Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2020 Dec 10.
Published in final edited form as: Arch Toxicol. 2018 Oct 24;93(1):25–35. doi: 10.1007/s00204-018-2332-7

Gut microbiome disruption altered the biotransformation and liver toxicity of arsenic in mice

Liang Chi 1, Jingchuan Xue 1, Pengcheng Tu 1, Yunjia Lai 1, Hongyu Ru 2, Kun Lu 1,*
PMCID: PMC7727877  NIHMSID: NIHMS1524786  PMID: 30357543

Abstract

The mammalian gut microbiome (GM) plays a critical role in xenobiotic biotransformation and can profoundly affect the toxic effects of xenobiotics. Previous in vitro studies have demonstrated that gut bacteria have the capability to metabolize arsenic (As); however, the specific roles of the gut microbiota in As metabolism in vivo and the toxic effects of As are largely unknown. Here, we administered sodium arsenite to conventionally raised mice (with normal microbiomes) and GM-disrupted mice with antibiotics to investigate the role of the gut microbiota in As biotransformation and its toxicity. We found that the urinary total As levels of GM-disrupted mice were much higher, but the fecal total As levels were lower, than the levels in the conventionally raised mice. In vitro experiments, in which the GM was incubated with As, also demonstrated that the gut bacteria could adsorb or take up As and thus reduce the free As levels in the culture medium. With the disruption of the gut microbiota, arsenic biotransformation was significantly perturbed. Of note, the urinary monomethylarsonic acid (MMA)/ dimethylarsinic acid (DMA) ratio, a biomarker of arsenic metabolism and toxicity, was markedly increased. Meanwhile, the expression of genes of one-carbon metabolism, including forl2, bhmt, and mthfr, was downregulated, and the liver S-adenosylmethionine (SAM) levels were significantly decreased in the As-treated GM-disrupted mice only. Moreover, As exposure altered the expression of genes of the p53 signaling pathway, and the expression of multiple genes associated with hepatocellular carcinoma (HCC) was also changed in the As-treated GM-disrupted mice only. Collectively, disruption of the GM enhances the effect of As on one-carbon metabolism, which could in turn affect As biotransformation. GM disruption also increases the toxic effects of As and may increase the risk of As-induced HCC in mice.

Introduction

Arsenic (As) is a ubiquitous contaminant that affects the health of millions of people and is associated with a series of human diseases, such as diabetes, cardiovascular diseases, and multiorgan cancers (Naujokas et al. 2013). Arsenic toxicity depends largely on its biotransformation and on the specific As species. While As metabolism in mammals is not fully understood, inorganic As is generally methylated to monomethylarsonic acid (MMA) and then transformed to dimethylarsinic acid (DMA), which is the end product of most of the As (Vahter and Concha 2001). Most of the As is excreted in the urine. Generally, pentavalent As species, such as iAsV, MMAV, and DMAV, are less toxic than trivalent As species, such as iAsIII, MMAIII and DMAIII (Hirano et al. 2004; Naranmandura et al. 2007; Van de Wiele et al. 2010), and the methylated pentavalent species MMAV and DMAV are less toxic than iAsV (Hirano et al. 2004). Consequently, in previous studies, the MMA/DMA ratio has been used a biomarker to assess arsenic biotransformation and toxicity, with higher numbers being associated with increased arsenic toxicity.

There are many factors that influence the biotransformation and toxic effects of As, including sex, host genetics, nutrition status, diet, and lifestyles (Chen et al. 2006; Lindberg et al. 2007; Yu et al. 2016a). More recently, the interaction between the gut microbiome (GM) and As has emerged as an important factor for understanding As toxicity. It has been demonstrated that As exposure can perturb the profile and metabolism of the gut microbial community (Lu et al. 2014b). On the other hand, there is increasing evidence to indicate that gut bacteria can influence As biotransformation in mammals. For example, different GM phenotypes, driven by host genetics or pathogenic infection, are associated with different urinary As profiles (Lu et al. 2013a;Lu et al. 2014a). Genes associated with As metabolism are widely distributed in the GM (Isokpehi et al. 2014), and multiple previous in vitro studies have successfully detected the products of As oxidation, reduction, methylation, and thiolation by incubating inorganic As with the mammalian gut microbiota (Chen et al. 1996; Kubachka et al. 2009; Kuroda et al. 2004; Pinyayev et al. 2011; Rowland and Davies 1981). However, to date, in vitro incubators cannot accurately simulate the complex environment of the mammalian gastrointestinal (GI) tract, and these incubators cannot always mirror the true gut bacterial composition and metabolic activity seen in the GI tract (Guerra et al. 2012; Rajilić-Stojanović et al. 2010). In addition, in vitro studies do not replicate the toxicological responses to As in hosts. Thus, the mechanism of As toxicity under physiological conditions and that via which the GM influences As biotransformation remain unknown. Here, we compared the difference in As metabolism in conventionally raised and GM-disrupted mice. Our findings indicate that the gut microbiota plays a role in decreasing the As load, promoting As methylation, and protecting the host from the liver toxicity of As.

Materials and Methods

Animals, arsenic exposure, and sample collection

Sodium arsenite was obtained from Fisher Scientific (Pittsburgh, PA). 7-week-old female C57BL/6 mice (SPF grade) were purchased from Jackson Laboratories (Bar Harbor, ME) and maintained at the University of North Carolina animal facility in static microisolator cages with Bed-O-Cob combination bedding under standard environmental conditions (22°C, 40–70% humidity, and a 12:12-hour light:dark cycle). A purified rodent diet was provided. The mice were observed for one week before the start of the experiment. To generate GM-disrupted animals, we treated the mice with antibiotics for 72 hours before As exposure (Theriot et al. 2016). Feces were collected, and fecal DNA was extracted using the PowerSoil DNA Isolation Kit (Mo Bio Laboratories). The GM disruption was validated by PCR using the universal primers iTRU-A 515 F and iTRU-1 806 R to target the V4 regions of the bacterial 16S rRNA gene (515F: 5’-GTGCCAGCMGCCGCGGTAA-3’; 806R: 5’-GGACTACHVGGGTWTCTAAT-3’) (Caporaso et al. 2012). No signal could be amplified from the fecal DNA of GM-disrupted animals, which indicated that most of the gut bacteria had been removed (data not shown). Then, the mice were divided into six groups: group A, 10 conventionally raised mice as the control group; group B, 10 conventionally raised mice treated with 250 ppb As; group C, 10 conventionally raised mice treated with 1 ppm As; group D, 10 GM-disrupted mice without any As treatment as a GM-disrupted control group; group E, 10 GM-disrupted mice treated with 250 ppb As; and group F, 10 GM-disrupted mice treated with 1 ppm As. Both As and antibiotics were added to the drinking water, and the drinking water was replenished twice a week. Antibiotic treatment was performed throughout As exposure to maintain GM disruption. Drinking water consumption was monitored, and no significant difference was observed among the groups. Urine samples were collected after two weeks of exposure. Mice were euthanized with carbon dioxide and necropsied after the 2-week As exposure. Tissues were collected, and half of tissues were treated with RNAlater (Thermo Fisher Scientific) and the other half were immediately frozen in liquid nitrogen. All the samples were stored at −80°C until further analysis. The experiment was approved by the University of North Carolina Institutional Animal Care and Use Committee, and all the mice were treated humanely.

Anaerobic culture

Approximately 2 grams of fresh fecal sample was added to 5 mL of nitrogen-saturated phosphate-buffered saline (PBS) in a Hungate culture tube. The samples were vortexed for 10 min and centrifuged at 1,000 rpm for 5 min. Then, 1 mL of the supernatant was added to 10 mL of autoclaved nitrogen-saturated culture medium with 1 ppm As. The culture medium used in this study was modified from a previous study (Olano-Martin et al. 2000) (1 L of medium contained the following: peptone water, 2 g; yeast extract, 2 g; NaCl, 0.1 g; K2HPO4, 0.04 g; KH2PO4, 0.04 g; MgSO4-7H2O, 0.01 g; CaC12–6H2O, 0.01 g; NaHCO4, 2 g; hemin, 0.005 g; bile salts, 0.5 g; L-cysteine-HCl, 0.5 g; Tween 80, 2 mL; vitamin K1, 10 μL; 0.025% resazurin, 4 mL. pH=6.8). The tubes were sealed with aluminum caps and incubated at 37°C with gentle shaking (150 rpm) for 7 days. Four replicates were used, and one negative control sample (1 ppm As-containing medium without fecal extract) was included. Before and after incubation, 1 mL of the bacteria-containing medium was sampled and centrifuged at 1,000 rpm for 5 min. The bacterial pellet and upper layer of the medium were used to measure the total As in the cells and medium, respectively.

Total As and As species measurement

Arsenic species and total As were measured by an Agilent 7500 ICP-MS (Santa Clara, CA). Sample preparation and detection were conducted as previously described with several modifications (Lu et al. 2013a; Sanz et al. 2005). Briefly, the bacterial pellet from the culture medium, 20 mg of feces or 50 mg of liver samples were incubated with 100 μL of high-purity HNO3 for 24 hours and then heated at 85°C for 24 hours. Then, the samples were cooled to room temperature, treated with 200 μL of 30% H2O2 and incubated at 85°C for 2 hours. The cooled samples were injected to an ICP-MS instrument at the proper dilution. To determine the total As in urine and culture medium, the samples diluted with the mobile phase solution were directly analyzed by ICP-MS. (NH4)H2PO4 was used as the mobile phase, and the pH was adjusted to 6.0 with 85% phosphoric acid. For measurement of the urinary As species, the collected urine samples were diluted, and As species were separated by a Hamilton PRP-X100 column in an Agilent 1260 HPLC instrument, and then, 1 μL of the sample was injected into the ICP-MS instrument. The mobile phase A was 10 mM ammonium carbonate and 10 mM Tris (pH 8.7), and the mobile phase B was 10 mM ammonium carbonate, 10 mM Tris, and 15 mM ammonium sulfate (pH 8.0). The gradient was as follows: (1) 0 to 5 min: 0% B to 100% B; (2) 5 to 11 min: 100% B; (3) 11-17 min: 100% A.

Liver RNA-Seq library preparation, sequencing and data analysis

The total liver RNA of control mice, GM-disrupted mice, 250 ppb As-treated conventionally raised mice and 250 ppb As-treated GM-disrupted mice were extracted using the RNeasy Mini Kit (Qiagen, Valencia, CA) according to the manufacturer’s instruction and then treated with DNase (DNA-free™ DNA Removal Kit, Thermo Fisher Scientific) to remove DNA contamination. RNA quality was determined by an Agilent 4200 TapeStation (Agilent Technologies). The RNA-Seq library was prepared using KAPA Stranded mRNA-Seq Kit (Kapa Biosystems) following the manufacturer’s protocol. The libraries were sequenced at the Georgia Genomics Facility using an Illumina NextSeq high-output flow cell. Approximately 10 million reads were generated for each sample. Sequencing data were aligned by HISAT2 (Kim et al. 2015) in the Galaxy server (https://usegalaxy.org/. Cufflinks was used to assemble the transcripts, estimate their abundances and detect the significant changes in transcript expression (Trapnell et al. 2010). Overrepresentation enrichment analysis (ORA) was conducted by WEB-based GEne SeT AnaLysis Toolkit (http://www.webgestalt.org/option.php# (Wang et al. 2017; Zhang et al. 2005) by mapping the gene to the KEGG database to analyze the enriched pathways.

Liver SAM measurement

Fifty milligrams of liver sample was lysed by a TissueLyser in 300 μL of a chloroform:methanol (2:1) solution for 5 min at 50 Hz. Then, 300 μL H2O was added, which was followed by vortexing and centrifugation at 15,000×g for 10 min. The upper and lower phases were transferred to a flat-bottom HPLC vial. Then, 100 μL of water and 100 μL of methanol were added to the pellet, and the pellet was lysed in the TissueLyser for 5 min at 50 Hz and centrifuged at 15,000×g for 10 min. The supernatant was collected and added to the vial to combine all of extracts, and the sample was dried by a SpeedVac. Next, 500 μL of 50% acetonitrile was added to dissolve the metabolites, which was followed by filtration with 0.22-μm filters. Identification and quantification of S-adenosylmethionine (SAM) was performed with a Thermo Finnigan TSQ Quantum electrospray triple quadrupole mass spectrometer (ESI-MS/MS). A Synergi 4 μm Fusion-RP 80A (50×2.00 mm) column was used. The injection volume was 10 μL, and the mobile phase comprised Milli-Q water containing 0.1% formic acid (A) and acetonitrile containing 0.1% formic acid (B). The mobile phase gradient was as follows: starting at 2% (v/v) B, held for 5 min, increased to 95% B over 6 min (total 11 min), reverted to 2% B over 0.5 min (total 11.5 min), held for 4.5 min (total 15 min); the total run time was 15 min. The flow rate of the mobile phase was 100 μL/min. The collision energy was set as 14 and 16 V for SAM (399.1/250) and d3-SAM (402/250), respectively. Nitrogen was used as both the curtain and collision gas. Data acquisition was performed at a scan speed of 40 ms and a resolving power of 0.70 full width at half maximum (FWHM).

Statistical analysis

A two-tailed Student’s t-test was used to determine the statistical significance of the total As, As species, and liver SAM levels for each group. A p-value of <0.05 was considered significantly different. Significant differences in FPKM values for individual transcripts between the control and other groups were determined statistically using Cuffdiff; a q-value <0.05 was considered indicative of a significant difference between two groups to address false positive hits from multiple comparisons.

Results

Urinary MMA and the MMA/DMA ratio significantly increased in As-exposed GM-disrupted mice

To investigate whether As metabolism was affected by GM disruption, we compared the urinary As profiles of conventionally raised mice and GM-disrupted mice. As expected, DMA was the most abundant As species and exhibited a dose-dependent increase (Fig. 1A). We found that GM-disrupted mice had higher levels of urinary DMA (~2-fold) than conventionally raised mice. Moreover, the MMA concentration increased much more in the GM-disrupted mice (~7-fold in the 250 ppb As-exposed group and ~20-fold in the 1 ppm As-exposed group) than in the conventionally raised mice (Fig. 1B). However, iAsV levels decreased significantly in the 1 ppm As-exposed GM-disrupted animals, while these levels did not change significantly in the 250 ppb group (Fig. 1C). Urinary iAsIII concentrations were not significantly different between the two types of mice, as shown in Fig. 1D. In particular, the MMA/DMA ratio, a biomarker of arsenic metabolism and toxicity, was much higher in the GM-disrupted mice than in the conventionally raised mice (Fig. 1E). These results indicate that As metabolism was altered in the GM-disrupted animals, and in particular, the MMA levels in the urine were greatly increased.

Fig. 1.

Fig. 1

Open in a new tab

Urinary As profiles of As-treated conventionally raised and GM-disrupted mice. DMA (A) and MMA (B) levels were significantly increased in GM-disrupted mice. iAsV levels were significantly decreased in 1ppm As-treated GM-disrupted mice compared with the levels in 1 ppm As-treated conventionally raised animals (C). There was no significant difference in iAsIII levels between the conventionally raised and GM-disrupted mice at the same As dose (D). The MMA/DMA ratio was greatly increased in GM-disrupted mice (E). (Microbiome: “+”, normal GM; “−” disrupted GM. N.S., no significant difference. n=10)

Urinary total As levels increased but fecal As levels decreased in GM-disrupted mice

To further examine the effect of the GM on arsenic excretion, we measured the total As in the urine, feces and liver. Consistent with previous data, the urinary total As levels in GM-disrupted mice were approximately 2-fold higher than those in As-treated conventionally raised mice (Fig. 2A). Liver As levels were not significantly different between conventionally raised mice and GM-disrupted mice (Fig. 2C). However, we found that the fecal As levels decreased significantly in GM-disrupted mice (Fig. 2B). These results indicate that the absence of the gut microbiota led to a high As load in the mice but did not increase As deposition in the livers.

Fig. 2.

Fig. 2

Open in a new tab

Total As in urine (A), feces (B) and liver (C). Urinary As levels were significantly increased (A) but fecal As levels were decreased (B) in GM-disrupted mice. Liver As levels were not significantly different between conventionally raised and GM-disrupted mice at the same dose (C). (Microbiome: “+”, normal GM; “−”, disrupted GM. N.S., no significant difference.n=10)

In vitro culture indicates that the gut microbiome takes up As

Based on the greatly increased As levels in the urine and the decreased As levels in the feces of GM-disrupted mice, we hypothesized that the GM could take up As, which reduces the absorbable As and thus decreases the total As load available to the host. To test our hypothesis, we incubated As with the GM in a batch culture system and measured the As absorption by the gut bacteria. After a 7-day incubation, we found that the As levels in the bacterial cells increased substantially (Fig. 3A). On the other hand, the As levels in the culture medium decreased significantly after incubation with the bacteria (Fig. 3B).

Fig. 3.

Fig. 3

Open in a new tab

Total As levels in bacterial cells (A) and in culture medium (B) before incubation and after a 1-week incubation. Arsenic was highly enriched in bacteria (A), and the As levels in the medium decreased accordingly (B) after incubation with gut bacteria (n=4).

Expression of genes associated with one-carbon metabolism and SAM concentrations decreased significantly in the livers of GM-disrupted mice

Given the changes in the levels of the As species, we next investigated whether the expression of genes associated with As metabolism was perturbed in the liver, the main organ for arsenic metabolism in the host. We did not observe significant changes in the expression of as3mt, which plays a central role in As methylation (Thomas 2007). However, several genes associated with one-carbon metabolism, including folr2, mthfr, and bhmt, were significantly downregulated in As-treated GM-disrupted mice but not in conventionally raised mice treated with arsenic (Fig. 4A). In addition, expression of the gclc gene, which encodes the enzyme that catalyzes the rate-limiting step of GSH synthesis (Shi et al. 1994), decreased in the As-treated GM-disrupted mice (Fig. 4). In addition, the gene gpx3, which utilizes GSH to reduce H2O2, was also downregulated in As-exposed GM-disrupted mice. GSH is a key factor of ROS detoxification and plays a crucial role in As metabolism and detoxification (Scott et al. 1993; Thomas 2007). Since one-carbon metabolism produces SAM, which is the methyl donor in As methylation, we then measured the SAM levels in the liver. There were no significant changes in SAM levels in either GM-disrupted mice or conventionally raised mice treated with 250 ppb arsenic (Fig. 4B). However, arsenic induced a significant decrease in liver SAM levels in GM-disrupted mice (p<0.02). These results indicate that As exposure may alter As biotransformation by influencing one-carbon metabolism and the cofactors of As methylation but not by directly altering the expression of as3mt. Notably, GM disruption sensitizes the effects of arsenic on one-carbon metabolism and the cofactors of As methylation.

Fig. 4.

Fig. 4

Open in a new tab

Significantly altered genes involved in folate one-carbon metabolism (q<0.05) and liver SAM levels. The 250-ppb As exposure specifically downregulated multiple genes associated with folate one-carbon metabolism and GSH synthesis in GM-disrupted mice (A). SAM levels were significantly decreased in the livers of the As-treated GM-disrupted mice (B). The framed image shows the related pathways involved in one-carbon metabolism and their relationships with As metabolism (C). (Microbiome: “+”, normal GM; “−”, disrupted GM. N.S., no significant difference. n=5)

As exposure altered the p53 signaling pathway in GM-disrupted mice and perturbed the expression of multiple HCC-related genes

Since the liver is a target organ of As toxicity, we further investigated the differential effects of As exposure on gene expression in the livers of conventionally raised mice and of those with disrupted microbiomes. The ORA indicated that the expression of multiple genes involved in the p53 signaling pathway was altered in As-exposed GM-disrupted animals (Fig. 5). Moreover, we found that the expression of multiple hepatocellular carcinoma (HCC)-related genes, including stard13, vegfa, azin1, spp1, htatip2, osgin1, alcam, ptma, trib2, and atoh8, was significantly altered in As-exposed GM-disrupted mice only (Table 1). According to previous studies, the overexpression or inhibition of these genes is involved in the HCC development (Table 1). In GM-disrupted mice, the change patterns of these genes induced by arsenic resemble those observed in HCC (Table 1), which indicates enhanced liver toxicity and a potentially increased risk of developing HCC.

Fig. 5.

Fig. 5

Open in a new tab

ORA analysis indicates that As (250 ppb) exposure altered the expression of multiple genes associated with the p53 signaling pathway in GM-disrupted mice. (Microbiome: “+”, normal GM; “−”, disrupted GM. n=5)

Table 1.

Alterations in the expression of key genes involved in HCC in GM-disrupted mice treated with 250 ppb As.

Gene name Alteration in As-exposed GM-disrupted mice q-value Potential role in HCC Alteration in HCC Reference
Vegfa Upregulated 0.009 Inducing vasculogenesis or angiogenesis and promoting cell growth Overexpressed (Chai et al. 2013; Llovet and Bruix 2008)
Azin1 Upregulated 0.002 Promoting HCC cell proliferation Overedited (Chen et al. 2013)
Spp1 Upregulated 0.002 Inhibiting HCC cell invasion and metastasis Overexpressed (Chen et al. 2014b; Shin et al. 2007)
Alcam Upregulated 0.022 Promoting cell invasion Overexpressed (Borlak et al. 2015)
Ptma Upregulated 0.008 Inhibiting apoptosis and promoting cell survival Overexpressed (Lin and Chao 2015;Magdalena et al. 2000;Wu et al. 1997)
Trib2 Upregulated 0.016 Promoting cell survival and transformation Overexpressed (Qiao et al. 2013;Wang et al. 2013)
Stard13 Downregulated 0.002 Suppressing cell growth and promoting apoptosis Underexpressed (Hanna et al. 2014; Zhang et al. 2017)
Atoh8 Downregulated 0.020 Inhibiting cell migration and invasion Underexpressed (Song et al. 2015; Zhao and Yu 2016)
Htatip2 Downregulated 0.035 Inhibiting angiogenesis to promote tumor growth Underexpressed (Lu et al. 2013b; Wang et al. 2014)
Osgin1 Downregulated 0.017 Inhibiting apoptosis Underexpressed (Jeng et al. 2015; Liu et al. 2014)
Open in a new tab

Discussion

In this study, we investigated As metabolism and its effects on the livers of GM-disrupted mice whose GMs had been largely removed by antibiotic treatment. The role of gut bacteria in As metabolism is being increasingly recognized. The ability of the gut microbiota to biotransform As has been demonstrated, and previous studies have suggested that As can be rapidly and completely absorbed from the gastrointestinal tract, indicating that the role of the GM in As metabolism is relatively unimportant (Hillman 2004). However, we found that mice that lacked the gut microbiome had much higher levels of urinary As but significantly lower levels of fecal As than conventionally raised mice (Fig. 1 and 2). This finding clearly indicates that the existence of the gut microbiome can significantly affect arsenic excretion and decrease the total As load in the host. Arsenic can be absorbed by gut bacteria via ion channels in the cell membrane (Isokpehi et al. 2014), thus reducing the total As available to host bodies. By in vitro experiments, we determined that the GM has strong As absorption or uptake ability (Fig. 3). Moreover, the gut microbiota can also influence As absorption by changing the physical and chemical environment in the GI tract or by altering host physiological conditions (Ruby et al. 1996). Further research needs to be conducted to reveal the mechanism via which the GM reduces the As load.

We found that the GM-disrupted mice had less urinary iAsV than the conventionally raised animals. In this study, we used arsenite (iAsIII) for exposure, and generally, in mammals, As is metabolized from iAsV to iAsIII and then to methylated As. However, arsenite oxidase, which converts iAsIII to iAsV is widely present in bacteria (Hamamura et al. 2009; Inskeep et al. 2007). Metagenomic sequencing demonstrated the existence of As oxidase in the GM, and in vitro studies have also shown that iAsIII can be oxidized to iAsV by human gut bacteria (Chen et al. 2014c; Yu et al. 2016b). Therefore, the low proportion of iAsV in GM-disrupted mice may be caused by the disruption of gut bacteria, which again highlights the influence of gut microbiota on the metabolic fate of As in mammals. In addition to the increase in total As, the MMA levels as well as the MMA/DMA ratio were significantly increased in the urine of GM-disrupted animals (Fig. 1). These data indicate that the As methylation efficiency was altered in the GM-disrupted animals.

In mammals, iAs is first methylated to MMA and then DMA (Hernández and Marcos 2008). AS3MT is the primary enzyme that performs As methylation in mice and humans (Thomas 2007). However, our current study indicates that the expression of as3mt did not differ significantly between the different groups (Fig. 4). Therefore, the alteration in the As methylation efficiency of GM-disrupted mice was not caused by the downregulation of as3mt expression; however, we cannot rule out the possibility that the catalytic activity of AS3MT was altered in the GM-disrupted mice. One-carbon metabolism plays a critical role in As biotransformation. As shown in Fig. 4C, the folate-associated one-carbon metabolism product SAM, which is the methyl group donor for AS3MT, participates in As methylation (Gamble et al. 2005). Folr2 encodes a folate-binding protein that functions as a folate receptor to transport folate into cells (Wibowo et al. 2013). Previous studies have shown that Folr2-knockout mice had lower plasma SAM levels and higher sensitivity to As-induced toxic effects than conventionally raised mice (Crandall and Vorce 2002; Spiegelstein et al. 2005). In this study, the expression of folr2 and the liver SAM levels were significantly reduced in GM-disrupted mice (Fig. 4). In addition, mthfr and bhmt encode two key enzymes in the one-carbon metabolic pathways (Lee et al. 2009), and both of these genes exhibited lower expression in As-exposed GM-disrupted mice than in the controls (Fig. 4). Therefore, alterations in the expression of genes associated with one-carbon metabolism might influence the production of SAM, which is a potential mechanism for the increase in urinary MMA observed in the GM-disrupted animals. Moreover, GSH also plays a critical role in As metabolism as well as in the detoxification of ROS (Flora 2011). Here, the gclc gene, which is a rate-limiting gene in GSH synthesis, was downregulated in the GM-disrupted mice (Fig. 4), and this gene might also affect As metabolism.

The urinary As profile, especially the MMA%, has been found to be associated with the risks of many As-related diseases. For example, case-control studies in Taiwan found that high urinary MMA% and low urinary DMA% were associated with a high risk of urothelial carcinoma (Huang et al. 2008; Pu et al. 2007). Likewise, patients with other As-related diseases, including skin cancer, lung cancer, skin lesions, and hypertension, have also been found have high urinary MMA% (Kile et al. 2011; Kuo et al. 2017; Li et al. 2013; Steinmaus et al. 2010; Wei et al. 2017). In the present study, we observed a large increase in MMA% in the urine of As-treated GM-disrupted animals, which may suggest that these mice suffered higher As toxic effects than conventionally raised mice.

It is known that As is a class I carcinogen and can induce multiorgan cancers, such as skin cancer, lung cancer, and bladder cancer (Naujokas et al. 2013). Although whether As can induce liver cancer was controversial for a long time, accumulating evidence indicates that the liver is a target organ of As carcinogenesis, and both HCC and hepatic angiosarcoma are frequently associated with As exposure (Liu and Waalkes 2008). For example, previous studies found that maternal gestational exposure to As increased the incidence of HCC in offspring (Waalkes et al. 2007; Waalkes et al. 2004). Epidemiological studies have also indicated the association of As exposure with liver cancer mortality (Liu and Waalkes 2008). In this study, As exposure altered the gene expression patterns in the livers of GM-disrupted animals, and ORA indicated that genes in the p53 signaling pathway were specifically altered (Fig. 5). As an important tumor-suppressing pathway, the p53 signaling pathway plays a critical regulatory role in the cell cycle and apoptosis under stress (Haupt et al. 2003). For example, the gene ccnd2 encodes the cyclin D2, which regulates cyclin-dependent kinases (CDKs) and promotes cell proliferation (Pines 1995). Overexpression of ccnd2 has been observed in some tumors (Houldsworth et al. 1997; Sicinski et al. 1996). In contrast, the cyclin G2, which is encoded by ccng2, is an inhibitor of cell cycle progression, and previous studies have indicated that ccng2 is downregulated in several tumors (Chen et al. 2014a; Choi et al. 2009; Hasegawa et al. 2015; Sun et al. 2014). In this study, ccnd2 was upregulated but ccng2 was downregulated in the GM-disrupted mice (Fig. 5). In addition, we found that As exposure altered the expression of many of genes that are directly associated with HCC in GM-disrupted mice, including stard13, vegfa, azin1, spp1, htatip2, osgin1, alcam, ptma, trib2, and atoh8 (Table 1). Importantly, multiple genes that play a positive role in HCC development or survival were upregulated in As-exposed GM-disrupted mice. For example, the vascular endothelial growth factor A (VegfA), which is encoded by vegfa, is a member of that VEGF family, which plays a central role in angiogenesis and neovascularization and in functions such as promotion of cell migration, and inhibition of apoptosis (Cursiefen et al. 2004; Ma et al. 2015). A previous study found that vegfa was highly expressed even in early hepatocarcinogenesis (Llovet and Bruix 2008), and miR-26-induced downregulation of vegfa could suppress tumor growth (Chai et al. 2013). Prothymosin alpha (PTMA), which is encoded by the ptma gene, plays essential roles in multiple physiological processes, such as cell proliferation, oxidative stress response and apoptosis (Gomez-Marquez et al. 1989; Karapetian et al. 2005; Malicet et al. 2006), and has been found to be able to protect HCC cells against sorafenib-induced cell death (Lin and Chao 2015). High expression of ptma has been observed in multiple cancers, including HCC, which makes PTMA a prognostic for cancer (Magdalena et al. 2000; Wu et al. 1997). Similarly, trib2, which encodes tribbles homolog 2 (TRIB2), is a target gene of Wnt signaling, and previous studies found that the trib2 gene was highly activated in liver cancer cells and plays a critical role in cancer cell survival and transformation (Qiao et al. 2013; Wang et al. 2013). On the other hand, some negatively regulated genes in GM-disrupted mice have low expression levels in HCC, including stard13, osgin1, htatip2, and atoh8 (Table 1). Stard13 can induce apoptosis and function as a tumor suppressor gene in many cancers, including HCC (Hanna et al. 2014; Zhang et al. 2017). A previous study found that the expression of stard13 was decreased in clinical HCC tissues (Zhang et al. 2017). Oxidative stress-induced growth inhibitor 1 (Osgin1), which is encoded by the gene osgin1, is also a tumor suppressor and acts by promoting apoptosis; this gene is also downregulated in HCC (Jeng et al. 2015; Liu et al. 2014). Similarly, atonal homolog 8 (Atoh8), which is encoded by atoh8, is a transcription factor and functions as a potential tumor suppressor; the expression of this gene has been seen to be inhibited in HCC (Song et al. 2015). Therefore, As exposure altered the gene expression pattern in GM-disrupted mouse livers to be similar to that observed in HCC tumors, suggesting that As exposure may increase the cancer risk in GM-disrupted mice.

Conclusion

Taken together, the results of the current study demonstrate that upon disruption of GM, the total As load and urinary MMA levels increased greatly, while the fecal As levels decreased significantly. Arsenic exposure did not alter as3mt expression in the livers of GM-disrupted mice but decreased the expression of multiple genes involved in one-carbon metabolism and reduced the liver SAM concentrations, which could influence As methylation. Moreover, in GM-disrupted animals, As disturbed the expression of genes, including the p53 signaling pathway, and up-regulated many oncogenes that exhibit high expression in HCC but down-regulated some tumor suppressor genes, which suggests an increased risk for HCC. Our study demonstrates the important role of the gut microbiota on the host As load, biotransformation and the toxic effects of As on the liver.

Acknowledgments

We thank the NIH grant (R01ES024950) for financial support for the project.

References

  1. Borlak J, Meier T, Halter R, Spanel R, Spanel-Borowski K (2005) Epidermal growth factor-induced hepatocellular carcinoma: gene expression profiles in precursor lesions, early stage and solitary tumours. Oncogene 24(11): 1809. [DOI] [PubMed] [Google Scholar]
  2. Caporaso JG, Lauber CL, Walters WA, et al. (2012) Ultra-high-throughput microbial community analysis on the Illumina HiSeq and MiSeq platforms. ISME J 6(8): 1621. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Chai Z-T, Kong J, Zhu X-D, et al. (2013) MicroRNA-26a inhibits angiogenesis by down-regulating VEGFA through the PIK3C2α/Akt/HIF-lα pathway in hepatocellular carcinoma. PLoS ONE 8(10):e77957. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Chen H, Yoshida K, Wanibuchi H, Fukushima S, Inoue Y, Endo G (1996) Methylation and demethylation of dimethylarsinic acid in rats following chronic oral exposure. Appl Organomet Chem 10(9):741–745 [Google Scholar]
  5. Chen J, Liu C, Wen H, et al. (2014a) Changes in the expression of cyclin G2 in esophageal cancer cell and its significance. Tumor Biology 35(4):3355–3362 [DOI] [PubMed] [Google Scholar]
  6. Chen L, Li Y, Lin CH, et al. (2013) Recoding RNA editing of AZIN1 predisposes to hepatocellular carcinoma. Nat Med 19(2):209–216 [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Chen R, Dong Y, Xie X, et al. (2014b) Screening candidate metastasis-associated genes in three-dimensional HCC spheroids with different metastasis potential. International journal of clinical and experimental pathology 7(5):2527. [PMC free article] [PubMed] [Google Scholar]
  8. Chen Y, Graziano JH, Parvez F, et al. (2006) Modification of risk of arsenic-induced skin lesions by sunlight exposure, smoking, and occupational exposures in Bangladesh. Epidemiology 17(4):459–467 [DOI] [PubMed] [Google Scholar]
  9. Chen Y, Qin N, Guo J, et al. (2014c) Functional gene arrays-based analysis of fecal microbiomes in patients with liver cirrhosis. BMC Genomics 15(1):753. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Choi M-G, Noh JH, An JY, et al. (2009) Expression Levels of Cyclin G2, But Not Cyclin E, Correlate With Gastric Cancer Progressionl. J Surg Res 157(2): 168–174 [DOI] [PubMed] [Google Scholar]
  11. Crandall LZ, Vorce RL (2002) Differential effects of arsenic on folate binding protein 2 (Folbp2) null and wild type fibroblasts. Toxicol Lett 136(l):43–54 [DOI] [PubMed] [Google Scholar]
  12. Cursiefen C, Chen L, Borges LP, et al. (2004) VEGF-A stimulates lymphangiogenesis and hemangiogenesis in inflammatory neovascularization via macrophage recruitment. J Clin Invest 113(7): 1040. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Flora SJ (2011) Arsenic-induced oxidative stress and its reversibility. Free Radic Biol Med 51(2):257–281 [DOI] [PubMed] [Google Scholar]
  14. Gamble MV, Liu X, Ahsan H, et al. (2005) Folate, homocysteine, and arsenic metabolism in arsenic-exposed individuals in Bangladesh. Environ Health Perspect 113(12): 1683. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Gomez-Marquez J, Segade F, Dosil M, Pichel JG, Bustelo XR, Freire M (1989) The expression of prothymosin alpha gene in T lymphocytes and leukemic lymphoid cells is tied to lymphocyte proliferation. J Biol Chem 264(15): 8451–8454 [PubMed] [Google Scholar]
  16. Guerra A, Etienne-Mesmin L, Livrelli V, Denis S, Blanquet-Diot S, Alric M (2012) Relevance and challenges in modeling human gastric and small intestinal digestion. Trends Biotechnol 30(11):591–600 [DOI] [PubMed] [Google Scholar]
  17. Hamamura N, Macur R, Korf S, et al. (2009) Linking microbial oxidation of arsenic with detection and phylogenetic analysis of arsenite oxidase genes in diverse geothermal environments. Environ Microbiol 11 (2):421–431 [DOI] [PubMed] [Google Scholar]
  18. Hanna S, Khalil B, Nasrallah A, et al. (2014) StarD13 is a tumor suppressor in breast cancer that regulates cell motility and invasion. Int J Oncol 44(5): 1499–1511 [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Hasegawa S, Nagano H, Konno M, et al. (2015) Cyclin G2: A novel independent prognostic marker in pancreatic cancer. Oncology letters 10(5):2986–2990 [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Haupt S, Berger M, Goldberg Z, Haupt Y (2003) Apoptosis-the p53 network. J Cell Sci 116(20):4077–4085 [DOI] [PubMed] [Google Scholar]
  21. Hernández A, Marcos R (2008) Genetic variations associated with interindividual sensitivity in the response to arsenic exposure. Pharmacogenomics 9(8): 1113–1132 [DOI] [PubMed] [Google Scholar]
  22. Hillman B (2004) Role of gut bacteria in human toxicology and pharmacology. CRC Press [Google Scholar]
  23. Hirano S, Kobayashi Y, Cui X, Kanno S, Hayakawa T, Shraim A (2004) The accumulation and toxicity of methylated arsenicals in endothelial cells: important roles of thiol compounds. Toxicol Appl Pharmacol 198(3):458–467 [DOI] [PubMed] [Google Scholar]
  24. Houldsworth J, Reuter V, Bosl GJ, Chaganti R (1997) Aberrant expression of cyclin D2 is an early event in human male germ cell tumorigenesis. Cell growth & differentiation: the molecular biology journal of the American Association for Cancer Research 8(3):293–299 [PubMed] [Google Scholar]
  25. Huang Y-K, Huang Y-L, Hsueh Y-M, et al. (2008) Arsenic exposure, urinary arsenic speciation, and the incidence of urothelial carcinoma: a twelve-year follow-up study. Cancer Causes & Control 19(8): 829–839 [DOI] [PubMed] [Google Scholar]
  26. Inskeep WP, Macur RE, Hamamura N, Warelow TP, Ward SA, Santini JM (2007) Detection, diversity and expression of aerobic bacterial arsenite oxidase genes. Environ Microbiol 9(4):934–943 [DOI] [PubMed] [Google Scholar]
  27. Isokpehi RD, Udensi UK, Simmons SS, et al. (2014) Evaluative profiling of arsenic sensing and regulatory systems in the human microbiome project genomes. Microbiology insights 7:25. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Jeng K-S, Chang C-F, Jeng W-J, Sheen I-S, Jeng C-J (2015) Heterogeneity of hepatocellular carcinoma contributes to cancer progression. Crit Rev Oncol Hematol 94(3):337–347 [DOI] [PubMed] [Google Scholar]
  29. Karapetian RN, Evstafieva AG, Abaeva IS, et al. (2005) Nuclear oncoprotein prothymosin a is a partner of Keapl: implications for expression of oxidative stress-protecting genes. Mol Cell Biol 25(3): 1089–1099 [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Kile ML, Hoffman E, Rodrigues EG, et al. (2011) A pathway-based analysis of urinary arsenic metabolites and skin lesions. Am J Epidemiol 173(7):778–786 [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Kim D, Langmead B, Salzberg SL (2015) HIS AT: a fast spliced aligner with low memory requirements. Nat Methods 12(4):357–360 [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Kubachka KM, Kohan MC, Conklin SD, Herbin-Davis K, Creed JT, Thomas DJ (2009) In vitro biotransformation of dimethylarsinic acid and trimethylarsine oxide by anaerobic microflora of mouse cecum analyzed by HPLC-ICP-MS and HPLC-ESI-MS. J Anal At Spectrom 24(8): 1062–1068 [Google Scholar]
  33. Kuo C-C, Moon KA, Wang S-L, Silbergeld E, Navas-Aden A (2017) The Association of Arsenic Metabolism with Cancer, Cardiovascular Disease and Diabetes: A Systematic Review of the Epidemiological Evidence Environmental Health Perspectives. Environ Health Perspect 87001:1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Kuroda K, Yoshida K, Yoshimura M, et al. (2004) Microbial metabolite of dimethylarsinic acid is highly toxic and genotoxic. Toxicol Appl Pharmacol 198(3):345–353 [DOI] [PubMed] [Google Scholar]
  35. Lee YL, Xu X, Wallenstein S, Chen J (2009) Gene expression profiles of the one-carbon metabolism pathway. Journal of Genetics and Genomics 36(5):277–282 [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Li X, Li B, Xi S, Zheng Q, Wang D, Sun G (2013) Association of urinary monomethylated arsenic concentration and risk of hypertension: a cross-sectional study from arsenic contaminated areas in northwestern China. Environ Health 12(1):37. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Lin Y-T, Chao CC-K (2015) Identification of the β-catenin/JNfCprothyrnosin-alpha axis as a novel target of sorafenib in hepatocellular carcinoma cells. Oncotarget 6(36):38999. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Lindberg A-L, Kumar R, Goessler W, et al. (2007) Metabolism of low-dose inorganic arsenic in a central European population: influence of sex and genetic polymorphisms. Environ Health Perspect 115(7):1081. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Liu J, Waalkes MP (2008) Liver is a target of arsenic carcinogenesis. Toxicol Sci 105(l):24–32 [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Liu M, Li Y, Chen L, et al. (2014) Allele-specific imbalance of oxidative stress-induced growth inhibitor 1 associates with progression of hepatocellular carcinoma. Gastroenterol 146(4): 1084–1096. e5 [DOI] [PubMed] [Google Scholar]
  41. Llovet JM, Bruix J (2008) Molecular targeted therapies in hepatocellular carcinoma. Hepatology 48(4):1312–1327 [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Lu K, Cable PH, Abo RP, et al. (2013a) Gut microbiome perturbations induced by bacterial infection affect arsenic biotransformation. Chem Res Toxicol 26(12): 1893–1903 [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Lu K, Mahbub R, Cable PH, et al. (2014a) Gut microbiome phenotypes driven by host genetics affect arsenic metabolism. Chem Res Toxicol 27(2): 172–174 [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Lu K, Ryan PA, Schlieper KA, et al. (2014b) Arsenic exposure perturbs the gut microbiome and its metabolic profile in mice: an integrated metagenomics and metabolomics analysis. Environ Health Perspect 122(3):284–291 [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Lu L, Sun H-C, Zhang W, et al. (2013b) Aspirin minimized the pro-metastasis effect of sorafenib and improved survival by up-regulating HTATIP2 in hepatocellular carcinoma. PLoS ONE 8(5):e65023. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Ma X, Shen D, Li H, et al. MicroRNA-185 inhibits cell proliferation and induces cell apoptosis by targeting VEGFA directly in von Hippel-Lindau-inactivated clear cell renal cell carcinoma In: Urologic Oncology: Seminars and Original Investigations, 2015. vol 33 Elsevier, p 169.el–169.ell [DOI] [PubMed] [Google Scholar]
  47. Magdalena C, Dominguez F, Loidi L, Puente J (2000) Tumour prothymosin alpha content, a potential prognostic marker for primary breast cancer. Br J Cancer 82(3):584. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Malicet C, Giroux V, Vasseur S, Dagorn JC, Neira JL, Iovanna JL (2006) Regulation of apoptosis by the p8/prothymosin a complex. Proc Natl Acad Sci U S A 103(8):2671–2676 [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Naranmandura H, Ibata K, Suzuki KT (2007) Toxicity of dimethylmonothioarsinic acid toward human epidermoid carcinoma A431 cells. Chem Res Toxicol 20(8): 1120–1125 [DOI] [PubMed] [Google Scholar]
  50. Naujokas MF, Anderson B, Ahsan H, et al. (2013) The broad scope of health effects from chronic arsenic exposure: update on a worldwide public health problem. Environ Health Perspect 121(3):295. [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Olano-Martin E, Mountzouris KC, Gibson GR, Rastall RA (2000) In vitro fermentability of dextran, oligodextran and maltodextrin by human gut bacteria. Br J Nutr 83(3):247–255 [DOI] [PubMed] [Google Scholar]
  52. Cyclins Pines J CDKs and cancer In: Semin Cancer Biol, 1995. vol 6 Elsevier, p 63–72 [DOI] [PubMed] [Google Scholar]
  53. Pinyayev TS, Kohan MJ, Herbin-Davis K, Creed JT, Thomas DJ (2011) Preabsorptive metabolism of sodium arsenate by anaerobic microbiota of mouse cecum forms a variety of methylated and thiolated arsenicals. Chem Res Toxicol 24(4):475–477 [DOI] [PubMed] [Google Scholar]
  54. Pu Y-S, Yang S-M, Huang Y-K, et al. (2007) Urinary arsenic profile affects the risk of urothelial carcinoma even at low arsenic exposure. Toxicol Appl Pharmacol 218(2):99–106 [DOI] [PubMed] [Google Scholar]
  55. Qiao Y, Zhang Y, Wang J (2013) Ubiquitin E3 ligase SCF β-TRCP regulates TRIB2 stability in liver cancer cells. Biochem Biophys Res Commun 441(3):555–559 [DOI] [PubMed] [Google Scholar]
  56. Rajilić-Stojanović M, Maathuis A, Heilig HG, Venema K, de Vos WM, Smidt H (2010) Evaluating the microbial diversity of an in vitro model of the human large intestine by phylogenetic microarray analysis. Microbiology 156(11):3270–3281 [DOI] [PubMed] [Google Scholar]
  57. Rowland I, Davies M (1981) In vitro metabolism of inorganic arsenic by the gastro - intestinal microflora of the rat. J Appl Toxicol 1(5):278–283 [DOI] [PubMed] [Google Scholar]
  58. Ruby MV, Davis A, Schoof R, Eberle S, Sellstone CM (1996) Estimation of lead and arsenic bioavailability using a physiologically based extraction test. Environ Sci Technol 30(2):422–430 [Google Scholar]
  59. Sanz E, Muñoz-Olivas R, Camara C (2005) A rapid and novel alternative to conventional sample treatment for arsenic speciation in rice using enzymatic ultrasonic probe. Anal Chim Acta 535(1):227–235 [Google Scholar]
  60. Scott N, Hatlelid KM, MacKenzie NE, Carter DE (1993) Reactions of arsenic (III) and arsenic (V) species with glutathione. Chem Res Toxicol 6(1): 102–106 [DOI] [PubMed] [Google Scholar]
  61. Shi MM, Kugelman A, Iwamoto T, Tian L, Forman HJ (1994) Quinone-induced oxidative stress elevates glutathione and induces gamma-glutamylcysteine synthetase activity in rat lung epithelial L2 cells. J Biol Chem 269(42):26512–26517 [PubMed] [Google Scholar]
  62. Shin HD, Park BL, Cheong HS, Yoon J-H, Kim YJ, Lee H-S (2007) SPP1 polymorphisms associated with HBV clearance and HCC occurrence. Int J Epidemiol 36(5): 1001–1008 [DOI] [PubMed] [Google Scholar]
  63. Sicinski P, Donaher JL, Geng Y, Parker SB (1996) Cyclin D2 is a FSH-responsive gene involved in gonadal cell proliferation and oncogenesis. Nature 384(6608):470. [DOI] [PubMed] [Google Scholar]
  64. Song Y, Pan G, Chen L, et al. (2015) Loss of ATOH8 increases stem cell features of hepatocellular carcinoma cells. Gastroenterol 149(4):1068–1081. e5 [DOI] [PubMed] [Google Scholar]
  65. Spiegelstein O, Lu X, Le XC, et al. (2005) Effects of dietary folate intake and folate binding protein-2 (Folbp2) on urinary speciation of sodium arsenate in mice. Environ Toxicol Pharmacol 19(1): 1–7 [DOI] [PubMed] [Google Scholar]
  66. Steinmaus C, Yuan Y, Kalman D, et al. (2010) Individual differences in arsenic metabolism and lung cancer in a case-control study in Cordoba, Argentina. Toxicol Appl Pharmacol 247(2): 138–145 [DOI] [PMC free article] [PubMed] [Google Scholar]
  67. Sun G, Zhang J, Hu W (2014) CCNG2 expression is downregulated in colorectal carcinoma and its clinical significance. Tumor Biology 35(4):3339–3346 [DOI] [PubMed] [Google Scholar]
  68. Theriot CM, Bowman AA, Young VB (2016) Antibiotic-induced alterations of the gut microbiota alter secondary bile acid production and allow for Clostridium difficile spore germination and outgrowth in the large intestine. MSphere 1(1):e00045–15 [DOI] [PMC free article] [PubMed] [Google Scholar]
  69. Thomas DJ (2007) Molecular processes in cellular arsenic metabolism. Toxicol Appl Pharmacol 222(3):365–373 [DOI] [PubMed] [Google Scholar]
  70. Trapnell C, Williams BA, Pertea G, et al. (2010) Transcript assembly and quantification by RNA-Seq reveals unannotated transcripts and isoform switching during cell differentiation. Nat Biotechnol 28(5):511–515 [DOI] [PMC free article] [PubMed] [Google Scholar]
  71. Vahter M, Concha G (2001) Role of metabolism in arsenic toxicity. Basic & Clinical Pharmacology & Toxicology 89(1): 1–5 [DOI] [PubMed] [Google Scholar]
  72. Van de Wiele T, Gallawa CM, Kubachka KM, et al. (2010) Arsenic metabolism by human gut microbiota upon in vitro digestion of contaminated soils. Environ Health Perspect 118:1004–1009 [DOI] [PMC free article] [PubMed] [Google Scholar]
  73. Waalkes MP, Liu J, Diwan BA (2007) Transplacental arsenic carcinogenesis in mice. Toxicol Appl Pharmacol 222(3):271–280 [DOI] [PMC free article] [PubMed] [Google Scholar]
  74. Waalkes MP, Ward JM, Diwan BA (2004) Induction of tumors of the liver, lung, ovary and adrenal in adult mice after brief maternal gestational exposure to inorganic arsenic: promotional effects of postnatal phorbol ester exposure on hepatic and pulmonary, but not dermal cancers. Carcinogenesis 25(1): 133–141 [DOI] [PubMed] [Google Scholar]
  75. Wang J, Park J-S, Wei Y, et al. (2013) TRIB2 acts downstream of Wnt/TCF in liver cancer cells to regulate YAP and C/EBPα function. Mol Cell 51 (2):211–225 [DOI] [PMC free article] [PubMed] [Google Scholar]
  76. Wang J, Vasaikar S, Shi Z, Greer M, Zhang B (2017) WebGestalt 2017: a more comprehensive, powerful, flexible and interactive gene set enrichment analysis toolkit. Nucleic Acids Res [DOI] [PMC free article] [PubMed] [Google Scholar]
  77. Wang W-Q, Liu L, Xu H-X, et al. (2014) The combination of HTATIP2 expression and microvessel density predicts converse survival of hepatocellular carcinoma with or without sorafenib. Oncotarget 5(11):3895. [DOI] [PMC free article] [PubMed] [Google Scholar]
  78. Wei B, Yu J, Wang J, et al. (2017) The relationships between arsenic methylation and both skin lesions and hypertension caused by chronic exposure to arsenic in drinking water. Environ Toxicol Pharmacol 53:89–94 [DOI] [PubMed] [Google Scholar]
  79. Wibowo AS, Singh M, Reeder KM, et al. (2013) Structures of human folate receptors reveal biological trafficking states and diversity in folate and antifolate recognition. Proc Natl Acad Sci USA 110(38): 15180–15188 [DOI] [PMC free article] [PubMed] [Google Scholar]
  80. Wu C, Habib N, Mitry R, Reitsma P, Van Deventer S, Chamuleau R (1997) Overexpression of hepatic prothymosin alpha, a novel marker for human hepatocellular carcinoma. Br J Cancer 76(9): 1199. [DOI] [PMC free article] [PubMed] [Google Scholar]
  81. Yu H, Liu S, Li M, Wu B (2016a) Influence of diet, vitamin, tea, trace elements and exogenous antioxidants on arsenic metabolism and toxicity. Environ Geochem Health 38(2):339–351 [DOI] [PubMed] [Google Scholar]
  82. Yu H, Wu B, Zhang X-X, et al. (2016b) Arsenic metabolism and toxicity influenced by ferric iron in simulated gastrointestinal tract and the roles of gut microbiota. Environ Sci Techno1 50(13):7189–7197 [DOI] [PubMed] [Google Scholar]
  83. Zhang B, Kirov S, Snoddy J (2005) WebGestalt: an integrated system for exploring gene sets in various biological contexts. Nucleic Acids Res 33(suppl_2):W741–W748 [DOI] [PMC free article] [PubMed] [Google Scholar]
  84. Zhang H, Wang F, Hu Y (2017) STARD13 promotes hepatocellular carcinoma apoptosis by acting as a ceRNA for Fas. Biotechnol Lett 39(2):207–217 [DOI] [PubMed] [Google Scholar]
  85. Zhao F, Yu J (2016) Unearthing a novel tumor suppressor function of ATOH8 in hepatocellular carcinoma: role in acquisition of cancer stem cell-like features. Translational Cancer Research 5(1):S91–S94 [Google Scholar]

RESOURCES