16S rRNA evaluation of the fecal microbiome

Following the manufacturer’s instructions, DNA was extracted from thawed frozen (−80°C) fecal samples collected during necropsy using the Norgen Biotek Corp Stool DNA isolation kit (Catalog number: 27600). Using 341F/805R primers, a 465 bp amplicon was generated by targeting the V3 and V4 regions of the 16S rRNA gene. Then, sequencing was performed on the Illumina MiSeq platform (according to the manufacturer’s specifications) at LC Sciences to yield 250 bp paired-end reads in either direction (Johnson et al., 2021).

The obtained raw data files were demultiplexed, filtered, and processed to merge paired-end reads into a single continuous sequence tag. In contrast to exploiting sequence similarity within the Quantitative Insights Into Microbial Ecology (QIIME) platform, OTU Clustering was performed using the Divisive Amplicon Denoising Algorithm (DADA2). DADA2 reduces background noise and corrects sequencing errors by filtering, dereplication, chimeric filtering, and other methods. This improves data accuracy, species resolution, and reliability of results (Callahan et al., 2016). Taxonomy annotation was performed using the ribosomal database project (RDP) classifier tool for alignment with the corresponding OTU tags (Wang et al., 2007; Cole et al., 2014). Alpha diversity was then determined based on the number of OTUs/species (Chao-1 index) and uniformity (Shannon index). With the aid of both the R vegan package (Oksanen et al., 2016) and GraphPad Prism 9 software, principal component analysis (PCA) based on the weighted UniFrac distance matrix was used to evaluate Beta diversity (test for phylogenetic relatedness) (Motulsky, 2020). Bacterial taxa data were reformatted to serve as an input on the online Galaxy/Hutlab web platform for linear discriminant analysis effect size (LEfSe) analysis for taxa discrimination found in the two studied groups. For this analysis, the cut-offs used were linear discrimination analysis (LDA) ≥ 2 and p-values for Kruskal–Wallis and Pairwise Wilcoxon tests set to ≤0.05 (Hutlab; Segata et al., 2011).

Plasma and CSF were collected at necropsy, frozen and shipped on dry ice for Luminex assays. Monoclonal antibodies were covalently bound to carboxylated Luminex beads following the carbodiimide procedure according to the manufacturer’s suggestion and washed and resuspended in PBS-0.5% Tween 20. Efficacy of coupling was confirmed with 1 μg/mL R-phycoerythrin goat anti-mouse IgG (H + L) antibody (Molecular Probes, Inc., Eugene, OR, USA). Commercial kits were run in individual plates with buffers and standards according to manufacturer’s directions. Following standardization, frozen samples were thawed quickly and 50 μL aliquots combined with coated beads. The acquisition gate was set between 8,000 and 13,500, and 100 events/region later acquired. Data were analyzed with the MasterPlex QT quantification software (MiraiBio Inc., Alameda, CA, USA). The following analytes were chosen for our analysis: macrophage inflammatory protein 1-alpha (MIP-1α), MIP-1β, Interleukin-6 (IL-6), interferon-γ-inducible protein 10 (IP-10), Interleukin-8 (IL-8), interferon-alpha (IFN-α), IFN-γ, Eotaxin, Interleukin-12p40 (IL-12p40), Interleukin-18 (IL-18), Interferon–inducible T-cell alpha chemoattractant (I-TAC), TNF-α, soluble CD40L (sCD40L), monokine induced by interferon-gamma (MIG), macrophage migration inhibitory factor (MIF), Interleukin-1 receptor a (IL-1Ra), Lymphatic Vessel Endothelial Receptor 1 (LYVE-1), Monocyte Chemoattractant Protein-1 (MCP-1), Regulated upon activation, normal T-cell expressed and presumably secreted (RANTES), Myeloperoxidase (MPO), Indoleamine 2, 3-dioxygenase (IDO), C-reactive protein (CRP) and perforin (Supplementary Table 2B). Measures for the levels of each analyte were interpolated from best fitting curves that were generated using commercial standards as previously described (Giavedoni, 2005).