Multi-parametric integration reveals exercise and prebiotics reverse microbiome mediated bone loss

Introduction: : The gut microbiome is a vast, readily modifiable, and impactful ecosystem consisting of the trillions of bacteria residing within the gut. Due to the rise of western diet and other environmental factors there is a rising prevalence in disrupted gut microbiome [1]. At the same time the rates of low bone mass and bone diseases are increasing in aging populations [2]. Recent work has shown the gut microbiome plays a role in bone homeostasis during and after skeletal maturity [3, 4]. Although there is a wealth of evidence illustrating the link between the gut microbiome and bone, the ability to reverse antibiotic driven bone changes through microbiome manipulation remains unknown. This lack of understanding hinders our ability to capitalize on this accessible ameliorative treatment for microbiome-mediated bone loss. Others have shown that moderate treadmill exercise may modulate gut health [5], and increases bone formation [6]. Additionally, others have shown the prebiotic inulin may improve both microbial diversity and bone health [7]. Since exercise and inulin both modulate the microbiome and bone, we hypothesize that intervention with one or both will reverse deleterious antibiotic-driven microbiome and bone changes in a mouse model of antibiotic induced microbiome disruption.

Materials and

Methods: : Under IACUC protocol 2104002138, we treated mice (n=5+ M/F per group) with 1g/L ampicillin + 0.5g/L neomycin from 5 to 16 weeks of age. At 16 weeks, they were assigned to continue Antibiotics ) Cessation of antibiotics, or cessation with an intervention up to 24 weeks. Interventions consisted of ceasing antibiotic treatment and moderate treadmill exercise at 13 m/min for 5 days/week (C+E), prebiotic supplementation of 10 g/L inulin in drinking water (C+P), or both exercise and prebiotic treatment (C+E+P) (Figure 1). Control mice (n=6 M/F) received normal drinking water from 5 to 24 weeks.
At euthanasia, we collected cecum contents for 16S rRNA sequencing (Zymo Research) and hindlimbs for micro-CT (Bruker Skyscan 1272). We segmented distal femoral metaphyseal trabecular bone (Dragonfly, ORS) and quantified [8] trabecular thickness, separation, number, and bone volume fraction (Figure 1).
We compared Shannon diversity among treatment groups and analyzed microbial community similarity using principal coordinate analysis (PCoA) of amplicon sequence variants (ASVs) using phyloseq package in R Studio. We performed pairwise comparisons on bone morphometric parameters. All univariate comparisons were conducted using the rstatix package.
We used Data Integration Analysis for Biomarker Discovery (DIABLO) [9] to integrate the 30 most abundant genera and bone morphometry and derive an Antibiotics-vs.- Control latent variable space. We assessed intervention effectiveness by projecting data from all intervention groups onto the Antibiotics-Control derived latent space using the predict function using mixOmics package. Statistical significance was defined at p< 0.05.

Results, Conclusions, and Discussions:: Shannon Index was significantly less than control in Antibiotics (p < 0.001), Cessation (p < 0.001) and C+P (p =0.045) groups but was not different with C+E and C+E+P groups (Figure 1). PCoA of ASVs revealed divergent microbiome compositions between control and continuous dosing (Figure 2). Cessation overlapped with both Antibiotics and Control groups, while C+E, C+P, and C+E+P separated from Antibiotics and overlapped with Control.
Metaphyseal bone volume fraction (p = 0.026) and trabecular thickness (p = 0.008 ) were significantly lower in male Antibiotics mice (Figure 2), but no significant differences were detected in female bone morphometry, suggesting a sex-based difference in bone response to microbiome manipulation. Notably, metaphyseal morphometric parameters from all intervention groups were not significantly different from the control. Since bone volume fraction and trabecular thickness in Antibiotics mice were less than Control, the exercise, prebiotic, and combination interventions successfully reversed antibiotic-mediated bone loss.
DIABLO mutli-omic integration between microbial composition and femoral metaphyseal bone morphometrics resulted in separation between Antibiotics and Control. DIABLO indicated a moderate positive correlation between microbes and metaphyseal bone morphometry (R = 0.56) (Figure 3). Projection of intervention groups onto the Antibiotics-vs.-Control latent variable space indicated that C+E, C+P, and C+E+P groups all shift toward Control while only 4 out of 6 of Cessation samples trend toward control (Figure 3). The correlation between microbes and morphometrics suggests that the microbiome plays a role in the reversal of antibiotic-driven bone loss and that additional intervention after cessation of antibiotics may be necessary for complete restoration of antibiotic-driven changes.
In this study, we demonstrated, through multi-parametric integration, that intervention via exercise, prebiotic treatment, or a combination of these reverses antibiotics-mediated microbiome disruption and associated bone loss in skeletally mature mice. Antibiotic cessation alone does not completely restore antibiotic-driven changes to both bone and gut microbiome. Future work will incorporate complementary computational analysis, including canonical correlation, to determine which specific microbes may drive these changes and will incorporate sex-based comparisons.

Acknowledgements and/or References (Optional):: References
[1] M. K. Zinocker and I. A. Lindseth, "The Western Diet-Microbiome-Host Interaction and Its Role in Metabolic Disease," Nutrients, vol. 10, no. 3, Mar 17 2018, doi: 10.3390/nu10030365.
[2] N. C. Wright et al., "The recent prevalence of osteoporosis and low bone mass in the United States based on bone mineral density at the femoral neck or lumbar spine," J Bone Miner Res, vol. 29, no. 11, pp. 2520-6, Nov 2014, doi: 10.1002/jbmr.2269.
[3] J. D. Guss et al., "The microbial metagenome and bone tissue composition in mice with microbiome-induced reductions in bone strength," Bone, vol. 127, pp. 146-154, Oct 2019, doi: 10.1016/j.bone.2019.06.010.
[4] C. Liu et al., "Microbiome-induced increases and decreases in bone matrix strength can be initiated after skeletal maturity," J Bone Miner Res, vol. 39, no. 11, pp. 1621-1632, Oct 29 2024, doi: 10.1093/jbmr/zjae157.
[5] W. Chen et al., "Moderate intensity treadmill exercise alters food preference via dopaminergic plasticity of ventral tegmental area-nucleus accumbens in obese mice," Neurosci Lett, vol. 641, pp. 56-61, Feb 22 2017, doi: 10.1016/j.neulet.2017.01.055.
[6] J. M. Wallace et al., "Exercise-induced changes in the cortical bone of growing mice are bone- and gender-specific," Bone, vol. 40, no. 4, pp. 1120-7, Apr 2007, doi: 10.1016/j.bone.2006.12.002.
[7] H. Bakirhan and E. Karabudak, "Effects of inulin on calcium metabolism and bone health," Int J Vitam Nutr Res, vol. 93, no. 1, pp. 85-96, Feb 2023, doi: 10.1024/0300-9831/a000700.
[8] M. L. Bouxsein, S. K. Boyd, B. A. Christiansen, R. E. Guldberg, K. J. Jepsen, and R. Muller, "Guidelines for assessment of bone microstructure in rodents using micro-computed tomography," J Bone Miner Res, vol. 25, no. 7, pp. 1468-86, Jul 2010, doi: 10.1002/jbmr.141.
[9] A. Singh et al., "DIABLO: an integrative approach for identifying key molecular drivers from multi-omics assays," Bioinformatics, vol. 35, no. 17, pp. 3055-3062, Sep 1 2019, doi: 10.1093/bioinformatics/bty1054.

This work was supported in part by a Defense Advanced Research Projects Agency Young Faculty Award, funded through the
Army Research Office as Contract W911NF2110372.