Ahmad, I., Usmani, N., Ahmad, M., 2016. Studies on biomarkers of oxidative stress and associated genotoxicity and histopathology in Channa punctatus from heavy metal polluted canal. Chemosphere 151, 210–219. Jeon, J.S., Oh, J.-J., Kwak, H.C., Yun, H.-Y., Kim, H.C., Kim, Y.-M., Oh, S.J., Kim, S.K., 2018. Agerelated changes in sulfur amino acid metabolism in male C57BL/6 mice. Biomol. Ther. 26, 167–174.

Jin, X., Xiang, Z., Liu, Q., Chen, Y., Lu, F., 2017. Polyethyleneimine-bacterial cellulose bioadsorbent for effective removal of copper and lead ions from aqueous solution. Bioresour. Technol. 244, 844 Pt 1. Kiran, M.G., Pakshirajan, K., Das, G., 2017. Heavy metal removal from multicomponent system by sulfate reducing bacteria: mechanism and cell surface characterization. J. Hazard. Mater. 324, 62–70. Kuroda, M., Kuroda, H., Oshima, T., Takeuchi, F., Mori, H., Hiramatsu, K., 2003. Twocomponent system VraSR positively modulates the regulation of cell-wall biosynthesis pathway in Staphylococcus aureus. Mol. Microbiol. 49, 807–821. Lee, S.-E., Yoo, D.-H., Son, J., Cho, K., 2006. Proteomic evaluation of cadmium toxicity on the midge Chironomus riparius Meigen larvae. PROTEOMICS 6, 945–957. Lei, D.-Y., Liu, Z., Peng, Y.-H., Liao, S.-B., Xu, H., 2014. Biosorption of copper, lead and nickel on immobilized Bacillus coagulans using experimental design methodologies. Ann. Microbiol. 64, 1371–1384. Li, D., Xu, X., Yu, H., Han, X., 2017. Characterization of Pb2+ biosorption by psychrotrophic strain Pseudomonas sp. I3 isolated from permafrost soil of Mohe wetland in Northeast China. J. Environ. Manag. 196, 8–15. Li, X., Lan, S.M., Zhu, Z.P., Zhang, C., Zeng, G.M., Liu, Y.G., et al., 2018. The bioenergetics mechanisms and applications of sulfate-reducing bacteria in remediation of pollutants in drainage: a review. Ecotoxicol. Environ. Saf. 158, 162–170. Lin, M.-Y., Yen, C.-L., 1999. Antioxidative ability of lactic acid bacteria. J. Agric. Food Chem. 47, 1460–1466. Luo, G., Huang, L., Su, Y., Qin, Y., Xu, X., Zhao, L., Yan, Q., 2016. flrA, flrB and flrC regulate adhesion by controlling the expression of critical virulence genes in Vibrio alginolyticus. Emerging. Microbes Infect. 5, e85. Mard, S.A., Veisi, A., Ahangarpour, A., Gharib-Naseri, M.K., 2015. Gastric acid induces mucosal H2S release in rats by upregulating mRNA and protein expression of cystathionine gamma lyase. J. Physiol. Sci. 65, 545–554. Miao, Z.-Y., He, H., Tan, T., Zhang, T., Tang, J.-L., Yang, Y.-C., Shi, K.-Y., Tang, J., 2018. Biotreatment of Mn2+ and Pb2+ with sulfate-reducing bacterium Desulfuromonas alkenivorans S-7. J. Environ. Eng. 144, 04017116. Mikowska, M., Świergosz-Kowalewska, R., 2018. DNA damage in a liver tissue of metal exposed Clethrionomys glareolus. Chemosphere 199, 625–629. Modis, K., Maskey, M., Johnson, P., Szabo, C., Hellmich, M.R., Chao, C., 2017. Abstract 2980: cystathionine-beta-synthase (CBS)-derived hydrogen sulfide (H2S) supports proliferation, migration and bioenergetics in colon cancer-associated fibroblasts (CAFs). Cancer Res. 77, 2980. Mohiuddin, K.M., Ogawa, Y., Zakir, H.M., Otomo, K., Shikazono, N., 2011. Heavy metals contamination in water and sediments of an urban river in a developing country. Int. J. Environ. Sci. Technol. 8, 723–736. Morales, K.A., Lasagna, M., Gribenko, A.V., Yoon, Y., Reinhart, G.D., Lee, J.C., Cho, W., Li, P., Igumenova, T.I., 2011. Pb2+as modulator of protein–membrane interactions. J. Am. Chem. Soc. 133, 10599–10611. Naik, M.M., Shamim, K., Dubey, S.K., 2012. Biological characterization of lead resistant bacteria to explore role of bacterial metallothionein in lead resistance. Curr. Sci. 103, 1113–1130. Neupane, K.P., Pecoraro, V.L., 2011. Pb-207 NMR spectroscopy reveals that Pb(II) coordinates with glutathione (GSH) and tris cysteine zinc finger proteins in a PbS3 coordination environment. J. Inorg. Biochem. 105, 1030–1034. Niu, Z.S., Pan, H., Guo, X.P., Lu, D.P., Feng, J.N., Chen, Y.R., et al., 2018. Sulphate-reducing bacteria (srb) in the Yangtze estuary sediments: abundance, distribution and implications for the bioavailibility of metals. Sci. Total Environ. 634, 296. Oladipo, O.G., Ezeokoli, O.T., Maboeta, M.S., Bezuidenhout, J.J., Tiedt, L.R., Jordaan, A., Bezuidenhout, C.C., 2018. Tolerance and growth kinetics of bacteria isolated from gold and gemstone mining sites in response to heavy metal concentrations. J. Environ. Manag. 212, 357–366. Orrù, L., Salvetti, E., Cattivelli, L., Lamontanara, A., Michelotti, V., Capozzi, V., et al., 2014. Draft genome sequence of bacillus coagulans gbi-30,6086, a widely used sporeforming probiotic strain. Genome Announc. 2 (e01080-14-e01080-14). Pal, M.K., Ghosh, T.C., Ghosh, J.K., 1990. Studies on the conformation of and metal ion binding by teichoic acid of Staphylococcus aureus. Biopolymers 30, 273–277. Pan, Y.-X., Luo, Z., Zhuo, M.-Q., Wei, C.-C., Chen, G.-H., Song, Y.-F., 2018. Oxidative stress and mitochondrial dysfunction mediated Cd-induced hepatic lipid accumulation in zebrafish Danio rerio. Aquat. Toxicol. 199, 12–20. Patzer, S.I., Hantke, K., 1998. The ZnuABC high-affinity zinc uptake system and its regulator Zur in Escherichia coli. Mol. Microbiol. 28, 1199–1210. Piotrowska-Seget, Z., Cycoń, M., Kozdrój, J., 2005. Metal-tolerant bacteria occurring in heavily polluted soil and mine spoil. Appl. Soil Ecol. 28, 237–246. Pruden, A., Messner, N., Pereyra, L., Hanson, R.E., Hiibel, S.R., Reardon, K.F., 2007. The effect of inoculum on the performance of sulfate-reducing columns treating heavy metal contaminated water. Water Res. 41, 904–914. Qing, Z., Zhu, L., Li, X., Yang, S., Zou, Z., Guo, J., Cao, Z., Yang, R., 2017. A target-lighted dsDNA-indicator for high-performance monitoring of mercury pollution and its antagonists screening. Environ. Sci. Technol. 51, 11884–11890. Qu, C., Du, H., Ma, M., Chen, W., Cai, P., Huang, Q., 2018. Pb sorption on montmorillonitebacteria composites: a combination study by XAFS, ITC and SCM. Chemosphere 200, 427–436. Quintelas, C., Fernandes, B., Castro, J., Figueiredo, H., Tavares, T., 2008. Biosorption of Cr (VI) by three different bacterial species supported on granular activated carbon—a comparative study. J. Hazard. Mater. 153, 799–809. Rehman, K., Fatima, F., Waheed, I., Akash, M.S.H., 2017. Prevalence of exposure of heavy metals and their impact on health consequences. J. Cell. Biochem. 119, 157–184. Ren, G., Jin, Y., Zhang, C., Gu, H., Qu, J., 2015. Characteristics of Bacillus sp. PZ-1 and its biosorption to Pb(II). Ecotoxicol. Environ. Saf. 117, 141–148. Romaidi, Ueki, T., 2016. Bioaccumulation of vanadium by vanadium-resistant bacteria isolated from the intestine of Ascidia sydneiensis samea. Mar. Biotechnol. 18, 359–371.

S.-C. Xing et al. / Science of the Total Environment 693 (2019) 133490 Ryan, R.P., Monchy, S., Cardinale, M., Taghavi, S., Crossman, L., Avison, M.B., Berg, G., van der Lelie, D., Dow, J.M., 2009. The versatility and adaptation of bacteria from the genus Stenotrophomonas. Nat. Rev. Microbiol. 7, 514–525. Salvetti, E., Orrù, L., Capozzi, V., Martina, A., Lamontanara, A., Keller, D., et al., 2016. Integrate genome-based assessment of safety for probiotic strains:bacillus coagulansgbi-30, 6086 as a case study. Appl. Microbiol. Biotechnol. 100, 4595–4605. Sheibani, L., Lechuga, T.J., Zhang, H., Hameed, A., Wing, D.A., Kumar, S., Rosenfeld, C.R., Chen, D.-B., 2017. Augmented H2S production via cystathionine-beta-synthase upregulation plays a role in pregnancy-associated uterine vasodilation. Biol. Reprod. 96, 664–672. Silverman, M., Simon, M., 1974. Flagellar rotation and the mechanism of bacterial motility. Nature 249, 73–74. Sirko, A., Zatyka, M., Sadowy, E., Hulanicka, D., 1995. Sulfate and thiosulfate transport in Escherichia coli K-12: evidence for a functional overlapping of sulfate- and thiosulfate-binding proteins. J. Bacteriol. 177, 4134–4136. Statham, T.M., Stark, S.C., Snape, I., Stevens, G.W., Mumford, K.A., 2016. A permeable reactive barrier (PRB) media sequence for the remediation of heavy metal and hydrocarbon contaminated water: a field assessment at Casey Station, Antarctica. Chemosphere 147, 368–375. Stevanato, P., Broccanello, C., Moliterni, V.M.C., Mandolino, G., Barone, V., Lucini, L., Bertoldo, G., Bertaggia, M., Cagnin, M., Pizzeghello, D., et al., 2018. Innovative approaches to evaluate sugar beet responses to changes in sulfate availability. Front. Plant Sci. 9, 14. Thomas, K.J., Rice, C.V., 2014. Revised model of calcium and magnesium binding to the bacterial cell wall. BioMetals 27, 1361–1370. Tian, F., Zhai, Q., Zhao, J., Liu, X., Wang, G., Zhang, H., Zhang, H., Chen, W., 2012. Lactobacillus plantarum CCFM8661 alleviates lead toxicity in mice. Biol. Trace Elem. Res. 150, 264–271. Trevors, J.T., 1999. Why on earth: self-assembly of the first bacterial cell to abundant and diverse bacterial species. World J. Microbiol. Biotechnol. 15, 297–304. Waite, C.C., Da, S.G., Bitencourt, J.A., Sabadini-Santos, E., Crapez, M.A., 2016. Copper and lead removal from aqueous solutions by bacterial consortia acting as biosorbents. Mar. Pollut. Bull. 109, 386–392. Wang, F., Burrage, A.M., Postel, S., Clark, R.E., Orlova, A., Sundberg, E.J., Kearns, D.B., Egelman, E.H., 2017. A structural model of flagellar filament switching across multiple bacterial species. Nat. Commun. 8, 960. Wang, K., Tang, S.F., Hou, X., 2018. Molecular mechanism investigation on the interactions of copper (ii) ions with glutathione peroxidase 6 from Arabidopsis thaliana. Spectrochim. Acta A Mol. Biomol. Spectrosc. 203, 428–433. Wickett, N.J., Forrest, L.L., Budke, J.M., Shaw, B., Goffinet, B., 2011. Frequent pseudogenization and loss of the plastid-encoded sulfate-transport gene cysA throughout the evolution of liverworts. Am. J. Bot. 98, 1263–1275.

15

Williams, S.B., Stewart, V., 2002. Microreview: functional similarities among twocomponent sensors and methyl-accepting chemotaxis proteins suggest a role for linker region amphipathic helices in transmembrane signal transduction. Mol. Microbiol. 33, 1093–1102. Wu, G., Fang, Y.-Z., Yang, S., Lupton, J.R., Turner, N.D., 2004. Glutathione metabolism and its implications for health. J. Nutr. 134, 489–492. Wu, W., Wu, P., Yang, F., Sun, D.-L., Zhang, D.-X., Zhou, Y.-K., 2018. Assessment of heavy metal pollution and human health risks in urban soils around an electronics manufacturing facility. Sci. Total Environ. 630, 53–61. Xia, L., Chen, S., Dahms, H.-U., Ying, X., Peng, X., 2016. Cadmium induced oxidative damage and apoptosis in the hepatopancreas of Meretrix meretrix. Ecotoxicology 25, 959–969. Xu, T., Zheng, L., Xu, L., Yin, L., Qi, Y., Xu, Y., Han, X., Peng, J., 2014. Protective effects of dioscin against alcohol-induced liver injury. Arch. Toxicol. 88, 739–753. Yang, Z., Zhang, Z., Chai, L., Wang, Y., Liu, Y., Xiao, R., 2016. Bioleaching remediation of heavy metal-contaminated soils using Burkholderia sp. Z-90. J. Hazard. Mater. 301, 145–152. Yang, Y., Zhou, X., Tie, B., Peng, L., Li, H., Wang, K., Zeng, Q., 2017. Comparison of three types of oil crop rotation systems for effective use and remediation of heavy metal contaminated agricultural soil. Chemosphere 188, 148–156. Yepiskoposyan, H., Egli, D., Fergestad, T., Selvaraj, A., Treiber, C., Multhaup, G., Georgiev, O., Schaffner, W., 2006. Transcriptome response to heavy metal stress in Drosophila reveals a new zinc transporter that confers resistance to zinc. Nucleic Acids Res. 34, 4866–4877. Yin, R., Zhai, Q., Yu, L., Xiao, Y., Wang, G., Yu, R., Tian, F., Chen, W., 2016. The binding characters study of lead removal by Lactobacillus plantarum CCFM8661. Eur. Food Res. Technol. 242, 1621–1629. Yoneda, A., Henson, W.R., Goldner, N.K., Park, K.J., Forsberg, K.J., Kim, S.J., Pesesky, M.W., Foston, M., Dantas, G., Moon, T.S., 2016. Comparative transcriptomics elucidates adaptive phenol tolerance and utilization in lipid-accumulating Rhodococcus opacus PD630. Nucleic Acids Res. 44, 2240–2254. Zatakia, H.M., Arapov, T.D., Meier, V.M., Scharf, B.E., 2017. Cellular stoichiometry of methyl-accepting chemotaxis proteins in Sinorhizobium meliloti. J. Bacteriol. 200, e00614–e00617. Zhai, Q., Wang, G., Zhao, J., Liu, X., Tian, F., Zhang, H., Chen, W., 2013. Protective effects of Lactobacillus plantarum CCFM8610 against acute cadmium toxicity in mice. Appl. Environ. Microbiol. 79, 1508–1515. Zheng, G., Fu, Y., He, C., 2014. Nucleic acid oxidation in DNA damage repair and epigenetics. Chem. Rev. 114, 4602–4620. Zoghi, A., Khosravi-Darani, K., Sohrabvandi, S., 2014. Surface binding of toxins and heavy metals by probiotics. Mini-Rev. Med. Chem. 14, 84–98.