Biomedicine & Pharmacotherapy 114 (2019) 108849
Contents lists available at ScienceDirect
Biomedicine & Pharmacotherapy journal homepage: www.elsevier.com/locate/biopha
Antibiotics exacerbated colitis by affecting the microbiota, Treg cells and SCFAs in IL10-deficient mice
T
Bo Shena, Jiajia Hua,c, Huan Songa, Zhengting Wanga, Jiangao Fanb, Yunwei Suna, , ⁎ Qijun Wanga,c, ⁎
a
Department of Gastroenterology of Ruijin Hospital, Shanghai Institute of Immunology, Shanghai Jiao Tong University School of Medicine, Shanghai, 200025, China Department of Gastroenterology, Xinhua Hospital Affiliated to Shanghai Jiao Tong University School of Medicine, Shanghai Key Lab of Pediatric Gastroenterology and Nutrition, Shanghai, 200092, China c Department of Pharmacology, Yale University School of Medicine, CT, 06511, USA b
ARTICLE INFO
ABSTRACT
Keywords: Colitis Antibiotics SCFAs Microbiota Treg
Many studies have shown that antibiotic therapy can attenuate colitis in IL10-deficient (IL10−/−) mice. However, these results have indicated that antibiotics were more successful in preventing, rather than treating established colitis. Those antibiotic treatments attempted to only partially alter the intestinal microbiota and to not eliminate it completely. Therefore, we treated IL10−/− mice with the multiantibiotic regimen that was used to develop a pseudo-germ-free mouse model to determine whether multi-antibiotics attenuated or exacerbated colitis. We evaluated the colitis in IL10−/− mice receiving the antibiotic treatment versus those receiving the water control; furthermore, we investigated the gut microbiota, the intestinal immune cell proportions and the cytokine secretion. Surprisingly, the IL10-/- mice receiving the antibiotic treatment had more severe intestinal colitis and a swollen cecum than those receiving the water control. Moreover, the abundance of microbiota and content of short-chain fatty acids (SCFAs) in the colon were dramatically decreased. Additionally, the proportions of Treg cells and Th1 cells in the colons of IL10-/- mice were also decreased. The mechanism may be that the decrease in the microbiota leads to a decrease in the proportions of Treg cells and SCFAs, which are necessary to maintain intestinal homeostasis. All changes lead to further exacerbated colitis in IL10−/− mice with antibiotic treatment.
1. Introduction A disruption in the microbiota has been associated with the development of colitis, especially for IL10-deficient mice (IL10−/−) [1]. The histological characteristics of colitis in IL10-/- mice is similar to that of human IBD [2,3]. As we all know, IL10-/- mice can develop spontaneous colitis under SPF or a conventional environment [2]. However, germfree IL10-/- mice do not develop spontaneous colitis [4], which shows that there is a close relationship between colitis and the intestinal microbiota. Furthermore, the possibility that antibiotic therapy attenuates colitis in IL10−/− mice has attracted much attention. In a study by K.L. Madsen et al. [5]. IL10−/− mice were treated with ciprofloxacin or with neomycin and metronidazole before IL10−/− mice developed colitis, and the treatment of antibiotics obviously prevented the development of colitis. For the treatment of IL10−/− mice with 8–12 weeks of established colitis, ciprofloxacin treatment resulted in the greatest
⁎
decrease in adherent or translocated bacteria; neomycin-metronidazole treatment resulted in a greater degree of healing than the degree of healing by the established colitis treatment with neomycin and metronidazole. We investigated whether it is true that the greater the decrease in the microbiota, the better the effects of antibiotic treatment in IL10−/− mice. In a study by F. Hoentjen et al. [6], antibiotics were also used to treat germ-free IL10−/− mice with the colonization of SPF bacteria. This study showed that antibiotics are a good therapy to attenuate colitis in IL10−/− mice. H.C. Rath et al. also demonstrated a similar result [7]. However, these results all indicated that antibiotics were better for prevention rather than for treatment, especially for established colitis [5–7]. For example, ciprofloxacin can play a preventive role in colitis, but for established colitis, the effect is minor [5]; metronidazole can play a preventive role in colitis induced by DSS but also has no effect on established colitis [8]. As proposed by H.C. Rath, some bacterial species may initiate inflammation while others, perhaps including a larger
Corresponding authors at: Shanghai Institute of Immunology, Shanghai Jiao Tong University School of Medicine, Shanghai, 200025, China. E-mail addresses:
[email protected] (Y. Sun),
[email protected] (Q. Wang).
https://doi.org/10.1016/j.biopha.2019.108849 Received 22 January 2019; Received in revised form 1 April 2019; Accepted 2 April 2019 0753-3322/ © 2019 The Authors. Published by Elsevier Masson SAS. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/BY-NC-ND/4.0/).
Biomedicine & Pharmacotherapy 114 (2019) 108849
B. Shen, et al.
spectrum of intestinal bacteria, perpetuate disease [7]. Therefore, we treated IL10−/− mice with the multiantibiotic regimen that was used to develop a pseudo-germ-free mouse model in an attempt to remove the gut microbiota as much as possible. With this method, multiple broad-spectrum antibiotics can decrease the intestinal microbiota and can even cause an almost germ-free state. We wanted to know whether multiantibiotic treatment further attenuated or exacerbated colitis in IL10−/− mice. At the same time, we also analyzed the intestinal microbiota, the immune cells and SCFAs to determine the reason for this phenomenon.
the SILVA-based bacterial reference sequence (version 128) according to the Standard Operating Procedure of the Mothur software for MiSeq data. All reads that passed the filtering procedure were classified into various taxonomic levels from the phylum to the genus by comparison with the SILVA database and the Bayesian classifier (with the confidence cutoff set at 80%). 2.5. Enzyme-linked immunosorbent assay (ELISA) According to the protocols of the ELISA kits (Invitrogen™, USA), an ELISA plate (Corning Coster 9018, USA) was coated with the target antibody and was incubated overnight at 4 °C. The levels of inflammatory cytokines, such as IFN-γ, IL-1β and IL-17, in colon samples were analyzed by ELISA Ready-SET-Go kits (eBioscience, San Diego, CA, USA) according to the manufacturer’s instructions. ELISA plates were analyzed at 450 nm.
2. Materials and methods 2.1. Animal models Male and female C3Bir.129P2(B6)-IL10tm1Cgn/Lt (C3H IL10−/−) mice were kind gifts from Prof. Zhinan Yin (Jinan University, Guangzhou, China). All mice were maintained under specific, pathogen-free conditions (8–12 weeks of age; 5 mice five housed per cage) in the Animal Resource Center at Shanghai Jiao Tong University School of Medicine. The animal experimentation protocols were approved by the Institutional Animal Care and Use Committee of Shanghai Jiao Tong University, School of Medicine.
2.6. Real-time PCR analyses of inflammatory factors in colonic samples The total RNA was extracted using TRIzol Reagent (Thermo Fisher Scientific, USA), and the cDNA was synthesized using a PrimeScript RT reagent Kit (Takara Biotechnology, Dalian, China). Real-time PCR was carried out using Hieff qPCR SYBR Green Master Mix (Yeasen, Shanghai, China) with a real-time PCR instrument (ViiA™ Real-Time PCR Instruments, Thermo Fisher Scientific, USA). The results were analyzed by the 2−ΔΔCT method. All primers were synthesized by Sunny (Shanghai, China) and had the following sequences: IL-1β: Forward primer: GAAATGCCACCTTTTGACAGTG Reverse primer: TGGATGCTCTCATCAGGACAG IL-17: Forward primer: TTTAACTCCCTTGGCGCAAAA Reverse primer: CTTTCCCTCCGCATTGACAC IFN-γ: Forward primer: ATGAACGCTACACACTGCATC Reverse primer: CCATCCTTTTGCCAGTTCCTC
2.2. Genotype identification At 4 weeks of age, the mouse tail tips were cut off by scissors. The length of the cut tip was less than 5 mm long. The samples were placed into 400 μl tissue digestive fluid with 10 μl proteinase K and were incubated in a 55 °C water bath for 3 h. The digestion was terminated by incubation at 95 °C for 5 min. Each PCR mixture contained 1 μl extracted DNA, 0.5 μl of 10 μM mutant primer (CCACACGCGTCACCTTA ATA), 0.5 μl of 10 μM wt primer (GTTATTGTCTTCCCGGCTGT), 1.0 μl of 10 μM common primer (CTTGCACTACCAAAGCCACA), 10 μl of 2x PCR buffer (I-5™ 2x Hi-Fi PCR Master Mix; Molecular Cloning Laboratories, South San Francisco, CA) and 7 μl of Milli-Q water. The PCR conditions were initially denatured at 95 °C for 5 min, followed by 35 cycles of: initial denaturation at 94 °C for 30 s, annealing at 64 °C for 1 min, and extension at 72 °C for 45 s. This was followed by a final extension at 72 °C for 3 min. The PCR products were separated with gel electrophoresis in a 1.5% DNA agarose gel. The genotypes and product sizes were as follows: mutant mouse = 312 bp, heterozygote mouse = 137 bp and 312 bp, and wild-type mouse = 137 bp.
2.7. Paraffin embedded section preparation and hematoxylin and eosin (H &E) staining The colon was cleaned with PBS and was fixed for more than 24 h with 10% buffered formalin. After fixation, the sample was dehydrated and embedded in paraffin. Then, the colon was sectioned (5 μm thick) at different cross-sections. Finally, the sample was stained with hematoxylin and eosin and was examined by microscopy. The sections were graded with a range from 0 to 3 for the amount of inflammation, and the depth of inflammation was graded with a range from 0 to 4 [[9]].
2.3. Antibiotic administration Adult male mice were administered the following: ampicillin (BBI life science, Shanghai, China) at 1.0 g/L, novobiocin sodium salt (BBI life science, Shanghai, China) at 1.0 g/L, metronidazole (Sigma-Aldrich, USA) at 1.0 g/L and vancomycin hydrochloride (BBI life science, Shanghai, China) at 0.5 g/L. These were administered in drinking water that was filtered through a 0.22 μm filter and was replaced every 2 days from the beginning to the end of the experiments.
2.8. Flow cytometry The colons that were harvested from the sacrificed mice were enzymatically digested for 55 min to a single-cell suspension, were washed in 1% bovine serum albumin and were stained for various surface markers: CD3-FITC (eBiosciences, San Diego, CA, USA), CD4-BV421 (BioLegend San Diego, CA, USA), CD8-APC-CY7 (eBiosciences, San Diego, CA), L/D-Fixable Viability Stain 700 (BD Biosciences, San Jose, CA) were used to stain for T cells; and CD45-AF700 (BD Biosciences, San Jose, CA), CD11b-FITC (BD Biosciences, San Jose, CA), Ly6GBV711 (BioLegend San Diego, CA, USA), Ly6C-BV785 (BioLegend San Diego, CA, USA), F4/80-PE-CY7 (eBiosciences, San Diego, CA, USA), CD11c-APC (eBiosciences, San Diego, CA, USA), MCHII-percpcy5.5 (BD Biosciences, San Jose, CA), SiglecF-PE (BD Biosciences, San Jose, CA), CD64-BV421 (BD Biosciences, San Jose, CA), and L/D- Fixable Viability Stain 780 (BD Biosciences, San Jose, CA) were used to stain for macrophages (1 × 106 cells with the manufacturer-suggested concentration of fluorochrome-conjugated antibody). Subsequently, they were fixed
2.4. 16S rRNA gene analysis of bacteria A QIAamp PowerFecal DNA Kit (QIAGEN, Germany) was used to isolate the DNA from stool samples. 16S rRNA sequencing was performed on an Illumina® MiSeq™II platform by fully sequencing the V1V3 16S rRNA gene regions, which were amplified using barcoded primers that were optimized for Serial Illumina Sequencing. The V1V3 hypervariable regions of the 16S rRNA gene were amplified using a two-step PCR strategy. Illumina HiSeq 2000 sequencing was performed at Shanghai Jiao Tong University, School of Medicine. NGS reads for each sample were decoded by using an in-house Java script, which is available upon request. These reads were then filtered and aligned to 2
Biomedicine & Pharmacotherapy 114 (2019) 108849
B. Shen, et al.
3.2. Pro-inflammatory cytokines increased in IL10−/− mice receiving antibiotic treatment
and permeabilized with BD Cytofix/Cytoperm (BD Biosciences, San Jose, CA) and were stained with the following intracellular antibodies: Foxp3-PE (eBiosciences, San Diego, CA, USA), IFN-γ-APC (eBiosciences, San Diego, CA, USA), and IL13-PE-CY7 (Invitrogen, Grand Island, NY, USA). The samples were analyzed with a BD LSRFortessa X-20 cell analyzer.
Due to the exacerbated colitis from antibiotic treatment in IL10−/− mice, we determined that there were pro-inflammatory cytokines in the colon. We isolated the colons from both the IL10−/− mice receiving the antibiotic treatment and from those receiving the water control. We then measured the inflammatory cytokine levels by ELISAs. The significant changes in the cytokine levels could be observed in IL10−/− mice receiving antibiotic treatment (Fig. 3A), indicating that much stronger inflammation may occur in IL10−/− mice receiving the antibiotic treatment. The mean level of IFN-γ was 68.3 ± 5.8 pg/50 mg for the water group and was 181.6 ± 33.5 pg/50 mg for the antibiotic group (P < 0.05). The mean level of IL-17 was 182.4 ± 5.6 pg/50 mg for the water group and was 506.9 ± 20.5 pg/50 mg for the antibiotic group (P < 0.001). The mean level of IL-1β was 51.0 ± 5.2 pg/50 mg for the water group and was 408.5 ± 40.0 pg/50 mg for the antibiotic group (P < 0.001). Consistent with the ELISA results, the higher mRNA expression levels of IFN-γ, IL-1β, IL-17 were also found in IL10−/ − mice receiving the antibiotic treatment (Fig. 3B).
2.9. Detection of the CH3COOH in the mouse feces An LCK365 kit (Hach Lange Gmbh, Germany) was used to detect CH3COOH in mouse fecal samples. Mouse feces (0.3 g) in 1 ml PBS were homogenized by vortexing and were then centrifuged at 10,000×g for 1 min. Following centrifugation, the supernatant was transferred to a new tube. After the addition of 25 μl LCK and 50 μl solution A (LCK 365 A), the tube was inverted a few times. The sample was heated at 100 °C. The highest standard (50 μl) was added to the appropriate tubes and was serially diluted by two-fold to construct a standard curve with six points (1,000, 500, 250, 125, 62.5, and 31.2 mg/l). Then, 50 μl of sample was added to the appropriate tubes, the tubes were inverted a few times, and the tubes were heated at 100 °C for 10 min. The tubes were allowed to cool to room temperature. Next, 50 μl solution B (LCK 365 B) was added to the tubes that were inverted a few times. Then, 50 μl solution C (LCK 365 C) was added to the same tubes. Finally, 250 μl solution D (LCK 365 D) was also added to the tubes. After 3 min, 100 μl of the solution from the tubes was added to a plate. The plate was analyzed at 497 nm.
3.3. The abundance of microbiota decreased in the IL10−/− mice receiving antibiotic treatment Considering the severity of the colitis in the IL10−/− mice receiving the antibiotic treatment, it was reasonable to suspect that the intestinal microbiota might be responsible for these effects. Therefore, we investigated the variety of the intestinal microbiota by 16S rRNA sequencing. The abundance of microbiota (Chao index of OTU level) in the colons of IL10−/− mice receiving the antibiotic treatment greatly decreased (Fig. 4A). The changes in diversity of microbiota (Shannon index of OTU level) were not big between the IL10−/− mice receiving the antibiotic treatment and those receiving the water control (Fig. 4B), implying that decrease of the microbiota abundance might be the main reason for the exacerbated colitis in the IL10−/− mice with multi-antibiotics administration. We also compared abundance and diversity of microbiota in wild type mice with water and antibiotics. The abundance of microbiota (Chao index of OTU level) in the colons of wild type mice receiving the antibiotic treatment also greatly decreased (Fig. 4G). The diversity of microbiota (Shannon index of OTU level) was also not big between the wild type mice receiving the antibiotic treatment and those receiving the water control (Fig. 4H).
2.10. Statistical analysis Statistical analysis was performed with IBM SPSS Statistics 19 software. A two-tailed Student’s t-test was performed as indicated. The results were considered statistically significant at *, P < 0.05; **, P < 0.01; ***, P < 0.001. 3. Results 3.1. Antibiotic treatment exacerbated colitis in IL10−/− mice We used a mixture of four antibiotics, i.e., 1.0 g/L ampicillin, 1.0 g/ L novobiocin, 1.0 g/L metronidazole, and 0.5 g/L vancomycin for the treatment of IL10−/− mice and used IL10−/− littermate mice treated with water as a control (Fig. 1A). The antibiotic mixture in water was administered for up to 19 days and was used to treat colitis. Surprisingly, the antibiotic treatment exacerbated colitis and caused a heavily swollen cecum. To examine the changes of colitis in IL10−/− mice, we dissected IL10-/- mice at day 15 after antibiotic treatment or the water control. Surprisingly, the colon length of the IL10-/- mice decreased with the treatment of antibiotics (Fig. 1B–C), and a significant difference in the colon length was observed between the IL10-/- mice treated with antibiotics and those treated with water (P < 0.01) (Fig. 1F). Meanwhile, a heavily swollen cecum was not observed in the IL10−/− mice treated with water (Fig. 1D), but was observed in the IL10−/− mice receiving the antibiotic treatment (Fig. 1E). We also examined the colon damage by a mouse endoscope, and observed severe inflammation and damage in the colon damage in IL10−/− mice by a mouse endoscope and observed severe inflammation and colon damage in IL10−/− mice, and these mice exhibited inflammatory vasculature, obvious signs of bleeding, and more ulcerates at day 15 during the antibiotic treatment (Fig. 2A). The Ulcerative Colitis Endoscopic Index was used to score the damage, and the mean score of 9.2 in the IL10−/− mice receiving antibiotic treatment was higher than the mean score of 4.2 in mice receiving water (P < 0.01) (Fig. 2B).H&E staining was also used to observe that the IL10-/- mice exhibited more severe epithelial damage and had higher levels of leukocyte infiltration with the antibiotic treatment (Fig. 2C–D). Taken together, these results indicate that the administration of antibiotics exacerbated colitis in IL10−/− mice.
3.4. The difference in the phyla and genera of microbiota between the IL10−/− mice treated with the water control and those treated with antibiotics The following phyla changes were observed in IL10−/− mice: Firmicutes, Bacteroidetes, unclassified_k_norank and Proteobacteria (Fig. 4C). In the IL10−/− mice treated with water, the phyla Firmicutes and Bacteroidetes dominated at 52.1% and 45.0% in the colon, respectively. In contrast, in IL10-/- mice receiving the antibiotic treatment, the phyla patterns were dominated by Proteobacteria at 96.7% (Fig. 4C). There were significant differences in the phyla of Proteobacteria, Firmicutes and Bacteroidetes. The phyla of Firmicutes and Bacteroidetes were significantly higher in IL10-/- mice receiving the water treatment than in mice receiving the antibiotic treatment (P < 0.05) (Fig. 4D). The phylum of Proteobacteria was significantly higher in IL10-/- mice receiving antibiotic treatment than in mice receiving the water treatment (P < 0.05) (Fig. 4D). There were microbiota genera differences between the IL10−/− mice receiving the antibiotic treatment and those receiving the water treatment. The IL10-/- mice receiving the antibiotic treatment had significantly higher levels of Pseudomonas than the levels of Pseudomonas in the IL10-/- mice receiving the water treatment (70.88% vs. 0.01%, 3
Biomedicine & Pharmacotherapy 114 (2019) 108849
B. Shen, et al.
Fig. 1. Phenotypes of IL10−/− mice treated with antibiotics and with water. A. The experimental timeline for mouse treatment with the administration of multiple antibiotics every two days. The multiple antibiotics included 1.0 g/L ampicillin, 1.0 g/L neomycin, 1.0 g/L metronidazole and 0.5 g/L vancomycin; the IL10−/− mice treated with water were used as a control. The experiments lasted a total of 19 days. B–C. Representative images of colons (IL10−/− mice with the antibiotic treatment and with the water treatment). D–E. Representative images of cecum (IL10−/− mice with the antibiotic and water treatments). F. The statistical analysis of colon length. The data are expressed as the mean ± SD. The statistical analyses were done with an unpaired t-test. **P < 0.01. G. The statistical analysis of cecal diameter. The data are expressed as the mean ± SD. The statistical analyses were done with an unpaired t-test. ***P < 0.001.
P < 0.05) (Fig. 4F). However, the Bacteroides levels were significantly lower in the antibiotic-treated IL10-/- mice than the levels in those treated with water (1.25% vs. 30.9%, P < 0.05) (Fig. 4F). There were significant differences between the mice treated with antibiotics and those treated with water for the ten following kinds of bacteria: unclassified_f_Lachnospiraceae (25.62% vs. 0.32%, P < 0.05), Lachnospiraceae_NK4A136_group (11.13% vs. 0.07%, P < 0.05), Dysgonomonas (8.70% vs. 0.19%, P < 0.05), unclassified_f_Ruminococcaceae (5.14% vs. 0.004%, P < 0.01), Turicibacter (2.56% vs. 0.002%, P < 0.01), unclassified_o_Clostridiales (2.07% vs. 0%, P < 0.05), Ruminiclostridium_6 (1.87% vs. 0%, P < 0.05), Parasutterella (1.71% vs. 0.02%, P < 0.01), norank_f_Lachnospiraceae (1.14% vs. 0.007%, P < 0.01), Ruminiclostridium (0.99% vs. 0%, P < 0.01) (Fig. 4F).
proportions of the different cell types, such as Treg cells, Th1 cells and macrophages, among lamina propria lymphocytes (LPLs) (Fig. 5A–E). We found that the proportions of Treg and Th1 cells decreased in IL10-/mice receiving the antibiotic treatment (Fig. 5A–C). The proportion of Treg cells was 39.6% in IL10-/- mice receiving the water treatment and was 33.9% in those receiving the antibiotic treatment (P < 0.01). At the same time, the proportion of Th1 cells was 10.2% in the watertreated IL10-/- mice and was 8.6% in the antibiotic-treated mice (P < 0.001). These results supported that the exacerbated colitis observed in the IL10-/- mice receiving the antibiotic treatment might be related to an imbalance in the types of immune cells. However, there was no significant difference in the percentage of macrophages in mice receiving the antibiotic treatment versus the percentage in those receiving the water treatment (Fig. 5E).
3.5. The proportions of Treg and Th1 cells decreased in antibiotic-treated IL10−/− mice
3.6. SCFAs levels in the colons of IL10−/− mice decreased with the treatment of antibiotics
Because of the severe effects from colitis in IL10−/− mice after the long-term antibiotic treatment, such as a shortened colon (Fig. 1C), swollen cecum (Fig. 1E), and inflamed vasculature (Fig. 2A–B), we explored the mechanisms underlying these phenomena in detail. We sacrificed the mice at day 15 of antibiotic treatment and analyzed the
Because of the changes in the microbiota composition (Fig. 4C–F), we investigated the metabolite changes in the intestines. Feces from the colons of wild-type and IL10−/− mice were isolated, and the total SCFA levels were assayed (Fig. 6). We found that the IL10-/- mice had a 4
Biomedicine & Pharmacotherapy 114 (2019) 108849
B. Shen, et al.
Fig. 2. Colonoscopy and H&E staining results of IL10−/− mice with the treatments of antibiotics and water. A. Representative colonoscopy results of the IL10−/− mice administered multiple antibiotics at day 15, and of the IL10−/− mice administered water as the control. B. The Ulcerative Colitis Endoscopic Index of Severity of the IL10−/− mice with the treatment of antibiotics and with water. The data are expressed as the mean ± SD. The statistical analyses were done with an unpaired t-test. **P < 0.01. C. Representative H&E histological sections of colons from the IL10-/- mice receiving the treatment of antibiotics and those receiving the water treatment. The original magnifications are 5x, 20x and 40 × . D. The inflammation score of the IL10−/− mice treated with antibiotics and of those treated with water. The data are expressed as the mean ± SD. The statistical analyses were done with an unpaired t-test. *P < 0.05.
dramatic change in the microbiota content in their intestines (Fig. 4C–F), which was accompanied by a sharp change in the colon acetate content with the treatment of antibiotics (P < 0.05) (Fig. 6). The mean acetate content of the colon was 1964.5 mg/L in the IL10-/mice treated with water. However, in the IL10-/- mice treated with antibiotics, the mean acetate content of the colon was 1142.8 mg/L.
Bacteroidetes dominated the gut microbial community, but the phyla of Proteobacteria is less abundant in the mammalian intestinal tract [12]. A sustained increase in the abundance of the phylum Proteobacteria caused an imbalance in the gut microbiota, where the natural human gut flora only have a minor proportion of this phylum [13]. In the natural gut flora, there is a balance of the gut microbiota with a high stability, and this flora interacts with the immune system. The immune system is capable of suppressing the uncontrolled expansion of Proteobacteria [13]. One study has shown that Proteobacteria were transiently dominant in the intestines of newborn mice, and this change directly led to the increase of proinflammatory factors [14]. Meanwhile, the role of Proteobacteria in gut inflammation has been confirmed in various animal models and has been shown to have a positive correlation [15–17]. In our study, the increased proportion of Proteobacteria in the gut reflected the unbalanced state of the gut microbial community in the antibiotic-treated IL10−/− mice, and this unbalanced state was an important factor that lead to the aggravation of colitis. Furthermore, significant differences were found in 12 different kinds of bacteria at the genus level after the antibiotic treatment, especially in the genera of Pseudomonas, Bacteroides and unclassified_f_Lachnosporaceae. Pseudomonas is an important opportunistic pathogen in the gut and has strong antibiotic resistance [18]. The proportion of Pseudomonas increased significantly after treatment with antibiotics, and this may play an important role in colitis exacerbation. However, Bacteroides was thought to be an important probiotic that benefits the host by preventing potential pathogens from colonizing the gut [19]. Meanwhile, unclassified_f_Lachnospiraceae can produce butyrate to protect the intestinal homeostasis [20]. The result suggested that the combination of antibiotics can remove a large number of harmful bacteria but can also remove a large number of intestinal probiotics, causing a disruption to the intestinal balance. Therefore, the contributions of the microbiota composition and abundance may be similar, or
4. Discussion Many studies have shown that antibiotic therapy may attenuate colitis in IL10−/− mice [5–7]. However, all these results have shown that antibiotics were more effective for prevention than for treatment, especially for the treatment of IL10−/− mice with established colitis [5–7]. Those antibiotic treatments only partially altered the intestinal microbiota and did eliminate it. Meanwhile, the resident microbiota play an important role in the initiation of colitis, and the disease outcome is dependent on the microbiota after antibiotic treatment [10]. Therefore, we hoped to treat the IL10−/− mice with a multi-antibiotic regimen, the same regimen that was used to develop the pseudo-germfree mouse model [11]. In this way, multiple broad-spectrum antibiotics decreased the intestinal microbiota as much as possible and even caused an almost germ-free state. Contrary to our expectations, the multi-antibiotic treatment exacerbated the colitis in the IL10−/− mice, rather than attenuating the colitis. The increased levels of inflammatory cytokines in the gut also proved that the colitis had been exacerbated. With the 16S rRNA sequencing of the microbiota in the antibiotic-treated and water-treated IL10−/− mice, we found that the microbiota abundance decreased significantly, but the diversity did not change significantly. With the antibiotic treatment, the phyla of Firmicutes and Bacteroidetes significantly decreased, and the phyla of Proteobacteria significantly increased. Studies have shown that the phyla of Firmicutes and 5
Biomedicine & Pharmacotherapy 114 (2019) 108849
B. Shen, et al.
Fig. 3. Cytokine expression in IL10−/− mouse colons with the treatments of antibiotics and water. A. The ELISA results of the inflammatory cytokines in the colon samples (including IFN-γ, IL-1β, IL-17); n = 5 mice/group. The data are expressed as the mean ± SD. The statistical analyses were done with an unpaired t-test. *P < 0.05; ***P < 0.001. B. The mRNA expression of inflammatory cytokines in the colon samples were analyzed with real-time PCR (including IFN-γ, IL-1β, IL-17); n = 5 mice/group. The data are expressed as the mean ± SD. The statistical analyses were done with an unpaired t-test. *P < 0.05; **P < 0.01.
even more important, than the contributions of pathogen colonization in developing colitis. This is consistent with the results in the study of Nabeetha A et al: the severity of disease was independent of antibioticinduced changes in the microbial community structure [10]. All the above results indicated the importance of the microbial community structure induced by antibiotics. Schulfer et al. also investigated the effect of antibiotic treatment in IL10−/− mice [21]. In comparison with the above study, we also found that 5 taxa had consistently decreased abundance in the IL10−/− mice treated with antibiotics, such as: Bacteroides, Lachnospiraceae, Ruminococcaceae, Turicibacter and Clostridiales. In the study by Schulfer et al., Bacteroides, Lachnospiraceae and Ruminococcaceae were significantly correlated with inflammation, but Turicibacter and Clostridiales were found to be important bacteria that protect the intestinal tract from progressive inflammation. In our study, there was a significant decrease in both the bacteria that protect the gut and those that exacerbate the
inflammation. These results suggest that the exacerbation of colitis induced by multiple antibiotics in the IL10−/− mice may not be closely related to specific microbiota, but this exacerbation may be related to the homeostasis of the whole intestinal flora. In the study by Schulfer et al., it was also noted that the use of antibiotics led to a decrease in the microbial diversity and abundance and that the antibiotic treatment of the intestinal flora further exacerbated the colitis. These conclusions are consistent with those of our study: the changes in the overall constitution of the intestinal flora caused by antibiotic treatment will destroy the intestinal homeostasis and will further lead to the occurrence of colitis. Studies have shown that the intestinal microbiota has a significant impact on the local and systemic immune system. Colonization by clearly defined and benign intestinal commensals can induce the activation and generation of Tregs in germ-free mice [22,23]. Even in the DSS model, it is necessary to induce efficient Tregs to maintain 6
Biomedicine & Pharmacotherapy 114 (2019) 108849
B. Shen, et al.
Fig. 4. The changes in the microbiota of IL10−/− and wild type mice with the treatments of antibiotics and water. A. The abundance of microbiota (Chao index of OTU) in mice receiving the treatment of antibiotics and in mice receiving the water treatment. The data are expressed as the mean ± SD. The statistical analyses were done with an unpaired t-test; *P < 0.05. B. The diversity of microbiota (Shannon index of OTU) in mice receiving the treatment of antibiotics and in mice receiving the water treatment. The data are expressed as the mean ± SD. NS, not statistically significant. C. The phylum levels of microbiota in the IL10−/− mice with the treatments of antibiotics and water. D. The genus level distribution of microbiota in the colon of IL10−/− mice. The data are expressed as the mean ± SD. The statistical analyses were done with the Wilcoxon rank-sum test bar plot. *P < 0.05. E. The genus levels of microbiota in the IL10−/− mice with the treatments of antibiotics and water. F. The genus level distribution of microbiota in the colons of IL10−/− mice. The data are expressed as the mean ± SD. The statistical analyses were done with the Wilcoxon rank-sum test bar plot; *P < 0.05. G. The abundance of microbiota (Chao index of OTU) in mice receiving the treatment of antibiotics and in wild type mice receiving the water treatment. The data are expressed as the mean ± SD. The statistical analyses were done with an unpaired t-test; **P < 0.01. H. The diversity of microbiota (Shannon index of OTU) in wild type mice receiving the treatment of antibiotics and in mice receiving the water treatment. The data are expressed as the mean ± SD. NS, not statistically significant.
intestinal homeostasis [24]. This was also demonstrated by the decreased proportion of Tregs in germ-free mice compared to that in SPF mice [23]. The results suggest that the global microbiota plays an important role in the regulation of Tregs. Meanwhile, additional murine studies have identified individual bacterial strains that have a very strong impact on the immune system, including Bacteroides fragilis, which can produce polysaccharide A (PSA) to mature the adaptive system and to induce Tregs [25]. A consortium of Clostridia species has also been reported to induce Treg cell maturation [22]. These studies are consistent with our results: the global intestinal microbiota content was largely cleared by multiantibiotics, which caused a decrease in the percentage of Tregs and exacerbated colitis. With an individual strain analysis, we found that the genus of Bacteroides declined significantly in our study. Our study demonstrated that antibiotic treatment must cause a decrease in the proportion of Tregs, both for the whole microbiota and for the individual strain. Interestingly, germ-free mice did not develop colitis in spite of the decreasing proportion of Tregs compared to the Treg proportions in SPF mice [4]. This could indicate that the decreased
proportion of Tregs cannot induce colitis in the absence of specific microbiota strains. However, in our model, the decrease of microbiota caused by antibiotic treatment further lead to the decrease in the proportion of Treg cells. The immune system was less able to suppress colitis. At the same time, colitis was still induced by the few remaining microbiota; therefore, colitis was further aggravated. As we all know, Crohn's disease is thought to be a Th1 mediated disease [26]. Th1 usually deals with infections by viruses and certain bacteria and tend to be pro-inflammatory [27,28]. In our study, antibiotic treatment led to a decrease in the proportion of Th1 cells in the colon. Considering that Th1 is an important proinflammatory cell, the proportion of Th1 cells decreases, and colitis should be alleviated. But in our study, colitis was still aggravated. This phenomenon may be explained the following reasons. Firstly, the percentage and absolute number of Th1 declines are small, in comparison with Treg cell. So they do not play a major role in the development of colitis. On the contrary, the changes in the overall constitution of the intestinal flora caused by antibiotic treatment play a more important role and the effect of exacerbating colitis is 7
Biomedicine & Pharmacotherapy 114 (2019) 108849
B. Shen, et al.
(caption on next page)
8
Biomedicine & Pharmacotherapy 114 (2019) 108849
B. Shen, et al.
Fig. 5. The CD4+ Foxp3+ T cell (Treg), CD4+ IFN-γ+ T cell (Th1) and macrophage proportion changes of IL10−/− mice with the treatments of antibiotics and water. A. The CD4+ Foxp3+ T cell (Treg) and the CD4+ IFN-γ+ T cell (Th1) cell proportions in the LPLs from the IL10−/− mice receiving the antibiotic treatment and the water treatment; n = 5 mice/group. B–C. The statistical analysis of Treg cells (left panel) and Th1 cells (right panel). The data are expressed as the mean ± SD from the five mice per group. The statistical analyses were done with and unpaired t-test. *P < 0.05; **P < 0.01; ***P < 0.001. D. The macrophage ratio in the LPLs from the IL10−/− mice receiving the antibiotic treatment and the water treatment; n = 5 mice/group. E. The statistical analysis of macrophage ratios. The data are expressed as the mean ± SD from the five mice per group. The statistical analyses were done with an unpaired t-test. NS, not statistically significant. Fig. 6 SCFAs decreased in the IL10−/− mice with the treatments of antibiotics and water and mode pattern.
The SCFA content changed in the colons of IL10−/− mice with the treatments of antibiotics and water; n = 5 mice/group. The data are expressed as the mean ± SD; *P < 0.05.
greater than that of attenuating colitis caused by Th1 cells. Microbiota in the colon not only affects the immune system but also affects the production of SCFAs, which are the metabolites produced by the microbiota [29]. SCFAs provide an important nutrient supply for colonocytes and enhance the barrier function of the intestinal mucosa, which can avoid the excessive activation of the immune system caused by the overwhelming influence on the immune system by the symbiotic microbiota [30]. At the same time, SCFAs can also promote the generation of Tregs. The acetylation pattern of certain histone residues at the Foxp3 promoter or at the conserved noncoding sequence 3 (CNS3, next to the Foxp3 promoter) were changed by SCFAs [31,32]. The
above studies are also consistent with the changes of SCFAs seen in our study. After treatment with multiantibiotics, the decrease in the symbiotic microbiota caused a lower production of SCFAs. Finally, the decrease of SCFAs further aggravated colitis by indirectly decreasing Tregs and by directly affecting the intestinal epithelium. However, there are some limitations in our study, especially in deciding CD4+FOXP3+ as an important marker of T regulatory cells in our experiments. The most specific and reliable marker for naturally occurring CD4+ T regulatory cells is the transcription factor Foxp3, which is expressed by the majority of CD25+CD4 + T cells and a fraction of CD25-CD4+ T cells [33]. There are certain differences
Fig. 7. Mode pattern. A: The decrease in the microbiota was caused by the multiantibiotic regimen and lead to a decrease in the proportion of Tregs and SCFAs, which are necessary to maintain intestinal homeostasis. All changes lead to further exacerbated colitis in IL10−/− mice. 9
Biomedicine & Pharmacotherapy 114 (2019) 108849
B. Shen, et al.
between the percentages of CD25+ Foxp3+ T cell population and the percentage of CD25+Foxp3+ T cell population [34]. But as shown in one study, the percentage of CD4+Foxp3+ as a marker of T regulatory cells is well correlated with the percentage of CD25+Foxp3+ [34]. In our study, we compared the IL10−/− mice, using CD4+Foxp3+ to define T regulatory cells. Because percentage of CD4+Foxp3+ and percentage of CD25+Foxp3+ have a good correlation, it does not affect the conclusions of the experiment. In summary, our study found that it was not feasible to attenuate colitis in IL10−/− mice with the multiantibiotic treatment regimen that was used to develop the pseudo-germ-free mouse model. In contrast, this treatment caused exacerbated colitis. The reason may be that the large amounts of antibiotics had completely destroyed the microbiota composition and abundance. Although a large number of harmful bacteria was removed, a large number of symbiotic microbiota was also cleared. The effects of the microbiota composition may be similar to, or may even be more important than, the effects of pathogen colonization in colitis development. In our opinion, blindly removing all symbiotic microbiota, rather than selectively removing the harmful bacteria, would lead to further exacerbated colitis. The mechanism may be that the decrease in microbiota leads to a decrease in the proportion of Tregs and SCFAs, which are necessary to maintain intestinal homeostasis. All changes lead to further exacerbated colitis in IL10−/− mice, which is similar to the effects demonstrated in the model (Fig. 7A).
[4] R.K. Sellon, S. Tonkonogy, M. Schultz, L.A. Dieleman, W. Grenther, E. Balish, D.M. Rennick, R.B. Sartor, Resident enteric bacteria are necessary for development of spontaneous colitis and immune system activation in interleukin-10-deficient mice, Infect. Immun. 66 (11) (1998) 5224–5231. [5] K.L. Madsen, J.S. Doyle, M.M. Tavernini, L.D. Jewell, R.P. Rennie, R.N. Fedorak, Antibiotic therapy attenuates colitis in interleukin 10 gene-deficient mice, Gastroenterology 118 (6) (2000) 1094–1105. [6] F. Hoentjen, H.J. Harmsen, H. Braat, C.D. Torrice, B.A. Mann, R.B. Sartor, L.A. Dieleman, Antibiotics with a selective aerobic or anaerobic spectrum have different therapeutic activities in various regions of the colon in interleukin 10 gene deficient mice, Gut 52 (12) (2003) 1721–1727. [7] H.C. Rath, M. Schultz, R. Freitag, L.A. Dieleman, F. Li, H.J. Linde, J. Scholmerich, R.B. Sartor, Different subsets of enteric bacteria induce and perpetuate experimental colitis in rats and mice, Infect. Immun. 69 (4) (2001) 2277–2285. [8] W. Hans, J. Scholmerich, V. Gross, W. Falk, The role of the resident intestinal flora in acute and chronic dextran sulfate sodium-induced colitis in mice, Eur. J. Gastroenterol. Hepatol. 12 (3) (2000) 267–273. [9] L.A. Dieleman, M.J. Palmen, H. Akol, E. Bloemena, A.S. Pena, S.G. Meuwissen, E.P. Van Rees, Chronic experimental colitis induced by dextran sulphate sodium (DSS) is characterized by Th1 and Th2 cytokines, Clin. Exp. Immunol. 114 (3) (1998) 385–391. [10] N.A. Nagalingam, C.J. Robinson, I.L. Bergin, K.A. Eaton, G.B. Huffnagle, V.B. Young, The effects of intestinal microbial community structure on disease manifestation in IL-10-/- mice infected with Helicobacter hepaticus, Microbiome 1 (1) (2013) 15. [11] A.C. Ericsson, A.R. Personett, G. Turner, R.A. Dorfmeyer, C.L. Franklin, Variable colonization after reciprocal fecal microbiota transfer between mice with low and high richness microbiota, Front. Microbiol. 8 (2017) 196. [12] F. Backhed, R.E. Ley, J.L. Sonnenburg, D.A. Peterson, J.I. Gordon, Host-bacterial mutualism in the human intestine, Science (New York, N.Y.) 307 (5717) (2005) 1915–1920. [13] N.R. Shin, T.W. Whon, J.W. Bae, Proteobacteria: microbial signature of dysbiosis in gut microbiota, Trends Biotechnol. 33 (9) (2015) 496–503. [14] J. Mirpuri, M. Raetz, C.R. Sturge, C.L. Wilhelm, A. Benson, R.C. Savani, L.V. Hooper, F. Yarovinsky, Proteobacteria-specific IgA regulates maturation of the intestinal microbiota, Gut Microbes 5 (1) (2014) 28–39. [15] T. Selvanantham, Q. Lin, C.X. Guo, A. Surendra, S. Fieve, N.K. Escalante, D.S. Guttman, C.J. Streutker, S.J. Robertson, D.J. Philpott, T. Mallevaey, NKT celldeficient mice harbor an altered microbiota that fuels intestinal inflammation during chemically induced colitis, J. Immunol. (Baltimore, Md.: 1950) 197 (11) (2016) 4464–4472. [16] N. Maharshak, C.D. Packey, M. Ellermann, S. Manick, J.P. Siddle, E.Y. Huh, S. Plevy, R.B. Sartor, I.M. Carroll, Altered enteric microbiota ecology in interleukin 10-deficient mice during development and progression of intestinal inflammation, Gut Microbes 4 (4) (2013) 316–324. [17] F.A. Carvalho, O. Koren, J.K. Goodrich, M.E. Johansson, I. Nalbantoglu, J.D. Aitken, Y. Su, B. Chassaing, W.A. Walters, A. Gonzalez, J.C. Clemente, T.C. Cullender, N. Barnich, A. Darfeuille-Michaud, M. Vijay-Kumar, R. Knight, R.E. Ley, A.T. Gewirtz, Transient inability to manage proteobacteria promotes chronic gut inflammation in TLR5-deficient mice, Cell Host Microbe 12 (2) (2012) 139–152. [18] P.D. Lister, D.J. Wolter, N.D. Hanson, Antibacterial-resistant Pseudomonas aeruginosa: clinical impact and complex regulation of chromosomally encoded resistance mechanisms, Clin. Microbiol. Rev. 22 (4) (2009) 582–610. [19] H.M. Wexler, Bacteroides: the good, the bad, and the nitty-gritty, Clin. Microbiol. Rev. 20 (4) (2007) 593–621. [20] R.J. Newton, J.L. Vandewalle, M.A. Borchardt, M.H. Gorelick, S.L. McLellan, Lachnospiraceae and Bacteroidales alternative fecal indicators reveal chronic human sewage contamination in an urban harbor, Appl. Environ. Microbiol. 77 (19) (2011) 6972–6981. [21] A.F. Schulfer, T. Battaglia, Y. Alvarez, L. Bijnens, V.E. Ruiz, M. Ho, S. Robinson, T. Ward, L.M. Cox, A.B. Rogers, D. Knights, R.B. Sartor, M.J. Blaser, Intergenerational transfer of antibiotic-perturbed microbiota enhances colitis in susceptible mice, Nat. Microbiol. 3 (2) (2018) 234–242. [22] K. Atarashi, T. Tanoue, T. Shima, A. Imaoka, T. Kuwahara, Y. Momose, G. Cheng, S. Yamasaki, T. Saito, Y. Ohba, T. Taniguchi, K. Takeda, S. Hori, I.I. Ivanov, Y. Umesaki, K. Itoh, K. Honda, Induction of colonic regulatory T cells by indigenous Clostridium species, Science (New York, N.Y.) 331 (6015) (2011) 337–341. [23] M.B. Geuking, J. Cahenzli, M.A. Lawson, D.C. Ng, E. Slack, S. Hapfelmeier, K.D. McCoy, A.J. Macpherson, Intestinal bacterial colonization induces mutualistic regulatory T cell responses, Immunity 34 (5) (2011) 794–806. [24] F. Boehm, M. Martin, R. Kesselring, G. Schiechl, E.K. Geissler, H.J. Schlitt, S. Fichtner-Feigl, Deletion of Foxp3+ regulatory T cells in genetically targeted mice supports development of intestinal inflammation, BMC Gastroenterol. 12 (2012) 97. [25] J.L. Round, S.K. Mazmanian, Inducible Foxp3+ regulatory T-cell development by a commensal bacterium of the intestinal microbiota, Proc. Natl. Acad. Sci. U. S. A. 107 (27) (2010) 12204–12209. [26] D.Q. Shih, S.R. Targan, D. McGovern, Recent advances in IBD pathogenesis: genetics and immunobiology, Curr. Gastroenterol. Rep. 10 (6) (2008) 568–575. [27] J. Zhu, W.E. Paul, CD4 T cells: fates, functions, and faults, Blood 112 (5) (2008) 1557–1569. [28] H. Bashyam, Th1/Th2 cross-regulation and the discovery of IL-10, J. Exp. Med. 204 (2) (2007) 237. [29] D.J. Morrison, T. Preston, Formation of short chain fatty acids by the gut microbiota and their impact on human metabolism, Gut Microbes 7 (3) (2016) 189–200. [30] M.A. Chapman, M.F. Grahn, P. Giamundo, P.R. O’Connell, D. Onwu, M. Hutton, J. Maudsley, B. Norton, J. Rogers, N.S. Williams, New technique to measure mucosal metabolism and its use to map substrate utilization in the healthy human large bowel, Br. J. Surg. 80 (4) (1993) 445–449. [31] N. Arpaia, C. Campbell, X. Fan, S. Dikiy, J. van der Veeken, P. deRoos, H. Liu, J.R. Cross, K. Pfeffer, P.J. Coffer, A.Y. Rudensky, Metabolites produced by
Funding This work was supported by grants from the Shanghai Municipal Natural Science Foundation (No. 16ZR1430700), Shanghai Health and Family Planning Commission Scientific Funding (No. 201540227) and National Natural Science Foundation of China (No. 31870021) to QJ Wang. The authors also gratefully acknowledge financial support from China Scholarship Council. Author contribution QJ Wang designed and managed the project. QJ Wang and B Shen performed all the experiments such as antibiotics treatment, 16S rRNA sequencing, Real-time PCR, H&E staining, ELISA, flow cytometry, and animal colonoscopy. QJ Wang, YW Sun and B Shen wrote the manuscript. JJ Hu, H Song, ZT Wang, JG Fan and YW Sun gave lots of suggestions for the experiment design and manuscript writing. All the authors read the paper. Competing interests All the authors declare no competing interests. Acknowledgments We thank Prof. Zhinan Yin (Jinan University) for kind donation of the IL10−/− mice. Many thanks to Dongping Chen (Shanghai Institute of Immunology) for suggestion of flow cytometry analysis, Huahong Fang (Shanghai Institute of Immunology) for suggestion of 16S rRNA sequence and J Qi (Fudan University) with assistance of 16S rRNA bioinformatics analysis. References [1] L.M. Keubler, M. Buettner, C. Hager, A. Bleich, A multihit model: colitis lessons from the interleukin-10-deficient mouse, Inflamm. Bowel Dis. 21 (8) (2015) 1967–1975. [2] R. Kuhn, J. Lohler, D. Rennick, K. Rajewsky, W. Muller, Interleukin-10-deficient mice develop chronic enterocolitis, Cell 75 (2) (1993) 263–274. [3] A. Bleich, M. Mahler, C. Most, E.H. Leiter, E. Liebler-Tenorio, C.O. Elson, H.J. Hedrich, B. Schlegelberger, J.P. Sundberg, Refined histopathologic scoring system improves power to detect colitis QTL in mice, Mamm. Genome 15 (11) (2004) 865–871.
10
Biomedicine & Pharmacotherapy 114 (2019) 108849
B. Shen, et al. commensal bacteria promote peripheral regulatory T-cell generation, Nature 504 (7480) (2013) 451–455. [32] Y. Furusawa, Y. Obata, S. Fukuda, T.A. Endo, G. Nakato, D. Takahashi, Y. Nakanishi, C. Uetake, K. Kato, T. Kato, M. Takahashi, N.N. Fukuda, S. Murakami, E. Miyauchi, S. Hino, K. Atarashi, S. Onawa, Y. Fujimura, T. Lockett, J.M. Clarke, D.L. Topping, M. Tomita, S. Hori, O. Ohara, T. Morita, H. Koseki, J. Kikuchi, K. Honda, K. Hase, H. Ohno, Commensal microbe-derived butyrate induces the
differentiation of colonic regulatory T cells, Nature 504 (7480) (2013) 446–450. [33] S. Sakaguchi, Naturally arising CD4+ regulatory t cells for immunologic self-tolerance and negative control of immune responses, Annu. Rev. Immunol. 22 (2004) 531–562. [34] A.L. Rodriguez-Perea, E.D. Arcia, C.M. Rueda, P.A. Velilla, Phenotypical characterization of regulatory T cells in humans and rodents, Clin. Exp. Immunol. 185 (3) (2016) 281–291.
11