Abstract
As human space exploration transitions toward longer-duration missions, understanding physiological adaptations to microgravity is crucial for maintaining crew health. While previous literature has primarily focused on the physical impacts of bone demineralization and skeletal unloading, utilizing the gut microbiome as a highly sensitive biosensor for systemic physiological shifts remains overlooked. This study aims to address this gap by identifying distinct microbial signatures that may predict a host's environmental exposure to microgravity. We hypothesized that microgravity exposure would produce specific microbial changes that could be used to train a Random Forest (RF) machine learning model to predict microgravity exposure in mice. To analyze if microgravity was producing the hypothesized effect on the microbiome, we utilized 16S rRNA sequencing data from the NASA Open Science Data Repository (dataset OSD-417) to analyze fecal samples from mice exposed to microgravity aboard the International Space Station (ISS), comparing them against ground control cohorts over 4.5-week and 9-week mission durations. Microbial community changes were assessed through alpha and beta diversity metrics, core microbiome analysis, differential abundance analysis, and indicator species analysis (ISA), which subsequently informed the training of the RF classification model to predict microgravity exposure. Our results demonstrated that while overall microbial richness and evenness remained relatively stable under microgravity conditions, some microbial taxa underwent significant restructuring. Prolonged microgravity exposure notably decreased the number of shared core microbiome taxa between flight and ground control mice, an effect that persisted even after the mice returned to Earth for recovery. Differential abundance analysis identified a trend towards enrichment of taxa under prolonged microgravity exposure. Furthermore, ISA identified specific taxa, such as Lachnoclostridium, Colidextribacter and Turicibacter, that were strongly associated with microgravity conditions. Leveraging these specific microbial shifts, the RF model successfully distinguished between microgravity-exposed and unexposed microbiomes with high accuracy. Ultimately, these findings indicate that the mice gut microbiome produces a robust predictive signal under microgravity.