Sage Journals: Discover world-class research

Abstract

Methotrexate (MTX)-induced intestinal mucosal injury in animals has been studied to understand how MTX can cause gastrointestinal disorders, but the pathogenesis of gastrointestinal disorders is still uncertain. We have attempted to reveal how dietary factors influence intestinal toxicity due to MTX. Mice were fed normal chow (NC) or a high-fat high-sucrose diet (HFHSD) before oral administration of MTX. While MTX significantly decreased the survival rates of mice fed HFHSD, the intestinal epithelial injury was detected. MTX excretion in the feces of mice fed HFHSD was reduced. Change of diets between NC and HFHSD influences the survival. The survival rates of the mice fed a high-sucrose diet or control diet were higher than those fed HFHSD. Higher survival rates were observed in mice fed a high-fat high-sucrose diet modified (HFHSD-M) in which casein was replaced by soybean-derived proteins. The survival rates of mice treated with vancomycin were lower than those administered neomycin. Microbiome and metabolome analyses on feces suggest a similarity of the intestinal environments of mice fed NC and HFHSD-M. HFHSD may modify MTX-induced toxicity in intestinal epithelia on account of an altered MTX distribution as a result of change in the intestinal environment.

Keywords

Methotrexate adverse effect intestinal mucosal injury microbiome metabolome

Introduction

Methotrexate (MTX) is a folate analog and one of the most frequently used disease-modifying antirheumatic drugs. MTX inhibits dihydrofolate reductase, prohibits nucleic acid synthesis, suppresses cell proliferation, and produces antirheumatic effects. MTX is used as an anticancer agent but is also prescribed for rheumatoid arthritis, psoriatic arthritis, polymyositis/dermatomyositis, adult-onset Still’s disease, and polymyalgia rheumatica.^1,2 MTX medication is sometimes discontinued because of side effects, such as bone marrow suppression, liver dysfunction, interstitial pneumonia, and gastrointestinal disorder.² Folinic acid is delivered to ameliorate critical side effects caused by the antifolate potency of MTX.³ It is suspected that dietary factors or obesity may modulate the distribution or side effects of MTX, but this remains undecided.^1,4

–7

MTX-induced intestinal mucosal injury in animals has been studied to understand the gastrointestinal disorder, the most common side effect of MTX. The influence of intestinal environment on pharmacokinetics and MTX-induced intestinal toxicity has been suggested. Elemental liquid diets alter the metabolism of MTX and increase the intestinal toxicity due to MTX. Concentrations of MTX in sera and bile are also increased. Replacement of amino acids with polypeptides decreases toxicity, but the addition of lipids or dietary fiber is ineffective.^8
–10 Another study has reported that dietary soybean proteins improve mucosal injury, while dietary fiber is only slightly effective.^11,12 It has been shown that dietary menhaden oil ameliorates MTX-induced intestinal mucosal injury.¹³ A methionine–choline-deficient diet also decreases the intestinal toxicity of MTX.¹⁴ On the other hand, antibiotic treatment aggravates MTX-induced intestinal mucosal injury and urinary MTX excretion is decreased.¹⁵ However, the relationship between the intestinal environment and the MTX toxicity remains uncertain. It has been determined that intestinal microbiota and metabolites influence the efficacy of various drugs.¹⁶ It has been suggested that diets change intestinal microbiota and as a result modulate the pharmacokinetics of MTX, leading to intestinal injury.¹⁷ In the present study, we tried to understand how dietary factors impact on intestinal toxicity due to MTX via change of the intestinal environment.

Materials and methods

Reagents

MTX, folinic acid calcium salt, sodium bicarbonate, carboxymethyl cellulose (CMC), naringenin, apigenin, neomycin sulfate, corn oil, nicotinic acid, l-glutamic acid, thiamin hydrochloride, β-alanine, and l-malic acid were purchased from Fujifilm Wako Pure Chemical Corp. (Osaka, Japan), and N-acetyl-l-glutamic acid and l-homoserine were purchased from Tokyo Chemical Industry Co., Ltd (Tokyo, Japan). ω-3 Fatty acid (Lotriga, 2 g of Lotriga containing 930 mg docosahexaenoic acid (DHA) and 750 mg eicosapentaenoic acid (EPA)) was purchased from Takeda Pharmaceutical Co., Ltd (Tokyo, Japan). EPA (Epadel) was bought from Mochida Pharmaceutical Co., Ltd (Tokyo, Japan). Vancomycin hydrochloride was purchased from Pfizer Inc. (New York City, New York, USA). Nnormal saline solution was acquired from Otsuka Pharmaceutical Co., Ltd (Tokyo, Japan).

Animals

C57BL/6J male mice (7–15 weeks old) were housed in a controlled environment in the Laboratory Animal Resource Center, University of Tsukuba, Japan, on a daily 14-h light/10-h dark cycle at 23.5 ± 2.5°C and 52.5 ± 12.5% relative humidity with free access to food and drinking water. All animal experiments were performed in accordance with the principles and guidelines for international animal care and under the protocols approved by the Institutional Animal Care and Use Committees of the University of Tsukuba.

Diets and MTX administration

Normal chow (NC, Certified Diet MF) and high-fat high-sucrose diet (HFHSD, F2HFHSD¹⁸) were obtained from Oriental Yeast Co. Ltd (Tokyo, Japan). High-sucrose diet (HSD), high-fat diet (HFD), and control diet (CD) were generated from HFHSD by replacing fat and/or sucrose by starch (Oriental Yeast, Supplementary Table S1). High-fat high-sucrose diet modified (HFHSD-M) was also made from HFHSD by replacing casein with soybean-derived proteins (Oriental Yeast, Supplementary Table S1).

Mice were fed ad libitum the aforementioned diets for 2 weeks prior to MTX administration by gavage. MTX was dissolved in 7% sodium bicarbonate and administered orally in a single dose of 3 mg/kg every day; bone marrow suppression is not observed at this concentration.¹⁹ Survival rates and body weight ratios were noted. Daily caloric intake per mouse was calculated from the food consumption values observed once a week. In common bile duct ligation (CBDL) experiment, CBDL or sham operation was performed on mice in the CBDL or sham operation group 1 day before administration of MTX and plasma MTX levels were measured on day 4 (24 h after MTX administration). In the food exchange experiments, mice were fed HFHSD 2 weeks prior to MTX administration and fed NC 1 day before administration, or fed NC 2 weeks prior to MTX administration and HFHSD 1 day before administration. In fasting experiments, mice fasted for 3 days after the start of MTX administration. Mice were treated with 1500 or 6700 mg/kg of ω-3 fatty acid, EPA, or corn oil 2 weeks before MTX administration. In the antibiotics treatment experiment, mice in the neomycin or vancomycin group were also treated with 1 g/L neomycin or 0.5 g/L vancomycin in drinking water 2 weeks before MTX administration. In the polyphenol treatment experiment, mice in the naringenin or apigenin group were treated with 80 mg/kg naringenin or apigenin 2 weeks before MTX administration. Mice were fed HFHSD 1 day before administration of MTX, as well as 500 mg/kg of candidate metabolites, including N-acetyl-l-glutamic acid, nicotinic acid, l-glutamic acid, thiamin hydrochloride, β-alanine, l-malic acid, or l-homoserine. Folinic acid was dissolved in normal saline solution and administered intraperitoneally in a single dose of 2.6 mg/kg every day.

MTX disposition

Blood samples were collected in BD Microtainer EDTA-2K (Becton, Dickinson and Co., Franklin Lakes, New Jersey, USA) on day 7 and cell blood counts were analyzed using Celltac alpha (Nihon Kohden, Tokyo, Japan). Plasma samples were prepared by the centrifugation of the heparinized blood at 800 × g for 20 min and collected at 1, 3, 6, and 12 h after MTX administration on day 4. Urine and fecal samples were collected 24 h after MTX administration on day 3 and bile was collected from the gallbladder on day 4 after 24 h fasting from day 3. MTX levels were analyzed using Nanopia eTDM MTX (Sekisui Medical, Tokyo, Japan) and Varioskan LUX multimode microplate reader (Thermo Fisher Scientific, Waltham, Massachusetts, USA).

Histological examination

After the 7-day treatment of MTX, intestinal organs from mice were fixed in 10% formalin in 0.01 M phosphate buffer (pH 7.2), embedded in paraffin, and cut into sections. The sections were stained with hematoxylin and eosin and histological score was evaluated, as previously described.^20
–22

Microbiome analyses

Fecal samples were collected from mice on day 3 of MTX administration and immediately frozen. DNA was extracted from the fecal samples with PureLink Microbiome DNA Purification Kit (Thermo Fisher Scientific). Amplicon library of V3–V4 region of 16S ribosomal ribonucleic acid (rRNA) gene was prepared, according to the manufacturer’s instructions with slight modifications (http://jp.support.illumina.com/content/dam/illumina-support/documents/documentation/chemistry_documentation/16s/16s-metagenomic-library-prep-guide-15044223-b.pdf). The V3–V4 region of 16S rRNA gene was amplified by two-step polymerase chain reaction (PCR) with NEB Next Q5 Hot Start HiFi PCR Master Mix (New England Biolabs Inc., Ipswich, Massachusetts, USA). Thermal cycling conditions of amplicon PCR consisted of initial denaturation at 98°C for 30 s, followed by 20 cycles at 98°C for 10 s, at 55°C for 15 s, and at 72°C for 30 s, followed by an extension at 72°C for 5 min. Thermal cycling conditions of index PCR with the index primers of Nextera Index Kit (Illumina, San Diego, California, USA) consisted of denaturation at 98°C for 30 s, followed by six cycles at 98°C for 10 s, at 55°C for 15 s, and at 72°C for 30 s, followed by an extension at 72°C for 5 min. The 16S rRNA gene amplicon sequencing was performed using MiSeq (Illumina).

Paired-end reads were assembled into artificial longer reads with the fastq-join (ver. 1.1.2-537).²³ Primer sequences were trimmed from the assembled reads with cutadapt (version 1.13).²⁴ Reads without primer sequences at both ends and those with average quality value >25 were discarded. After quality filtering processes, 20,000 reads per sample were randomly chosen and used for further analysis. The randomly sampled reads were combined into one file and clustered into operational taxonomic units by a cutoff value of 97% identity using the UPARSE pipeline at USEARCH (version 9.2.64_i86linux32).²⁵ The representative sequence of each operational taxonomic unit was assigned to our custom 16S rRNA sequences database with the BLAST+ (ver. 2.3.0+).²⁶ Shannon indices were calculated with the R (version 3.2.2) vegan package (version 2.3-5, https://cran.r-project.org/web/packages/vegan/). Unweighted UniFrac distances among samples were calculated with the PyCogent library (ver. 1.5.3)²⁷ of Python 2.7.5.

Metabolome analyses

Fecal samples were collected from mice on the day of MTX administration and pooled in each group to analyze the metabolomic change in feces by a diet which may cause a different reaction to MTX. Ionic and nonionic metabolites were extracted from pooled fecal samples with water and methanol, respectively. Metabolome analysis was performed at Human Metabolome Technologies (Tsuruoka, Japan) using capillary electrophoresis time-of-flight mass spectrometry for ionic metabolites and liquid chromatography time-of-flight mass spectrometry for nonionic metabolites based on the methods described previously.^28
–30

MetaboAnalyst 4.0 (http://www.metaboanalyst.ca/MetaboAnalyst/faces/home.xhtml) was used for principal component analysis (PCA) and pathway analysis.³¹ All the zero values were replaced with half of the minimum positive values in the original data and data scaling was performed by mean centering, dividing by the standard deviation (SD) of each variable. The impact of the average mouse survival periods of each group on the metabolic pathway was evaluated with Mus musculus pathway library, global test, and relative-betweenness centrality; p values from the enrichment analysis and pathway impact values from the pathway topology analysis were also obtained. Hierarchical cluster analysis was performed with IBM SPSS Statistics 25 (International Business Machines Corporation, Armonk, New York, USA).

Statistical analysis

The experimental results were represented as mean ± SD and analyzed by Mann–Whitney’s U test, log-rank test, or one-way analysis of variance using IBM SPSS Statistics 25. The value of p < 0.05 was considered statistically significant.

Results

The effects of HFHSD on MTX disposition and resulting toxicity

MTX was administered to mice fed NC or HFHSD. Diarrhea and tarry stool were observed in mice fed HFHSD after the administration of MTX. The survival rates of mice fed HFHSD decreased after the administration of MTX (Figure 1(d), average survival period ± SD (day), 7.00 ± 0.82 (MTX + HFHSD), 14.00 ± 0.00 (MTX + NC), p = 2.23 × 10⁻⁴). The body weight ratio and the caloric intake of mice fed HFHSD also decreased (Figure 1(h) and (l)). Although 3 mg/kg MTX is not toxic for mice fed NC (Figure 1(c)), it proved toxic for mice fed HFHSD (Figure 1(d)). The results of a histological examination of the intestine are presented in Supplementary Figure S1. MTX administration caused severe mucosal injury in the small intestine, especially in the jejunum, of mice fed HFHSD. The intestinal villi were destroyed in these mice. Histological scores in the jejunum showed the most serious mucosal injury in mice fed HFHSD (average ± SD, 0.50 ± 0.00 (MTX + NC), 4.90 ± 0.22 (MTX + HFHSD), p = 0.004). Thus, severe mucosal injury in the jejunum is estimated to be the cause of death in mice fed HFHSD.

Figure 1.

MTX treatment decreases survival rate, body weight ratio, and caloric intake of mice fed HFHSD. (a) to (d) The survival rate in each mouse group. The difference was tested by the log-rank test (MTX + HFHSD vs. MTX + NC). (e)–(h) Body weight ratio in each mouse group. The difference was tested by one-way ANOVA; *p < 0.05 and **p < 0.01, respectively (mean ± SD, MTX + HFHSD vs. MTX + NC). (i) to (l) Caloric intake in each mouse group (vehicle + NC: n = 5, vehicle + HFHSD: n = 7, MTX + NC: n = 5, and MTX + HFHSD: n = 7). The results are representative of two independent experiments. MTX: methotrexate; NC: normal chow; HFHSD: high-fat high-sucrose diet; ANOVA: analysis of variance; SD: standard deviation.

Blood cell count was observed 1 week after MTX administration (Supplementary Table S2). The number of white blood cells decreased in mice fed HFHSD. MTX concentration in plasma and bile was also investigated (Figure 2). Plasma MTX concentration in mice fed HFHSD was slightly higher than those fed NC 12 h after administration (average ± SD (μmol/L), 0.009 ± 0.012 (MTX + NC), 0.124 ± 0.135 (MTX + HFHSD), p = 0.028). The concentration of MTX in bile of mice fed HFHSD was also higher (average ± SD (μmol/L), 6.349 ± 3.013 (MTX + NC), 30.099 ± 42.759 (MTX + HFHSD), p = 0.041). The amount of MTX excreted in the feces of mice treated with HFHSD was very much lower (Figure 2(c), average ± SD (nmol), 95.242 ± 35.592 (MTX + NC), 15.050 ± 8.466 (MTX + HFHSD), p = 0.004), though the amount of MTX excreted in urine was comparable (average ± SD (nmol), 27.583 ± 15.598 (MTX + NC), 14.667 ± 25.083 (MTX + HFHSD), p = 0.055). Thus, fecal MTX excretion was reduced in mice fed HFHSD.

Figure 2.

MTX disposition. (a) MTX concentration in plasma from 1 h to 12 h following MTX administration on day 4. The difference was tested by one-way ANOVA; *p < 0.05 (mean ± SD, MTX + NC: n = 6, MTX + HFHSD: n = 6). (b) MTX concentration in bile on day 4. The difference was tested by Mann–Whitney’s U test (mean ± SD, MTX + NC: n = 12, MTX + HFHSD: n = 10). MTX excretion in (c) feces and (d) urine for 24 h on day 3 (mean ± SD, MTX + NC: n = 6, MTX + HFHSD: n = 6). MTX: methotrexate; NC: normal chow; HFHSD: high-fat high-sucrose diet; ANOVA: analysis of variance; SD: standard deviation.

After CBDL or sham operation, the plasma MTX concentration was measured (Supplementary Figure S2). Plasma MTX concentration in mice with CBDL tended to be higher but did not reach statistical significance. The amount of MTX excreted in feces or urine was comparable. The histological scores showed serious mucosal injury in mice that had undergone CBDL (average ± SD, 2.50 ± 0.50 (MTX + HFHSD + CBDL), 0.75 ± 0.65 (MTX + HFHSD + sham operation), p = 0.008).

The effects of folinic acid on MTX toxicity

Folinic acid and MTX were simultaneously administered to mice fed HFHSD (Supplementary Figure S3A and S3E). The survival rates of mice treated with folinic acid and MTX were higher than for mice treated with MTX alone (average survival period (day) ± SD, 14.60 ± 3.51 (MTX + HFHSD), 19.00 ± 0.00 (MTX + HFHSD + folinic acid), p = 0.002). The body weight ratio of mice treated with folinic acid and MTX did not decrease. Thus, folinic acid relieved intestinal mucosal injury caused by MTX.

Effects of diet change on MTX toxicity

Some mice were fed HFHSD 2 weeks prior to MTX administration and fed NC 1 day before administration (Supplementary Figure S3B and S3F). The survival rate of mice-fed diets changed from HFHSD to NC was higher (average survival period (day) ± SD, 9.71 ± 0.95 (MTX + HFHSD), 25.00 ± 0.00 (MTX + HFHSD → NC), p = 2.13 × 10⁻⁴). Some mice were fed NC 2 weeks prior to MTX administration and fed HFHSD 1 day before administration (Supplementary Figure S3C and S3G). The survival rates of mice fed diets changed from NC to HFHSD were lower (25.00 ± 0.00 (MTX + NC), 12.29 ± 2.69 (MTX + NC → HFHSD), p = 1.40 × 10⁻⁴). In fasting experiments, the survival rates of mice that underwent fasting were higher, compared with mice without fasting (6.43 ± 0.98 (MTX + HFHSD), 9.14 ± 0.38 (MTX + HFHSD [Fasting]), p = 1.61 × 10⁻⁴). Thus, change of diets influences intestinal toxicity due to MTX.

Effect on MTX toxicity of dietary ingredients

MTX was administered to mice fed HSD, HFD, and CD (Figure 3(a) to (c)). The survival rates of mice fed HSD (average survival period (day) ± SD, 13.40 ± 4.88, p = 0.024) or CD (12.00 ± 2.34, p = 0.027) were higher than for those fed HFHSD (9.55 ± 1.51), though they were comparable to mice fed HFD (8.60 ± 1.67, p = 0.614). No difference was observed in the body weight ratio of these mice (Figure 3(e) to (g)). Additionally, MTX was administered to mice fed HFHSD-M (Figure 3(d)). The survival rates of mice fed HFHSD-M were higher (9.29 ± 1.38 (MTX + HFHSD), 15.29 ± 5.35 (MTX + HFHSD-M), p = 0.002).

Figure 3.

The effect of diet ingredients. (a) to (d) The survival rate in each mouse group. The difference was tested by log-rank test. (e)–(h) Body weight ratio in each mouse group. Difference was tested by one-way ANOVA (mean ± SD, MTX + HFHSD [(a), (b), (c), (e), (f), and (g)]: n = 11, MTX + HFD: n = 5, MTX + HSD: n = 5, MTX + CD: n = 5, MTX + HFHSD [(d) and (h)]: n = 7, and MTX + HFHSD-M: n = 7). The results are representative of two independent experiments. MTX: methotrexate; HFHSD: high-fat high-sucrose diet; HFD: high-fat diet; HSD: high-sucrose diet; CD: control diet; HFHSD-M: high-fat high-sucrose diet modified; ANOVA: analysis of variance; SD: standard deviation.

Mice fed HFHSD were treated with 6700 mg/kg of ω-3 fatty acid, EPA, or corn oil, before administration of MTX (Figure 4(a) and (b)). The survival rates of mice treated with ω-3 fatty acid (14.00 ± 5.69, p = 0.001) or EPA (10.86 ± 4.98, p = 0.006) were higher than those treated with corn oil (6.00 ± 1.15). Mice fed HFHSD-M were also treated with 1500 mg/kg of ω-3 fatty acid, before administration of MTX (Figure 4(c)), but no significant difference was observed (13.71 ± 5.55 (MTX + HFHSD-M + ω-3 fatty acid), 12.71 ± 4.11 (MTX + HFHSD-M), p = 0.824). Mice fed HFHSD were treated with polyphenols, naringenin (8.29 ± 4.15 (MTX + HFHSD + naringenin), 7.08 ± 2.02 (MTX + HFHSD), p = 0.481) or apigenin (7.00 ± 1.29, p = 0.723), before administration of MTX (Supplementary Figure S4A, S4B, S4E, and S4F), but no significant difference was observed with these postbiotics known to have an effect on energy expenditure.³² Thus, dietary lipid and protein affected intestinal toxicity due to MTX.

Figure 4.

The effect of lipids. (a) to (c) The survival rate in each mouse group. The difference was tested by the log-rank test. (d) to (f) Body weight ratio in each mouse group. The difference was tested by one-way ANOVA and asterisks; *p < 0.05 and **p < 0.01, respectively (mean ± SD, MTX + HFHSD + ω-3 fatty acid: n = 7, MTX + HFHSD + EPA: n = 7, MTX + HFHSD + corn oil: n = 7, MTX + HFHSD-M + ω-3 fatty acid: n = 7, and MTX + HFHSD-M: n = 7). The results are representative of two independent experiments. MTX: methotrexate; HFHSD: high-fat high-sucrose diet; HFHSD-M: high-fat high-sucrose diet modified; EPA: eicosapentaenoic acid; ANOVA: analysis of variance; SD: standard deviation.

The effects of antibiotics on MTX toxicity

Mice fed NC were treated with vancomycin or neomycin before administration of MTX (Supplementary Figure S4C and S4G). The survival rates of mice treated with vancomycin were lower than those treated with neomycin (average survival period (day) ± SD, 14.40 ± 3.58 (MTX + NC + vancomycin), 17.00 ± 0.00 (MTX + NC + neomycin), p = 0.003). Mice fed HFHSD were also treated with vancomycin before administration of MTX (Supplementary Figure S4D and S4H). The survival rates of mice treated with vancomycin (8.00 ± 1.00 (MTX + HFHSD + vancomycin), p = 0.007) were lower than for untreated mice (9.85 ± 1.07 (MTX + HFHSD)). Thus, the administration of vancomycin influenced MTX toxicity, but neomycin did not have an effect.

The effect of diet on gut microbiota and metabolome

Feces were collected from mice fed various diets and gut microbiota was analyzed (Figure 5). Results from principal coordinate analysis using unweighted UniFrac successfully discriminated between samples from mice fed NC (G1) or HFHSD (G3, Figure 5(a)). The samples of mice fed HFHSD-M (G5) or changed diets (G9 and G10) were also located between the previous two samples. Fecal samples of mice treated with vancomycin (G2 and G11) were differentiated from that of untreated mice. Hierarchical clustering analysis of microbiota suggested that the characteristics of the microbiota in the feces of mice fed NC (G1) were similar to those of mice fed HFHSD-M (G5, Figure 5(b)). Shannon indices decreased in feces of mice fed HFHSD (G3) or treated with vancomycin (G2). Shannon indices of feces of mice fed HFHSD-M were relatively higher (G5, Figure 5(c)). The bacterial composition of mice fecal samples at the family level was very different in mice treated with vancomycin (G2, Figure 5(d)). Thus, the gut microbiome of mice fed HFHSD-M is closely related to that of mice fed NC, though vancomycin treatment totally alters the composition.

Figure 5.

The effect of diets on gut microbiota. (a) Scores plots for PCoA using unweighted UniFrac distance matrix of microbiota of mice fecal samples. (b) Hierarchical clustering analysis of the microbiota of mice fecal samples. (c) Shannon indices of microbiota of mice fecal samples. (d) The bacterial composition of mice fecal samples at the family level. G1: mice fed NC, G2: mice fed NC and treated with 0.5g/L vancomycin, G3: mice fed HFHSD, G4: mice fed HFHSD and treated with ω-3 fatty acid (1500 mg/kg), G5: mice fed HFHSD-M, G6: mice fed HFHSD and treated with EPA (1500 mg/kg), G7: mice fed HFHSD and treated with naringenin (80 mg/kg), G8: mice fed HFHSD and treated with apigenin (80 mg/kg), G9: mice fed HFHSD for 13 days and fed NC 24 h before starting MTX, G10: mice fed NC for 13 days and fed with HFHSD 24 h before starting MTX, and G11: mice fed HFHSD and treated with 0.5g/L vancomycin. MTX: methotrexate; NC: normal chow; HFHSD: high-fat high-sucrose diet; HFHSD-M: high-fat high-sucrose diet modified; EPA: eicosapentaenoic acid; PCoA: principal coordinate analysis.

Metabolomic analysis was also performed on the pooled fecal samples of mice fed various diets. Results from PCA also discriminated between samples of mice fed NC or HFHSD (Figure 6(a)) and the sample of mice fed HFHSD-M was located between these two groups. The samples from mice treated with vancomycin were differentiated from other samples. Hierarchical clustering analysis of fecal metabolites suggested that the characteristics of metabolites in feces of mice fed NC are similar to those in mice fed HFHSD-M (Figure 6(b)). The impact of dietary factors on metabolic pathways was also analyzed (Figure 6(c)), based on the average relative survival period of each group (NC: 2.78, NC vancomycin: 1.55, HFHSD: 1, HFHSD ω3 fatty acid: 1.38, HFHSD-M: 1.72, HFHSD EPA: 0.93, HFHSD naringenin: 1.17, and HFHSD apigenin: 0.99). The pathways of purine metabolism, β-alanine metabolism, thiamine metabolism, biotin metabolism, d-glutamine, and d-glutamate metabolism were overrepresented in the pathway analysis. It is suggested that diets influence vitamin metabolism, amino acid turnover, and nucleic acid metabolism. Candidate metabolites were selected based on the average relative survival period of each group and the effects of these candidate metabolites on MTX toxicity were evaluated (Supplementary Table S3, Supplementary Figure S5, average survival period (day) ± SD, 12.0 ± 3.94, p = 0.250 (MTX + HFHSD + N-acetyl-l-glutamic acid), 8.40 ± 0.55, p = 0.630 (MTX + HFHSD + nicotinic acid), 9.40 ± 1.14, p = 0.604 (MTX + HFHSD + l-glutamic acid), 9.40 ± 0.89, p = 0.680 (MTX + HFHSD + thiamin hydrochloride), 12.20 ± 2.77, p = 0.128 (MTX + HFHSD + β-alanine), 9.80 ± 1.64, p = 0.930 (MTX + HFHSD + l-malic acid), 11.60 ± 0.89, p = 0.054 (MTX + HFHSD + l-homoserine), and 9.40 ± 1.34 (MTX + HFHSD + CMC)). However, no metabolite significantly increased the survival rates of mice.

Figure 6.

The effects of diets on the gut metabolites. (a) Scores plots for PCA to comparing mice groups. (b) Hierarchical clustering analysis of metabolites in the feces of each mice group. (c) Metabolomic pathway analysis based on the metabolites in feces of mice groups. NC: normal chow; HFHSD: high-fat high-sucrose diet; HFHSD-M: high-fat high-sucrose diet modified; EPA: eicosapentaenoic acid; PCA: principal component analysis.

Discussion

The administration of MTX is frequently stopped because of side effects.² It is thought that diets or obesity influences the distribution and side effects of MTX.^1,4

–7 However, this remains undetermined. Effects of diets or antibiotics on the severity of MTX-induced intestinal mucosal injury have been reported. These results suggest the possibility of the impact of intestinal environment on the pharmacokinetics and pharmacodynamics of MTX, but these mechanisms are not clearly understood. The present study has tried to clarify these mechanisms.

Feeding HFHSD increased the severity of mucosal injury in the small intestine, especially in the jejunum, and it was relieved by the administration of folinic acid (Figure 1, Supplementary Figures S1 and S3), suggesting folate-antagonistic action of MTX to damage intestinal epithelial cells. It is known that orally administered MTX is mainly absorbed from the proton-coupled folate transporter on jejunal epithelial cells and that MTX is excreted by breast cancer resistance protein (BCRP) or multidrug resistance-associated protein 2 (MRP2) on the luminal membrane of jejunal epithelial cells.^33,34 It is speculated that feeding HFHSD changes the intestinal environment and increases the amount of some intestinal metabolites that may inhibit the function of BCRP or MRP2, followed by decreased fecal MTX excretion. Accumulated MTX in jejunal epithelial cells triggers epithelial cell injury and malabsorption.

The methionine–choline-deficient diet causes nonalcoholic steatohepatitis and enhances the hepatocellular toxicity of MTX, but reduces intestinal injury due to MTX.¹⁴ Fecal MTX excretion has been reduced by methionine–choline-deficient diet, similar to the effects of HFHSD in the present study, though the effect of diets on the MTX intestinal toxicity was found to be completely opposite. CBDL was predicted to reduce excretion of MTX from bile and was expected to reduce the severity of mucosal injury. However, MTX-induced intestinal mucosal injury was aggravated in mice with CBDL, compared with mice that underwent the sham operation (Supplementary Figure S2). It has been reported that hyperbilirubinemia reduces MTX excretion by MRP2 on the jejunal epithelial cell,³⁴ suggesting that hyperbilirubinemia caused by CBDL exacerbates MTX-induced intestinal mucosal injury.

The survival rates of mice fed diets changed from HFHSD to NC were higher (Supplementary Figure S3). The survival rates of mice fed diets changed from NC to HFHSD were lower. The survival rates of mice that had undergone fasting were higher. In the gut microbiome analysis, the samples of mice fed changed diets (G9 and G10, Figure 5(a)) were located between those fed NC (G1) and HFHSD (G3). It has been reported that gut microbiota rapidly change after the change of diet,^32,35 suggesting that the rapid modification in the intestinal environment caused by the change of diet alters the intestinal toxicity caused by MTX.

Feeding elemental liquid diets increases MTX intestinal toxicity, and the replacement of amino acids with polypeptides decreases the toxicity, but the addition of lipid or dietary fiber is ineffective.^8,9 The quality and quantity of protein modulate the intestinal toxicity due to MTX.^11,12,36 Dietary menhaden oil or DHA improves MTX-induced intestinal mucosal injury.^13,37 Menhaden oil contains ω-3 fatty acids, such as DHA or EPA, and corn oil contains ω-6 and ω-9 fatty acids. In the present study, mice fed HFHSD-M could survive longer and the lipid component in HFHSD was revealed to contribute to MTX intestinal toxicity (Figure 3). It was also found that ω-3 fatty acid or EPA increased survival rates, compared with corn oil (Figure 4). Thus, the change of intestinal environment caused by the quality of different proteins may increase the intestinal toxicity due to MTX and the effect of quality and quantity of different lipids are also implied.

It has been found that treatment with neomycin increased the lethality of MTX-induced intestinal mucosal injury.¹⁵ Urinary MTX excretion is decreased and fecal MTX excretion is increased. It has also been shown that intestinal sterilization by oral administration of vancomycin and imipenem increases the toxicity of MTX.³⁸ In the present study, the treatment with vancomycin reduced the survival rates, but the treatment with neomycin did not (Supplementary Figure S4). The different doses of neomycin may explain the discrepancy between these studies. Since treatment with vancomycin decreased the survival rate of mice fed HFHSD, the mechanisms of exacerbation due to HFHSD diet and vancomycin may be different. Vancomycin is one of the glycopeptide antibiotics and blocks the cell wall construction of Gram-positive bacteria. Neomycin is an aminoglycoside antibiotic and blocks protein translation of Gram-negative bacteria. Vancomycin and neomycin are hardly absorbed in the intestine. The bacteria in the antimicrobial spectrum of vancomycin and not in the spectrum of neomycin may contribute to the relief of MTX-induced intestinal mucosal injury.

The microbiome analysis showed a similarity of profile in mice fed NC and HFHSD-M (Figure 5). Metabolome analysis also suggested similarity in the profile of mice fed NC and HFHSD-M and the involvement of vitamin metabolism, amino acid turnover, and nucleic acid metabolism in the pathogenesis of MTX-induced intestinal mucosal injury (Figure 6). The gut commensal microbes may metabolize MTX^39
–41 and MTX changes the gut microbiota.^42,43 It seems rational that immunological responses are involved in the pathogenesis of intestinal injury induced by MTX,^38,44 because dietary factors and gut microbiota influence the development of immunological cells in the gut.^45
–47 It has been reported that some bacteria^42,48,49 or metabolites including amino acids, polyphenols, or vitamins^{21,50

–54} influence MTX-induced intestinal mucosal injury. However, each strain or metabolite administered alone produces only a limited range of effects on the intestinal toxicity of MTX (Supplementary Figure S5). The combinational effects of candidate bacteria and metabolites may be investigated by bacterial transplantation and metabolite administration experiments.

The results of the present study suggest that protein or lipid in HFHSD alters the intestinal environment, influencing the pathogenesis of MTX-induced intestinal mucosal injury. It has been revealed that the gut microbiome and its metabolites modulate the efficacy of drugs.¹⁶ The change of gut microbiota influences the pharmacokinetics and pharmacodynamics of MTX via its metabolites. If the combinatorial therapy with candidate bacteria and metabolites for reducing the toxicity of MTX is established, side effects of MTX may become preventable and a new therapeutic strategy for rheumatoid arthritis can be developed.

Supplemental material

Supplemental Material, MTX_Fig#6-S - Modulation of methotrexate-induced intestinal mucosal injury by dietary factors

Supplemental Material, MTX_Fig#6-S for Modulation of methotrexate-induced intestinal mucosal injury by dietary factors by T Higuchi, M Yoshimura, S Oka, K Tanaka, T Naito, S Yuhara, E Warabi, S Mizuno, M Ono, S Takahashi, S Tohma, N Tsuchiya and H Furukawa in Human & Experimental Toxicology

Supplemental material

Supplemental Material, MTX_Table#6-S - Modulation of methotrexate-induced intestinal mucosal injury by dietary factors

Supplemental Material, MTX_Table#6-S for Modulation of methotrexate-induced intestinal mucosal injury by dietary factors by T Higuchi, M Yoshimura, S Oka, K Tanaka, T Naito, S Yuhara, E Warabi, S Mizuno, M Ono, S Takahashi, S Tohma, N Tsuchiya and H Furukawa in Human & Experimental Toxicology

Footnotes

Acknowledgments

We thank Mr Keita Yamashita (Department of Clinical Laboratory, Tsukuba Medical Center Hospital) for technical instruction and Ms Mayumi Yokoyama and Ms Satomi Hanawa (Clinical Research Center for Allergy and Rheumatology, National Hospital Organization Sagamihara National Hospital) for secretarial assistance.

Author contributions

TH, MY, EW, and HF conceived and designed the experiments. TH, MY, SO, KT, TN, and HF performed the experiments. TH, MY, SY, and HF analyzed the data. TH, MY, MO, and HF examined the pathological findings. TH, MY, SM, STa, SToh, NT, and HF contributed reagents/materials/analysis tools. TH, MY, and HF contributed to the writing of the manuscript. All authors read and approved the final manuscript. TH and MY contributed equally to this work.

Declaration of conflicting interests

The author(s) declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: HF has the following conflicts, and the following funders are supported wholly or in part by the indicated pharmaceutical companies. The Takeda Science Foundation is supported by an endowment from Takeda Pharmaceutical Company. HF was supported by research grants from Bristol-Myers Squibb Co. HF received honoraria from Ajinomoto Co., Inc., Daiichi Sankyo Co., Ltd, Dainippon Sumitomo Pharma Co., Ltd, Pfizer Japan Inc., and Takeda Pharmaceutical Company, Luminex Japan Corporation Ltd, and Ayumi Pharmaceutical Corporation. KT, TN, and SY are employees of the Miraca Research Institute. The other authors declare no financial or commercial conflict of interest.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by Grants-in-Aid for Scientific Research (B, C) (Grant Numbers 26293123, 18K08402, and 18K10985) from the Japan Society for the Promotion of Science, Research Grants from Takeda Science Foundation, and Bristol-Myers K.K. RA Clinical Investigation Grant from Bristol-Myers Squibb Co. The funders had no role in study design, data collection and analysis, decision to publish, or preparing the manuscript.

ORCID iD

H Furukawa

Supplemental material

Supplemental material for this article is available online.

References

Visser

van der Heijde

. Risk and management of liver toxicity during methotrexate treatment in rheumatoid and psoriatic arthritis: a systematic review of the literature. Clin Exp Rheumatol 2009; 27: 1017–1025.

Kinder

Hassell

Brand

, et al. The treatment of inflammatory arthritis with methotrexate in clinical practice: treatment duration and incidence of adverse drug reactions. Rheumatology (Oxford) 2005; 44: 61–66.

Goldman

Matherly

. Biochemical factors in the selectivity of leucovorin rescue: selective inhibition of leucovorin reactivation of dihydrofolate reductase and leucovorin utilization in purine and pyrimidine biosynthesis by methotrexate and dihydrofolate polyglutamates. NCI Monogr 1987; 5: 17–26.

Pinkerton

Welshman

Glasgow

, et al.

Can food influence the absorption of methotrexate in children with acute lymphoblastic leukaemia?

Lancet 1980; 2: 944–946.

Madanat

Awidi

Shaheen

, et al. Effects of food and gender on the pharmacokinetics of methotrexate in children. Res Commun Chem Pathol Pharmacol 1987; 55: 279–282.

Kozloski

De Vito

Kisicki

, et al. The effect of food on the absorption of methotrexate sodium tablets in healthy volunteers. Arthritis Rheum 1992; 35: 761–764.

Suzuki

Uehara

Tajima

, et al. Elevation of serum hepatic aminotransferases during treatment of rheumatoid arthritis with low-dose methotrexate. Risk factors and response to folic acid. Scand J Rheumatol 1999; 28: 273–281.

McAnena

Rossi

Mehta

, et al. Alteration of methotrexate metabolism in rats by administration of an elemental liquid diet. I. Changes in drug enterohepatic circulation. Cancer 1987; 59: 31–37.

McAnena

Harvey

Bonau

, et al. Alteration of methotrexate toxicity in rats by manipulation of dietary components. Gastroenterology 1987; 92: 354–360.

10.

Mitsugi

Nakamura

Kashiwabara

, et al. Protection against methotrexate toxicity by a soybean protein- and omega-3 fatty acid-containing diet: comparative study with a casein-containing diet. Oncol Rep 2004; 12: 41–45.

11.

Funk

Baker

. Effect of fiber, protein source and time of feeding on methotrexate toxicity in rats. J Nutr 1991; 121: 1673–1683.

12.

Funk

Baker

. Effect of soy products on methotrexate toxicity in rats. J Nutr 1991; 121: 1684–1692.

13.

Vanderhoof

Blackwood

Mohammadpour

, et al. Effect of dietary menhaden oil on normal growth and development and on ameliorating mucosal injury in rats. Am J Clin Nutr 1991; 54: 346–350.

14.

Hardwick

Clarke

Lake

, et al. Increased susceptibility to methotrexate-induced toxicity in nonalcoholic steatohepatitis. Toxicol Sci 2014; 142: 45–55.

15.

Zaharko

Bruckner

Oliverio

. Antibiotics alter methotrexate metabolism and excretion. Science 1969; 166: 887–888.

16.

Alexander

Wilson

Teare

, et al. Gut microbiota modulation of chemotherapy efficacy and toxicity. Nat Rev Gastroenterol Hepatol 2017; 14: 356–365.

17.

Kehoe

Harvey

Daly

. Alteration of chemotherapy toxicity using a chemically defined liquid diet in rats. Cancer Res 1986; 46: 4047–4052.

18.

Maeda

Shimomura

Kishida

, et al. Diet-induced insulin resistance in mice lacking adiponectin/ACRP30. Nat Med 2002; 8: 731–737.

19.

Ohbayashi

Suzuki

Yashiro

, et al. Induction of pulmonary fibrosis by methotrexate treatment in mice lung in vivo and in vitro. J Toxicol Sci 2010; 35: 653–661.

20.

Chiu

McArdle

Brown

, et al. Intestinal mucosal lesion in low-flow states. I. A morphological, hemodynamic, and metabolic reappraisal. Arch Surg 1970; 101: 478–483.

21.

Kolli

Kanakasabapathy

Faith

, et al. A preclinical study on the protective effect of melatonin against methotrexate-induced small intestinal damage: effect mediated by attenuation of nitrosative stress, protein tyrosine nitration, and PARP activation. Cancer Chemother Pharmacol 2013; 71: 1209–1218.

22.

Kolli

Natarajan

Isaac

, et al. Mitochondrial dysfunction and respiratory chain defects in a rodent model of methotrexate-induced enteritis. Hum Exp Toxicol 2014; 33: 1051–1065.

23.

Aronesty

. Comparison of sequencing utility programs. Open Bioinforma J 2013; 7: 1–8.

24.

Martin

. Cutadapt removes adapter sequences from high-throughput sequencing reads. EMBnet J 2011; 17: 10–12.

25.

Edgar

. UPARSE: highly accurate OTU sequences from microbial amplicon reads. Nat Methods 2013; 10: 996–998.

26.

Camacho

Coulouris

Avagyan

, et al. BLAST+: architecture and applications. BMC Bioinformatics 2009; 10: 421.

27.

Knight

Maxwell

Birmingham

, et al. PyCogent: a toolkit for making sense from sequence. Genome Biol 2007; 8: R171.

28.

Ohashi

Hirayama

Ishikawa

, et al. Depiction of metabolome changes in histidine-starved Escherichia coli by CE-TOFMS. Mol Biosyst 2008; 4: 135–147.

29.

Ooga

Sato

Nagashima

, et al. Metabolomic anatomy of an animal model revealing homeostatic imbalances in dyslipidaemia. Mol Biosyst 2011; 7: 1217–1223.

30.

Sugimoto

Wong

Hirayama

, et al. Capillary electrophoresis mass spectrometry-based saliva metabolomics identified oral, breast and pancreatic cancer-specific profiles. Metabolomics 2010; 6: 78–95.

31.

Chong

Soufan

, et al. MetaboAnalyst 4.0: towards more transparent and integrative metabolomics analysis. Nucleic Acids Res 2018; 46: W486–W494.

32.

Thaiss

Itav

Rothschild

, et al. Persistent microbiome alterations modulate the rate of post-dieting weight regain. Nature 2016; 540: 544–551.

33.

Qiu

Jansen

Sakaris

, et al. Identification of an intestinal folate transporter and the molecular basis for hereditary folate malabsorption. Cell 2006; 127: 917–928.

34.

Yokooji

. Role of ABC efflux transporters in the oral bioavailability and drug-induced intestinal toxicity. Yakugaku Zasshi 2013; 133: 815–822.

35.

Harvey

McAnena

Mehta

, et al. Reversibility of elemental liquid diet-enhanced methotrexate toxicity by refeeding with chow. J Parenter Enteral Nutr 1987; 11: 119–123.

36.

Dunki Jacobs

Ruevekamp

Varossiau

, et al. Alterations in serum levels, anti-tumor activity and toxicity of methotrexate in rats after a short period of nutritional depletion. Eur J Cancer Clin Oncol 1989; 25: 415–422.

37.

Horie

Nakamaru

Masubuchi

. Docosahexaenoic acid exhibits a potent protection of small intestine from methotrexate-induced damage in mice. Life Sci 1998; 62: 1333–1338.

38.

Frank

Hennenberg

Eyking

, et al. TLR signaling modulates side effects of anticancer therapy in the small intestine. J Immunol 2015; 194: 1983–1995.

39.

Goldman

Levy

. The enzymatic hydrolysis of folate analogues. Biochem Pharmacol 1968; 17: 2265–2270.

40.

Shane

Stokstad

. Transport and metabolism of folates by bacteria. J Biol Chem 1975; 250: 2243–2253.

41.

Nayak

O’Loughlin

Fischbach

, et al. Methotrexate is an antibacterial drug metabolized by human gut bacteria. Arthritis Rheum 2016; 68: 1536.

42.

Zhou

Xia

Wang

, et al. Induction and amelioration of methotrexate-induced gastrointestinal toxicity are related to immune response and gut microbiota. EBioMedicine 2018; 33: 122–133.

43.

Fijlstra

Ferdous

Koning

, et al. Substantial decreases in the number and diversity of microbiota during chemotherapy-induced gastrointestinal mucositis in a rat model. Support Care Cancer 2015; 23: 1513–1522.

44.

de Koning

van Dieren

Lindenbergh-Kortleve

, et al. Contributions of mucosal immune cells to methotrexate-induced mucositis. Int Immunol 2006; 18: 941–949.

45.

Atarashi

Tanoue

Shima

, et al. Induction of colonic regulatory T cells by indigenous Clostridium species. Science 2011; 331: 337–341.

46.

Furusawa

Obata

Fukuda

, et al. Commensal microbe-derived butyrate induces the differentiation of colonic regulatory T cells. Nature 2013; 504: 446–450.

47.

Kinoshita

Kayama

Kusu

, et al. Dietary folic acid promotes survival of Foxp3+ regulatory T cells in the colon. J Immunol 2012; 189: 2869–2878.

48.

Mao

Nobaek

Kasravi

, et al. The effects of Lactobacillus strains and oat fiber on methotrexate-induced enterocolitis in rats. Gastroenterology 1996; 111: 334–344.

49.

Duman

Kumral

Ercan

, et al. Saccharomyces boulardii ameliorates clarithromycin- and methotrexate-induced intestinal and hepatic injury in rats. Br J Nutr 2013; 110: 493–499.

50.

Shou

Lieberman

Hofmann

, et al. Dietary manipulation of methotrexate-induced enterocolitis. J Parenter Enteral Nutr 1991; 15: 307–312.

51.

Warden

Noltorp

Francis

, et al. Vitamin A deficiency exacerbates methotrexate-induced jejunal injury in rats. J Nutr 1997; 127: 770–776.

52.

Sukhotnik

Mogilner

Karry

, et al. Effect of oral glutamine on enterocyte turnover during methotrexate-induced mucositis in rats. Digestion 2009; 79: 5–13.

53.

Sukhotnik

Pollak

Coran

, et al. Glutamine attenuates the inhibitory effect of methotrexate on TLR signaling during intestinal chemotherapy-induced mucositis in a rat. Nutr Metab (Lond) 2014; 11: 17.

54.

Cetiner

Sener

Sehirli

, et al. Taurine protects against methotrexate-induced toxicity and inhibits leukocyte death. Toxicol Appl Pharmacol 2005; 209: 39–50.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

1.25 MB

0.16 MB