Bioinspired rhyolite particles as a functional material for lipid regulation in diet-induced obesity

Many mammals engage in soil-eating behavior, ingesting minerals derived from rocks to obtain nutrients and promote detoxification in the body. Inspired by this animal behavior, we applied rhyolite particles, a natural mineral, as a modulator of lipid metabolism in high-fat diets (HFD) to investigate its effects in regulating lipid digestion and preventing obesity risk, comparing it with activated bamboo charcoal In vitro digestion models, rhyolite particles, regardless of particle size, inhibited the activity of pancreatic lipase and α-amylase, reducing the release of free fatty acids. Similarly, in vivo, rhyolite supplementation suppressed weight gain, improved plasma and hepatic lipid profiles, and maintained the integrity of intestinal and liver tissue. The bacterial composition of the gut microbiota also improved to a near-normal state. Our findings suggest that rhyolite particles, inspired by soil-eating behavior, are a promising natural material for preventing diet-induced obesity through lipid regulation and interaction with the gut microbiota.

According to the World Health Organization (WHO), 1.9 billion adults were overweight in 2016, with over 650 million classified as obese, leading to 4 million deaths per year (Rakhra et al., 2020; Afshin et al., 2017). Obesity rates in the United States have continued to climb steadily in recent years. The Western diet, particularly high-fat diets (HFD), is the main source of the elevated obesity in the United States. Obesity not only promotes metabolic dysfunction but also elevates the risk of type II diabetes, cardiovascular disease, nonalcoholic fatty liver disease, chronic kidney disorders, several forms of cancer, and musculoskeletal problems (Everard et al., 2013; Ghaben et al., 2019). Current therapeutic strategies, such as anti-obesity medications and bariatric surgery, effectively alleviate obesity (Lagerros et al., 2013; Apovian et al., 2013). However, the long-term use of these treatments is associated with high cost and may induce inflammation, toxicity, and gut microbiota disorder (Almalki et al., 2018).

Given the high costs, serious adverse effects, and potential long-term toxicity of bariatric surgery and pharmacotherapy, there is a pressing need for simpler, low-risk adjuncts to existing obesity treatment. Targeting the absorption of dietary fat, one of the major determinants of excess caloric consumption and body fat accumulation, has emerged as a potentially successful strategy (Wit et al., 2022; Tang et al., 2024). Non-digestible lipid-binding agents can sequester dietary fats in the gastrointestinal tract, thereby reducing their absorption and subsequent energy intake. Novel therapeutic approaches have reported using functional foods and nutraceuticals to promote healthier lifestyles and address the physiological effects of obesity, with recent advances highlighting the potential of emerging nanonutraceuticals against obesity (Ni et al., 2025). Activated charcoal has been investigated for the same purpose, but it is imperfect due to its non-selective adsorption of nutrients and an environmentally harmful production process (Watson et al., 1986; Bikharudin et al., 2025; Chen et al., 2018; Dai et al., 2018; González-García, 2018; Kılıç et al., 2012). Thus, the discovery of a natural, sustainable, and biologically compatible alternative for dietary fat sequestration is a critical objective for developing new anti-obesity strategies.

We draw inspiration from the geophagy behavior of animals, where they consume the soils or rock for supplement and physiological functions, such as improving the digestion, promoting natural detoxification, and modulating the gut environment (Vogel et al., 2019; Houston et al., 2008; Holdø et al., 2002; Sach et al., 2019). Recent studies further reported that geophagy can influence the gut microbiota composition by altering luminal pH, providing mineral surfaces for microbial adhesion, and adsorbing metabolites that shape microbial activity (Neubauer et al., 2019; Prasai et al., 2016; Gerstner et al., 2018; Wang et al., 2012; Chalvatzi et al., 2016), highlighting that geophagy or ingested minerals can exert biological effects beyond the nutrient supplementation. Rhyolite, a volcanic mineral abundant in soils worldwide, has long served in cosmetics, food additives, antibacterial agents, deodorants, and thermal storage applications (https://www.materra.jp/). Its therapeutics potential arises from its silica-rich material composition, which composed of roughly 45%–50% K-feldspar, 25%–30% quartz, 10%–15% albite plagioclase, 3%–7% biotite, and 1%–3% opaque oxides (Elliot, 2018) and its high specific surface area, which together enable adsorption of organic molecules, bile salts, and toxins within the GI tract (Nadziakiewicza et al., 2019; Elliot et al., 2020; Oladele et al., 2025; Howell et al., 2019). Despite these properties, rhyolite remains underexplored as a dietary supplement for suppressing body weight in HFD model by regulating lipid profiles.

Inspired by animal geophagy behavior, we hypothesized that orally administered rhyolite particles could act as a bioinspired, functional mineral for lipid regulation, suppressing weight gain in the HFD-mice model. We first evaluated rhyolite of different particle sizes using an in vitro gastrointestinal model to quantify the effects on digestive enzyme activity, pH, and bile salt interactions. Subsequently, we evaluated the therapeutic efficacy of these supplementary rhyolites in reducing obesity. We compared it to activated bamboo charcoal (ABC) using a mouse model of HFD-induced obesity, assessing body weight, fat deposition, liver and intestinal function, and plasma biochemical data from the treated mice in this study.

Materials

Unless otherwise specified, all chemicals were reagent grade and purchased from FUJIFILM Wako Pure Chemical Corp (Osaka, Japan). Rhyolite powder (Materra, Ehime, Japan) with different particle sizes (Rhyolite-A, Rhyolite-B, and Rhyolite-C) was prepared using the milling treatment.

Preparation of rhyolite particles and activated bamboo charcoal

Rhyolite particles were milled using a planetary ball mill (P-6, Fritsch GmbH, Idar-Oberstein, Germany) using wet conditions at 200 rpm with a ratio of Zirconia ball-to-rhyolite powder about 5:1. Rhyolite-A was obtained with a ball size of 5 mm for 1 h, Rhyolite-B was then milled with a ball size of 2 mm for 3 h, and Rhyolite-C was used with a ball size of 2 mm for 12 h. The particles milled were washed several times with ethanol-water and dried at 105 °C overnight. The physicochemical characteristics of rhyolite milled particles were analyzed by using scanning electron microscopy (SEM, S-4800, Hitachi, Japan), X-ray diffractometer (XRD, MiniFlex, Rigaku, Tokyo, Japan), X-ray fluorescence spectrometry (XRF, JIS R2212-1 Fireclay Refractory, Tokyo, Japan), Fourier-transform infrared spectrophotometer (FT-IR, IR Affinity-1S, Shimadzu, Kyoto, Japan), and the specific surface area (SSA) was determined by the Brunauer–Emmett–Teller method (BET, Macsorb HM model-1208, Mountech, Tokyo, Japan).

Moreover, the preparation of activated bamboo charcoal (ABC) was followed by the method of Banat et al. (2003), which briefly involved an amount of bamboo charcoal powder (BC) (Taketora, Kochi, Japan) with the impregnation weight ratio of BC to 30 wt% KOH solutions of 1:3 at 130 °C for 15 h, and the residue was washed with hot Milli-Q water several times to adjust the pH to become neutral, and the sample was dried at 120 °C overnight. The powder was activated at 800 °C for 1 h and yielded ABC powder. The physicochemical properties of ABC particles showed a particle size of about 10.60 µm (Supplementary Figure S1a), micropore structure (Supplementary Figure S1b), and XRD pattern (Supplementary Figure S1c), which demonstrated characteristics identical to ABC particles.

Gastrointestinal digestion using in vitro obesity model

To simulate fed-state-gastric chyme entering the duodenum, the intestinal digestion phase was conducted at pH 5.0, which relevant the physiologically postprandial pH range (pH 4.5–5.5), commonly used in standardized in vitro digestion models such as INFOGEST. Importantly, individuals with obesity often exhibit gastric pH due to increased gastric volume, altered acid secretion, and buffering by larger meal sizes, postprandial gastric pH to rise, making pH 5.0 closely resemble the gastric chyme conditions relevant to obesity-associated digestion (Steenackers et al., 2021). Briefly, a two-step in vitro digestion model, simulating the gastric and intestinal phases following a high-fat, high-carbohydrate condition, was prepared using FaSSIF/FeSSIF/FaSSGF powder (Biorelevant.com Ltd., London, United Kingdom). Medium-chain triglycerides (MCT; 625 mg; Tricarpin, Tokyo Chemical Industry, Japan), starch (625 mg), and cellulose (31 mg) were dispersed in fed-state simulated gastric fluid (FeSSGF; 10 mL) under continuous stirring at 600 rpm for 10 min at 37 °C (Chen et al., 2022). Rhyolite particles (10% w/w relative to lipid; 62.5 mg) were then added. The pH was adjusted to 5.00 ± 0.01 before initiating gastric lipolysis with Candida antarctica lipase (100 μL; 600 TBU) and maintained for 30 min at constant pH via pH-stat titration (0.1 M NaOH or 0.1 M HCl). Afterward, fed-state simulated intestinal fluid (FeSSIF, 20 mL) was added, and pH was auto-titrated to 6.00 ± 0.01 over 10 min. Intestinal lipolysis was initiated by adding pancreatin extract (2 mL; ∼2000 TBU lipase, ∼1600 TBU α-amylase) and monitored for 60 min, with free fatty acid (FFA) release quantified from NaOH consumption required to maintain pH 6.00 ± 0.01. In this equation, V NaOH represents the volume of NaOH needed to neutralize the FFAs produced (in mL), m NaOH represents the molarity of the NaOH solution (0.1 M), W Lipid represents the total weight of lipid that was initially present in the reaction vessel (625 mg), and M Lipid represents the molecular weight of MCT (470.68 g/mol). Lipase activity was determined using Equation 2, where AULC rhyolite represents the area under the lipolysis curve in the presence of rhyolite particles, and AULC no ryholite corresponds to that obtained without rhyolite particles.

An optical microscope (ECLIPSE Ti2, Nikon, Tokyo, Japan) was employed to observe a lipid droplet in the intestinal phase on a microscope slide, and the lipid droplet was estimated using ImageJ software. Additionally, the rhyolite that precipitated after the intestinal phase was washed several times with deionized water and ethanol to remove any unbound organic material, then collected by centrifugation (10,000 rpm, 10 min) and dried under vacuum conditions. FT-IR characterization was used to identify the functional groups associated with rhyolite particles after in vitro digestion.

In vivo studies

Among the different of particle sizes of rhyolite, Rhyolite-C showed the highest lipid-binding capacity with the strongest interaction with bile salts, and enhanced on digestive enzyme activity in the simulated gastrointestinal obesity model. Based on these results, and considering the established relationship between smaller particle size, higher specific surface area, and enhanced sorbent activity, Rhyolite-C was selected as the material candidate for in vivo evaluation in this study. An HFD containing 5 wt% of rhyolite-C (HFD+5% rhyolite-C-diet) was used against obesity in HFD mice. The HFDs (containing 2.2% dextrinized cornstarch, 60% kcal from fat, D12492; Research Diets, United States of America) contained the same amount of lard and soy oil as the primary source of fat in each diet. The food pellet was stored at −20 °C before use.

A total of 28 of 6-week-old male ICR mice were obtained from Japan SLC (Shizuoka, Japan). The mice were housed in a fixed cage and kept at a temperature of 22 °C ± 2 °C, relative humidity of 50% ± 10%, and a natural light-dark cycle (12-h light/12-h dark cycle). The mice were split into four groups and given different types of food for 12 weeks: (1) normal diets (AIN-93G, Oriental Yeast, Tokyo, Japan), (2) a high-fat diet (HFD), (3) a high-fat diet with 5% Rhyolite-C, and (4) a high-fat diet with 5% ABC for comparison. The total sample size for each group was set to 7 mice. Animal experiments were conducted according to Okayama University’s Guidelines for Animal Research, with approval from the Animal Care and Use Committee (Approval No. OKU-2022505). The study was obtained in compliance with ARRIVE guidelines. At the end of the treatment, after a 15-h fasting period, the blood obtained from the submandibular vein was collected using an anticoagulant tube (Terumo, Tokyo, Japan). It was promptly converted to plasma by centrifuging at 3,500 rpm for 15 min and stored at −80 °C. We sacrificed the mice using an overdose of isoflurane to collect the liver and colon for further analysis. The fresh feces were collected at the end of the treatment and immediately frozen to −80 °C prior to fecal microbiota analysis.

Plasma and liver biochemical parameters

Plasma blood collected was analyzed biochemically for alkaline phosphatase (ALP), and alanine aminotransferase (ALT), aspartate aminotransferase (AST), glucose (Glu), triglycerides (TG), high-density lipoprotein cholesterol (HDLC), total cholesterol (T-chol), free fatty acid (FFA) and total protein were determined using commercial assay kits and a biochemical analyzer (CHEM 7000, Fuji Film, Tokyo, Japan). The liver digestion for biochemical analysis was followed by Zhang et al. (2018). Briefly, the liver was weighed and treated with 1 mol/L NaOH (1:9 w/v) at 80 °C for 10 min, then neutralized (1 mol/L HCl) and incubated at 85 °C for 2 h, then re-neutralized using 5 mol/L NaOH. Subsequently, FFA, T-chol, and TG levels were determined as in the above methods.

Histopathological studies

The small intestine, colon, and liver were preserved in 4% paraformaldehyde (PFA) and then embedded in paraffin. Several 5 μm-thick tissue slices were sectioned and treated with hematoxylin and eosin (H&E) stain to enable morphological analysis. The photos were examined using an optical microscope (Nikon Co., Ltd., Tokyo, Japan).

Fecal microbiota analysis

DNA extraction

Prior to DNA extraction, frozen feces samples were thawed on ice, and stool samples in preservation tubes were agitated for 10 s. DNA was isolated from the fecal samples using the ZymoBIOMICS™ DNA/RNA Miniprep Kit (Zymo Research) according to the manufacturer’s instructions. DNA concentrations were quantified utilizing a Qubit™ dsDNA HS test kit in conjunction with a Qubit 4 Fluorometer (Thermo Fisher Scientific, Waltham, MA, United States). Importantly, microbial DNA was extracted from individual fecal samples collected from each mouse, and no pooling of samples was performed to ensure the alpha and beta diversity analysis indicated individual variability.

Amplicon sequencing of 16 S rRNA gene fragments

Partial 16 S rRNA genes (V4–V5 sections) were amplified by PCR with the primer sets 530F and 907R4 (Caporaso et al., 2012), which feature overhanging adapters at the 5′ends. PCR amplification, enzymatic purification, and incorporation of multiplexing indices and Illumina sequencing adapters, followed by purification using magnetic beads, were executed as previously outlined (Hirai et al., 2017). Amplicon sequencing was conducted with an Illumina MiSeq platform and MiSeq v3 reagent (Illumina, United States) for 300 bp paired-end reads, adhering to Illumina’s standard technique. FASTQ files produced by MiSeq underwent analysis through the QIIME2 pipeline (Bolyen et al., 2019). Pair-end raw FASTQ files were demultiplexed utilizing the demux plugin, according to their unique barcodes (Caporaso et al., 2010). Sequences from each sample were demultiplexed and processed with the dada2 plugin to generate a feature table (Callahan et al., 2016). The feature-classifier plugin was utilized to align feature sequences with a pre-trained SILVA-138 99% database, resulting in the generation of a taxonomy table (Bokulich et al., 2018). Processed files were visualized using QIIME2 View (https://view.qiime2.org).

PICRUSt2 (Douglas et al., 2020) was used to predict the functional profile of microorganisms in each microbiome, with functional annotation of gene sequences performed based on Kyoto Encyclopedia of Genes and Genomes (KEGG) orthology (https://www.kegg.jp/kegg/kegg1.html). The gene abundances of the microbiomes from the HFD, ABC, and Rhyolite groups were compared with those of the microbiomes from the control group. Genes with significant differences in the abundance among microbiomes were further analyzed using Welch’s t-test. Differences in the abundance were considered significant with a false discovery rate cut-off of 0.05. The differences in abundance of each gene were mapped onto the pathway map using KEGG Mapper Color (https://www.genome.jp/kegg/mapper/color.html). Genes exhibiting high (>2-fold) and low (<0.5-fold) abundance were colored red and blue, respectively.

Statistical analysis

Statistical analysis with one-way ANOVA and Tukey-Kramer tests was conducted using GraphPad Prism software (version 10.0; GraphPad Software, CA, United States of America). The alpha level was set at 0.05.

Physicochemical characterization and cytotoxicity of rhyolite particles

The physicochemical properties of rhyolite particles of various sizes (Rhyolite-A, Rhyolite-B, and Rhyolite-C) are shown in Figure 1. The scanning electron microscopy (SEM) image revealed that the milling process significantly affected the size of rhyolite particles, with irregular quartz microparticles being the most prevalent (Figure 1A). This observation may be related to the energy accumulated by collision and friction during the particle-milling treatment, which reduces the particle size of rhyolite and results in irregular shapes. The X-ray diffraction (XRD) pattern indicated that the mineral phases of the rhyolite were quartz, orthoclase, cristobalite, and sillimanite (Figure 1B). The intensity of the rhyolite peak at 26.42° decreased to 79.13% after milling, possibly due to the formation of an amorphous phase. Furthermore, the specific surface area increased as particle size decreased, as the surface parameter of rhyolite was related to the projected size of the corresponding particle. The smallest particle size (0.67 µm) was found in Rhyolite-C, corresponding to the highest specific surface area of 18.96 m 2 /g (Figure 1C).

Figure 1D shows the infrared (IR) spectra of Rhyolite-A, Rhyolite-B, and Rhyolite-C in the wavenumber range of 400–1,600 cm -1. The bands at 1,000–1,200 cm -1 are attributed to the antisymmetric stretching of Si-O vibrations. The Si-O bending and stretching vibrations in tetrahedra SiO 4 are ascribed to bands at approximately 450 cm -1 and 700–900 cm -1, respectively. The vibrational elongation of silanol (OH bonds) causes broad peaks at 3,100–3,540 cm -1 (Li et al., 2015). X-ray fluorescence (XRF) was used to analyze the primary metal oxides in the rhyolite (Supplementary Table S1), wherein SiO 2 had the highest concentration, followed by aluminum trioxide (Al 2 O 3), calcium oxide (CaO), sodium oxide (Na 2 O), and potassium oxide (K 2 O).

In addition to physicochemical characterization, in vitro cytotoxicity assays were performed using Caco-2 cells. Caco-2 cells were exposed to rhyolite particles at varying concentrations for 48 h and compared with ABC. Caco-2 cell viability showed particle size dependence of rhyolite at concentrations of 500 and 1,000 μg/mL (Supplementary Figure S3). Furthermore, rhyolite particles showed superior cell viability compared to ABC, indicating the absence of harmful effects associated with rhyolite particles. These findings showed the biocompatibility and safety of rhyolite.

In vitro digestion studies

The presence of rhyolite particles significantly altered nutrient digestion profiles in a particle size-dependent manner during stimulated fed-state gastrointestinal digestion (Figure 2). Lipid hydrolysis was decreased in all rhyolite-treated groups compared with the no-rhyolite control (Figures 2A,B), including free fatty acid and lipase activity at the end of the intestinal phase, regardless of their particle sizes. Starch hydrolysis followed a similar inhibitory pattern. Glucose release and the area-under-the-starch digestion curve (AUSC) were also significantly reduced by rhyolite, with smaller particles causing greater suppression (Supplementary Figure S2a). Correspondingly, amylase activity decreased (Supplementary Figure S2c), suggesting that rhyolite may impede enzyme–substrate interactions or substrate accessibility during carbohydrate hydrolysis.

To further investigate the particle size effect of rhyolite in vitro, emulsion formation was analyzed. It was shown that the smaller the rhyolite particle size, the smaller the emulsion size and the greater the number of droplets (Figures 2C,D). This suggests that smaller particle sizes have a higher specific surface area, which may allow them to capture more lipids. Furthermore, FT-IR spectroscopy of the precipitated rhyolite after intestinal digestion (Figure 2E) showed characteristic Si–O–Si stretching and bending vibrations, indicating that the core silicate structure remained intact. The peak shifts and intensity changes, specifically in the bands linked to surface hydroxyl groups and adsorbed organic matter, corresponding to C–H stretching (∼2,850–2,930 cm -1), C=O stretching (∼1,740 cm -1), and N–H bending (∼1,540 cm -1), indicating the presence of lipid and carbohydrate residues on the particle surface.

Taken together, these findings support complementary inhibitory mechanism, including (i) physical adsorption of the substrates onto the rhyolite surface, as evidenced by FT-IR results of adsorbed lipid and carbohydrate residues, reducing their aqueous-phase availability; (ii) adsorption of digestive enzymes or steric hindrance from particle–substrate complexes, impairing enzyme–substrate interactions, which is consistent with other aluminosilicate minerals (Sheng et al., 2022; Zimmerman et al., 2010; Chen et al., 2016). In addition, emulsification might affect lipid hydrolysis. Rhyolite-C was found to produce a large amount of emulsion with small particle size among Rhyolite-A and Rhyolite-B. The large amount of lipid trapping by many small rhyolite particles would be one of the factors suppressing lipid degradation.

Evaluation of rhyolite particles for suppressing body weight in HFD-induced obesity mice

Extensive studies have reported that an increased intake of lipids is directly associated with weight gain and an increased risk of obesity (Bray et al., 2002; Wang et al., 2020; Beulen et al., 2018; Nouri et al., 2023; Hall et al., 2019). In this study, we evaluated the effects of oral administration of rhyolite particles on body weight, plasma lipid regulation, and liver function in a HFD mouse model. In vivo studies showed that the HFD group exhibited substantial weight gain and increased obesity risk. In contrast, the HFD +5% Rhyolite-C diet and HFD +5% ABC diet significantly suppressed body weight gain compared to the control group (Figure 3A). No differences in food intake were observed among the groups (Figure 3B). However, 5% of the HFD was replaced with non-nutritive rhyolite, which reduced the caloric density of the diet from 4.03 to 3.83 kcal/g (Supplementary Table S2). This decrease in energy density raises the possibility that part of the observed reduction in weight gain may be attributable to caloric dilution rather than solely to the functional properties of rhyolite. The present study did not include an isocaloric pair-feeding control, in which control animals are restricted to consume the same caloric intake as the treated group. Future studies should incorporate isocaloric diet formulations or pair-feeding approaches to precisely isolate the metabolic effects of rhyolite from the physical effects of reduced caloric intake.

Furthermore, H&E staining revealed that the livers of the high-fat diet group showed significantly increased lipid accumulation compared to the control group, the high-fat diet +5% rhyolite C diet group, and the high-fat diet +5% ABC diet group, as the HFD group exhibited a large fatty liver-like appearance in its overall appearance (Figures 3C–E). H&E staining of the liver confirmed that rhyolite particles regulated lipid metabolism because fewer lipid droplets were observed in the liver with HFD treatment. This was likely linked to the fact that emulsification of particles in the digestive tract promotes the excretion of lipids from the body and promotes balanced fat metabolism. In other words, supplementing with rhyolite appears to prevent excessive lipid accumulation in the liver and maintain a healthy liver morphology by reducing the influx of excess lipids into the bloodstream.

H&E staining of the small intestine and colon was performed to investigate the potential pro-inflammatory effects of the rhyolite diet after oral administration for 12 weeks. Results showed a normal distribution morphology of crypts and a potentially protective effect in the rhyolite-diet groups compared to the HFD group (Figures 3F,G), suggesting that rhyolite particles may act as a physical barrier, limiting the direct exposure of the colonic epithelium to harmful lipid degradation products and free bile acids (Zheng et al., 2017), both of which were higher under HFD-fed conditions.

Evaluation of plasma and liver biochemical parameters

Plasma biochemical parameters were compared between the control group and the HFD groups with rhyolite or ABC diets (Figures 4A–I). Serum liver function markers (ALP, ALT, AST) increased in mice ingested with HFD. Lipid markers such as total cholesterol, triglycerides, and free fatty acids also increased. However, simultaneous supply of activated charcoal or Rhyolite-C to HFD-ingested mice significantly reduced these increases. Ultimately, no significant difference was observed between the normal values of the control mice and those of the Rhyolite-C-supplied mice. Similar results were obtained for total cholesterol, triglycerides, and free fatty acids in the liver (Figures 5A–C). Although ABC exhibited a larger specific surface area (SSA) than rhyolite particles (Supplementary Figure S1d), both ABC and rhyolite demonstrated equivalent efficacy in removing lipids from the plasma and liver. Moreover, rhyolite particles modulate hepatic triglyceride and cholesterol levels, potentially influencing lipid metabolism and enhancing lipid clearance pathways in the liver. While our findings suggest that rhyolite supplementation may influence lipid metabolism in the high-fat diet. The mechanistic interpretation remains preliminary. The observed changes in lipid parameters could be partially explained by caloric dilution, since the inclusion of rhyolite in the diet may have reduced overall energy density. We have therefore tempered the mechanistic claims, which require validation in controlled studies for caloric intake. Importantly, after 12 weeks of treatment with Rhyolite-C, no mortality was observed. The biochemical analysis showed a comparable effect of rhyolite supplementation to the control group during the treatment for 12 weeks, which supports the tolerability of rhyolite particles under the conditions tested. Moreover, silica, as the major component of rhyolite, has been applied via oral administration in murine models and has shown no mortality, with normal hematological and biochemical profiles (Li et al., 2026). Nevertheless, comprehensive safety characterization, including acute and chronic toxicity studies, will be essential before considering long-term oral administration for such purposes.

Fecal gut microbiota communities

HFD altered a characteristic shift in gut microbial composition, including an increase in Firmicutes and a reduction in Bacteroidota, resulting in an elevated Firmicutes / Bacteroidota (F/B) ratio, a microbial signature commonly associated with enhanced energy harvest and diet-induced obesity (Ley et al., 2006). Firmicutes are highly efficient at absorbing energy. In today’s dietary environment, they can easily store excess energy as fat, contributing to obesity. Whereas Bacteroidetes help suppress excessive nutrient absorption, contributing to weight loss and improvement of diabetes (Turnbaugh et al., 2006). In the HFD group, the gene abundance of Firmicutes increased significantly, and the gene abundance of Bacteroidota decreased compared to the other groups (Figures 6A,B). This shift resulted in an increased F/B ratio (Supplementary Figure S4). This also indicates that Firmicutes can efficiently store energy by breaking down bile salts, whereas Bacteroidota is more strongly affected by the pressure from bile acids. The environmental change was clearly reflected in the fecal microbiota, and this trend continued for some time in the ABC and rhyolite groups.

Furthermore, gut microbiota functions are also affected by HFD, specifically leading to gene expression responsible for fatty acid metabolism and digestion (Ley et al., 2006; Turnbaugh et al., 2006). The volcano plot showed that the HFD group had approximately 2,400 genes that showed significantly different abundances compared to the control group; however, the change in gene abundance was reduced to approximately 1,000 in the rhyolite diet groups (Figure 6C). Comparing the entire metabolic map, the HFD group exhibited an enrichment of genes associated with amino acid metabolism compared to the control group. Furthermore, within fatty acid metabolism, the HFD group was characterized by a higher prevalence of elongation pathways rather than degradation or biosynthesis. In contrast, the rhyolite diet group showed a higher abundance of genes related to amino sugar and nucleotide sugar metabolism, while displaying lower activity in both the degradation and biosynthesis of fatty acids (Supplementary Figure S5). The fecal microbiota composition indicates that rhyolite particles alter genes associated with fatty acid synthesis, reflecting a complex interaction between the host and microbiota. These findings indicate that dietary rhyolite particles significantly influence host metabolic regulation via interactions with the gut microbiota, affecting gene expression and physiological adaptation beyond nutrient absorption. Additionally, future work requires the investigations whether rhyolite alters gut-microbiota compositions and host-microbe interactions.

Inspired by the soil-eating behavior of wild animals, this study demonstrated that rhyolite particles exhibit novel functions as lipid regulators, potentially offering a promising strategy for obesity treatment. The rhyolite particles altered the composition of the gut microbiota and gene expression related to fatty acid metabolism, suggesting metabolic effects beyond nutrient binding. Rhyolite particles are a safe, effective, and environmentally friendly material, and are expected to have applications in biomedical fields, particularly in controlling dietary lipids and reducing the risk of obesity.