Glioma and temozolomide induced alterations in gut microbiome

The gut microbiome is fundamental in neurogenesis processes. Alterations in microbial constituents promote inflammation and immunosuppression. Recently, in immune-oncology, specific microbial taxa have been described to enhance the effects of therapeutic modalities. However, the effects of microbial dysbiosis on glioma are still unknown. The aim of this study was to explore the effects of glioma development and Temozolomide (TMZ) on fecal microbiome in mice and humans. C57BL/6 mice were implanted with GL261/Sham and given TMZ/Saline. Fecal samples were collected longitudinally and analyzed by 16S rRNA sequencing. Fecal samples were collected from healthy controls as well as glioma patients at diagnosis, before and after chemoradiation. Compared to healthy controls, mice and glioma patients demonstrated significant differences in beta diversity, Firmicutes/Bacteroides (F/B) ratio, and increase of Verrucomicrobia phylum and Akkermansia genus. These changes were not observed following TMZ in mice. TMZ treatment in the non-tumor bearing mouse-model diminished the F/B ratio, increase Muribaculaceae family and decrease Ruminococcaceae family. Nevertheless, there were no changes in Verrucomicrobia/Akkermansia. Glioma development leads to gut dysbiosis in a mouse-model, which was not observed in the setting of TMZ. These findings seem translational to humans and warrant further study.


GL261 mouse model of glioma
For intracranial mouse xenograft studies, both young and aged mice were implanted with GL261 cell lines maintained in our laboratory. Briefly, GL261 cells were cultured in Dulbecco's Modified Eagle's Minimal with 10% FBS, penicillin, and streptomycin in a humidified atmosphere with 5% CO2 at 37°C. Mice were anesthetized and stabilized in a stereotactic frame, a burr hole was drilled 2mm lateral and 1mm anterior to bregma in the right hemisphere, to a depth of 3.5mm. 20 GL261 (1×10 5 cells) in 2µl of Hank's buffered salt solution (HBSS) or the same amount of HBSS were implanted over 5 minutes using autoinjectors as a tumor-bearing group or sham group respectively.

Temozolomide
100mg of Temozolomide (Sigma-Aldrich, MO, USA) was dissolved in 1.5ml of DMSO (Sigma-Aldrich, MO, USA) and sonicated three times. The solution was further dissolved in 38.5ml of sterile saline for a final working concentration of TMZ 2.5mg/ml.

Gut Permeability Assay
On day 42, aged mice were fasted for 5 hours prior to sacrifice. One hour prior to sacrifice mice were orally gavaged with 6mg/10g body weight of 4kDa FITC dextran (Sigma-Aldrich, MO, USA), DMSO was used as control to ensure that the differences were not due to solvent of the drug. One hour later blood samples were collected, centrifuged at 4 C at 3000g for 6 minutes and stored at -80 C until use. Samples were analyzed using a fluorescence spectrometer at an excitation wavelength of 428nm and an emission wavelength of 535nm. 21 Microbial DNA extraction and the 16S rRNA gene sequencing 16S rRNA gene compositional analysis provides a summary of the composition and structure of the bacterial component of the microbiome. Genomic bacterial DNA extraction methods were optimized to maximize the yield of bacterial DNA while keeping background amplification to a minimum. 16S rRNA gene sequencing methods were adapted from the methods developed for the Earth Microbiome Project and NIH-Human Microbiome Project. 22-24 Briefly, bacterial genomic DNA was extracted using the Qiagen MagAttract Power Soil DNA Kit. The 16S rRNA V4 region was amplified by PCR and sequenced on the MiSeq platform (Illumina, Inc. CA, USA) using the 2x250 bp paired-end protocol yielding pair-end reads that overlap almost completely. The primers used for amplification contain adapters for MiSeq sequencing and singleindex barcodes so that the PCR products may be pooled and sequenced directly, 24 targeting at least 10,000 reads per sample. CMMR16S (variable region 4 [v4]) rRNA gene pipeline data incorporates phylogenetic and alignment-based approaches to maximize data resolution. The read pairs are demultiplexed based on unique molecular barcodes added via PCR during library generation, then merged using USEARCH v7.0.1090. 25 The subsequent analysis steps of the pipeline leverage custom analytic packages developed at the Alkek Center for Metagenomics and Microbiome Research (CMMR) at Baylor College of Medicine to produce summary statistics and quality control measurements for each sequencing run, as well as multi-run reports and datamerging capabilities for validating built-in controls and characterizing microbial communities across large numbers of samples or sample groups. 16Sv4rRNA sequences are clustered into OTUs at a similarity cutoff value of 97% using the UPARSE algorithm. 26 OTUs are subsequently mapped to an optimized version of the SILVA Database 27 containing only sequences from the v4 region of the 16S rRNA gene to determine taxonomies. Abundances are recovered by mapping the demultiplexed reads to the UPARSE OTUs. A custom script constructs an OTU table from the output files generated in the previous two steps for downstream analyses using a visualization toolkit also developed at the CMMR named ATIMA (Agile Toolkit for Incisive Microbial Analyses).

Patient characteristics and fecal sample collection
This prospective study was approved by the institutional review board of our institution. The study was conducted from January 2018 to July 2019. Patients in whom a fecal sample was obtained prior to surgical resection of a newly diagnosed glioma at our institution were included in the study. Patients with recent antibiotic exposure (30 days) for other related conditions, under 18-year of age, history of cancer, and presence of gastrointestinal diseases (e.g. inflammatory bowel disease) were excluded from the study. All tumors were examined by H&E and immunohistochemistry by a board-certified neuropathologist. Patients' demographics, clinical and follow-up information, including age, gender, race, Karnofsky performance score (KPS), body mass index (BMI), diagnosis according to the 2016 WHO classification of brain tumors, tumor volume, radiographic and volumetric extent of resection (EOR), adjuvant therapy, progression, and survival were obtained from the electronic medical record (EMR) and stored in a prospective REDCap database, which was later exported for data analysis (Table 1). Patients underwent biopsy or maximum safe tumor resection at the discretion of the treating neurosurgeon (YE, NT) followed by radiation therapy with concomitant TMZ (75mg/m2) in the majority of the cases. Following this, the maintenance dose of TMZ according to the Stupp protocol was implemented. 28 IDH1 p.R13H immunohistochemistry was performed with the mutant protein-specific antibody (1:40; H09 monoclonal, Dianova, Hamburg, Germany) in a Dako Omnis (Agilent, California, USA) autostainer. The imaging analysis consisted of the volumetric measurement of three tumor components: Enhancing tumor, necrosis, and non-enhancing FLAIR hyperintensity. The enhancing tumor volume and necrosis volume were obtained from standard-of-care MRIs on routine post-contrast 2D or 3D T1-weighted images before and after the resection. FLAIR signal abnormality volume was measured on standard 2D or 3D T2-FLAIR images. Volumes were calculated using a semi-automated software (TeraRecon, Aquarius Intuition viewer). By using the threshold tool, the software selects volumes with a specific range of signal intensity values; then, the tumor boundaries are delineated utilizing the Region-of-interest (ROI) tool. The software automatically calculated the volume using computational algorithms. In the case of multiple lesions, the sum of all individual lesions was used. Fecal samples were collected from patients at three time points: before resection (Pre-Sx) and prior to prophylactic surgical single dose of cephalosporin per-protocol; prior to initiation of chemoradiation (Pre-Tx), which was collected at least 2-4 weeks after antibiotic administration to avoid antibiotic-induced biome changes; 33 and after completion of 6-weeks of chemoradiation (Post-Tx). Several common factors are known to impinge on the human microbiome composition including various environmental and social factors such as, lifestyle, diet, medications, place of residence, etc. 29,30 To minimize such confounding factors and to increase our sample size, twenty-one fecal samples (9-household family members and 12-non-related individuals) were utilized as controls following the same exclusion criteria as enrolled patients (antibiotic exposure, under 18-year of age, history of cancer, and presence of gastrointestinal diseases).
Progression was analyzed in patients with more than 6 months of follow-up and with information regarding tumor progression or death. Fecal samples for DNA extraction and microbiome sequencing for most patients and control samples were collected using microbial collection and stabilization kits (OMNIgene Gut; DNA Genotek), however 12 controls samples were collected after rectal examination for a non-related condition with immediate freezing of the fecal sample at -80 C. After which, microbial DNA extraction and the 16S rRNA gene sequencing were performed.

Statistical analysis
The ATIMA (Agile Toolkit for Incisive Microbial Analyses) software was used to perform the downstream analyses. ATIMA is an R software suite combining publicly available packages (i.e. APE and VEGAN) and purpose written code to import sample data and identify trends in taxa abundance, alpha-diversity, and beta-diversity using weighted Bray-Curtis PCoA with sample metadata. The significance of categorical variables was determined using the non-parametric Mann-Whitney U test for two category comparisons or the Kruskal-Wallis test when comparing three or more categories. Correlation between two continuous variables is determined with R's base "lm" function for linear regression models, where p-values indicate the probability that the slope of the regression line is zero. PCoA plots employ the Monte Carlo permutation test to estimate p-values. All p-values were adjusted for multiple comparisons with the FDR algorithm.
Firmicutes to Bacteroides ratio (F/B) was calculated in normal Gaussian distribution (verified by the Anderson-Darling test, D'Agostino-Pearson omnibus normality test, Shapiro-Wilk normality test, and Kolmogorov-Smirnov normality test with Dallal-Wilkinson-Lillie for p-value) with paired t-test. Meanwhile, when no normal Gaussian distribution was present, the Mann-Whitney test was performed. PFS was calculated using Kaplan-Meier curves with Log-rank (Mantel-cox) test and the Gehan-Breslow-Wilcoxon test. Kaplan-Meier curves for PFS were generated using GraphPad Prism (version 8.2.1 for Mac, GraphPad, CA, USA). Analysis at phylum, family, and genus level in the aged-mice was performed using a paired t-test, due to the small sample size of this group prior to confirmation of normal distribution (Shapiro-Wilk normality test). All analysis was performed by a blinded investigator and mice were randomly assigned to treatment groups. OS analysis was not performed due to the short follow-up and status (alive) of most patients at the time of analysis (Table 1).