Study designThe effects of SCFAs alone on experimental and control mice were first investigated in a pilot study (N = 3 for each group). We then conducted two-stage experiments. In the first stage, we compared the effects of SCFAs on an IMQ-induced psoriasis-like inflammation mouse model. Experimental mice were divided into four groups (N = 16): those given drinking water with SCFAs plus topical phosphate buffered saline (PBS) (n = 3), those given SCFAs plus IMQ (n = 5), those given ordinary drinking water without SCFAs plus topical PBS (n = 3), and those given ordinary drinking water without SCFAs plus topical IMQ (n = 5) (Supplementary Fig. S1A). In the second stage, the effects of anti-IL17 treatment on IMQ mice were compared in the presence and absence of SCFAs. Experimental mice were divided into eight groups (n = 40) depending on SCFA supplementation or anti-IL17 treatment. Non-SCFA supplementation groups included IMQ plus IgG-isotype (n = 5), IMQ plus anti-IL17 (n = 5), PBS plus anti-IL17 (n = 5), and PBS plus IgG-isotype (n = 5). SCFA groups included SCFAs with IMQ plus IgG-isotype (n = 5), SCFAs with IMQ plus anti-IL17 (n = 5), SCFAs with PBS plus anti-IL17 (n = 5), and SCFAs with PBS plus IgG-isotype (n = 6) (Supplementary Fig. S1B). To ensure consistent results, each experiment was replicate as needed. Comparisons among groups were performed for skin thickness, organ weights, inflammatory cytokine profiles, and fecal microbiota. Animal management and treatment protocols are described below.Animal managementAll experimental procedures were approved by the Institutional Animal Care and Utilization Committee (IACUC) of Taichung Veterans General Hospital (La-1091737) and performed in accordance with institutional guidelines. Male C57BL/6JNarl mice, aged four weeks, were obtained from the National Laboratory Animal Center (Taipei, Taiwan). They were initially housed in separate cages (3–5 mice per cage, L320*W215*H130mm, SHINETEH) with free access to water and food. The mice in the non-SCFA supplementation groups were provided with standard chow (5001, Laboratory Rodent Diet, USA) and the mice in the SCFA groups were provided with standard chow plus SCFA supplements (Sigma, Germany). Housing conditions included a constant temperature of 22 ± 1 °C and a regular 12-h light/dark cycle (lights on 8 PM to 8 AM, with Snai-chip #1511 bedding, Young Li). After three days of acclimatization, the 4-week-3-day-old mice received antibiotics for two days and standard chow (5001, Laboratory Rodent Diet) for 21 days or standard chow with SCFA supplements. The 7-week-5-day-old mice were reallocated to different groups in each arm in separate experiments.Prior to experiments, the mice were randomly distributed among the groups to ensure a comparable average body weight in each group. Investigators were not blinded during the experiments and analyses.IMQ treatment and administrationA daily topical dose of 80 mg of commercially available IMQ cream (5%) (Aldara; 3 M Health Care, Leicestershire, UK) was applied to the shaved back and right ear for seven consecutive days, translating into a daily dose of 4 mg of the active compound21. Control mice were treated similarly with a control topical PBS solution.AntibioticsTo avoid background gut microbiota signals, Ciprofloxacin (0.2 g/L) (17,850, Sigma) and Metronidazole (1 g/L) (9,002,409, Cayman)22 were added to drinking water two days before SCFA-supplemented diet or standard chow diet.SCFA supplementTo explore the effects of SCFAs on inflammation in IMQ-treated mice, experimental mice were provided with standard chow and drinking water supplemented with SCFAs (Sigma Aldrich, Germany) at a final concentration of 150 mM sodium propionate (P1880, Sigma) and 150 mM sodium butyrate (303,410 Sigma) for three weeks before and throughout IMQ treatment23,24. The control mice received standard chow and ordinary drinking water only.Treatment with IL-17 antibody or isotype of anti-IL-17 antibodyUltra-LEAF™ Purified anti-mouse anti-IL-17A Antibody (506,945, Biolegend), 100 µg/mouse, was given intraperitoneally every other day during the experimental period, with or without IMQ treatment. To avoid background non-specific signals from target antibodies25, we used IgG-isotype as a negative control. Control mice were injected with isotype (Ultra-LEAF™ Purified Rat IgG1, κ Isotype Ctrl Antibody, 400,431, Biolegend), 100 µg/mouse, every other day in addition to IMQ or PBS application.Skin thickness, body weight, and tissue weight measurementsTo avoid investigator bias, we measured the thicknesses of the right ear and back skin by digital caliper daily during the seven days of IMQ treatment. Skin samples from the ear and back, as well as samples of spleen, liver, and fat, were collected at the end of the experiment and fixed in 10% formaldehyde for H&E staining (Sigma, Germany, Leica Autostainer XL ST5010). Immunohistochemical staining was performed on skin and intestinal epithelia, for tight junction, anti-Claudin 3 antibody (AFFINITY, AF0129) and anti-Claudin 2 Antibody (ab53032, abcam). Each skin section was measured at three random spots and the average skin thickness was calculated. We calculated the thicknesses of the epidemis and dermis of the different treatment groups using Image J software. (National Institute of Health, NIH).Serum cytokine measurementsSerum samples were collected after the experiments. The cytokine levels of mouse serum (20 × diluted) were measured on Multiplex cytokine bead array assay with MILLIPLEX MAP Mouse Cytokine/Chemokine Magnetic Bead Panel (MCYTMAG-70 KPX32; Millipore). All measurements were performed with a MAGPIX instrument (Luminex, Austin, TX), following the manufacturer’s instructions. Data were collected with xPonent software (Luminex). Mann–Whitney U test was used to calculate the significance levels between treatment groups. All results are expressed as mean ± SEM (Standard Error of the Mean). Moreover, p values less than 0.05 were considered significant.Stool sample collection, DNA extraction, library construction, and metagenomic sequencingStool samples were collected and examined before and after IMQ (or PBS) treatment. Stool samples were obtained from each mouse after physical restraint, directly from the rectal aperture to avoid contamination. An Eppendorf tube pre-filled with Inhibitex Buffer, provided in the QIAamp Fast DNA Stool Mini Kit (Qiagen), was used to collect and stabilize the fecal specimen. The collection tube was carefully inserted into the rectal opening, ensuring minimal distress to the tissue. The sample was then secured in the tube, which was subsequently sealed and labeled with the appropriate identification details. Samples were stored at − 20 °C until DNA extraction was performed. The collected fecal sample (about 200 mg) was used for total bacterial DNA extraction with QIAamp Fast DNA Stool Mini Kit (Qiagen, MD, USA) according to the manufacturer’s instructions. The quality and quantity of DNA were determined with NanoDrop ND-1000 (Thermo Scientific, Wilmington, DE, USA) and DNA was stored at − 80 °C before library construction and sequencing.Extracted DNA (about 500 ng) was fragmented to approximately 350 base pairs by Covaris S2 system (Covaris, Inc., Woburn, MA, USA) and then subjected to library construction with Illumina DNA Prep Kit (Illumina, San Diego, CA). Sequencing was performed using Illumina NovaSeq 6000 platform, resulting in paired-end (PE) reads of 150 bp in length.Metagenomic raw read processingOn a per-sample basis, raw read quality control was performed using the Kneaddata pipeline (https://github.com/biobakery/kneaddata), which integrates Trimmomatic26 for trimming Illumina adaptors and low-quality regions and filtering short reads, as well as Bowtie227 for identifying and removing host contamination from human hg38 build (or mouse mm10 build) and PhiX genome. Trimmomatic options included “ILLUMINACLIP:NexteraPE-PE.fa:2:30:10” (identifying and removing adapters), “SLIDINGWINDOW:4:20” (trimming low quality region), and “MINLEN:50” (discarding reads shorter than 50 base pairs). After that, reads with low-complexity region and repeated sequences were identified and removed using Komplexity software (https://github.com/eclarke/komplexity) with default settings.Metagenome assembly, annotation, and binningMetagenome assembly was performed using MEGAHIT28 to assemble the clean reads into contigs. In terms of taxonomic identification, there were two approaches: (1) read-based taxonomy, which analyzes QC-passed reads using Sourmash29 to estimate relative abundance of taxa based on GTDB taxonomy30 and (2) contig-based taxonomy, which predicts taxonomy of contigs using MMSeqs231,32 easy-taxonomy pipeline and estimates abundance by contigs depth, derived by the jgi_summarize_bam_contig_depths script analyzing bowtie2 alignment that maps the clean reads onto contig.Open reading frame (ORF) prediction and functional annotation [including the Clusters of Orthologous Groups (COGs) family and Enzyme Commission (EC) number assignment] was performed by subjecting contigs to Prokka33. ORFs were also analyzed to identify corresponding KEGG Orthology (KO) number, Carbohydrate Active enZyme (CAZy) family, antibiotics gene, and virulence factor using MMSeqs2 easy-search pipeline. MinPath34 was used for pathway reconstruction by analyzing KO (for KEGG pathway) and EC (for MetaCyc pathway)35 profiles of each sample. The abundance of gene families (or categories) was estimated by accumulating ORF depths, which were calculated by tpm_table python script (https://github.com/EnvGen/toolbox) based on the number of unique reads mapped on each ORF and presented in units of transcript per million (TPM)36.Metagenomics binning was performed by MetaBAT37 to cluster contigs into genome “bins” [i.e., metagenome-assembled genomes (MAGs)]. The quality was assessed and taxonomically identified by CheckM lineage-specific workflow38. In addition, the classification of MAGs was based on a voting approach using the contig-based taxonomy of binned contigs, deepening the CheckM-predicted taxonomy in most cases.Statistical analysis, bioinformatics analysis, and microbial gene function predictionGraphPad Prism software 10 (GraphPad Software Inc., San Diego, CA, USA) was used to analyze experimental and clinical data. One-way analysis of variance was used for inter-group comparisons. P < 0.05 was considered statistically significant. Other statistical analyses were performed using R (http://www.r-project.org/), unless otherwise specified. The read- or contig-based taxonomic profiles were imported and handled by R package phyloseq39 and processed for alpha diversity estimation. Beta diversity was analyzed and visualized by principal coordinate analysis (PcoA) via the R package ade440 based on Bray–Curtis distance of species-level relative abundance profile. As for functional profiles, alpha diversity was represented by the number of observed entities. Beta diversity was also analyzed by PcoA based on Bray–Curtis distance of functional genes, families, or relative abundance profiles.Between-group inertia percentages were tested based on the Monte-Carlo test (with 10,000 permutations) using Bray-Curtis41, with p values less than 0.05 considered significant. To identify organismal features differentiating communities of fecal microbiota between experimental and control mice, linear discriminant analysis (LDA) effect size (LefSe)42 was applied with α of 0.05 (Kruskal–Wallis and Wilcoxon tests) and effect size threshold of 2 on linear discriminant analysis (LDA) using the stand-alone implementation (https://bitbucket.org/nsegata/lefse).The abundances of various gene families of experimental and control mice were scaled by total sum per sample and subjected to enrichment analysis of two-group comparison using several tests, including Wilcoxon signed rank test and ANOVA rank test or Kruskal–Wallis test with a Benjamin-Hochberg false discovery rate (FDR) correction to adjust p values for multiple testing.The Bioinformatics analyses mentioned above were carried out by Germark Biotechnology Co., Ltd. (Taichung, Taiwan).The experiments were carried out in accordance with the protocols and approval of the IACUC Taichung Veterans General Hospital, La-1091737.Approval of animal study and animal welfareThe experiments were carried out in accordance with the protocols approved by the IACUC of Taichung Veterans General Hospital, La-1091737. The present study followed national guidelines of the 3Rs for humane animal treatment and complied with relevant legislation of the Ministry of Agriculture, Taiwan. The study is reported in accordance with ARRIVE guidelines.
