Introduction

Diabetes mellitus is a chronic metabolic disease with a globally increasing prevalence every year, making it a major global health issue1. In addition to affecting quality of life, diabetes significantly increases the risk of complications and affects patient prognosis2.

Recent systematic research has shown that short-chain fatty acids (SCFAs) such as butyrate, acetate, and propionate contribute to the improvement of blood glucose levels and dyslipidemia in diabetic mouse models, and their application for diabetes treatment is expected3. SCFAs are the primary products of the bacterial fermentation of dietary fiber, performed by gut microbes, that positively impact the host. In particular, acetate, the most abundant SCFA, may enhance GLP-1 and PYY secretion in the colon and improve metabolic health by modulating adipose tissue function, increasing insulin sensitivity and oxidative capacity, promoting satiety, and regulating insulin secretion. Hence, potential therapies to increase gut microbial fermentation and acetate production have been extensively investigated4.

Cellulose is a naturally occurring polysaccharide and is usually the main component of plant cell walls. Irrespective of species, ~50% of solid cell wall components of wood are cellulose. Hence, cellulose is one of the most abundant global organic resources. However, there are limitations to its utilization related to processing because cellulose is insoluble in water and other simple common organic solvents. One approach to addressing this processing issue of cellulose is chemical modification to make it soluble in water or a common solvent. Among such chemical modifications, an approach that has been widely implemented for decades commercially is cellulose acetylation, chemically turning cellulose into cellulose acetate (CA)5. Ordinary CA has not been utilized in the food industry because it lacks water solubility. Although not yet implemented commercially, water-soluble CA is achieved by controlling the degree of chemical modification (degree of acetylation) within a certain range. For cellulose, the degree of acetylation necessary to make it water soluble is roughly 15–30%, based on the hydroxyl groups available for acetylation, whereas that for ordinary CA is 80–95%6. Water-soluble CA with a degree of acetylation of 15–30% is sometimes called WSCA. WSCA, owing to its water solubility, could serve as a food additive. When hydrolyzed, WSCA leads to naturally occurring cellulose and acetate. The degradation of WSCA by enzyme systems (naturally occurring enzyme cocktails) expressed in wood cellulolytic fungus systems is well-studied7. These enzyme systems usually include acetyl esterases for naturally occurring acetylated polysaccharides such as xylan; these are also capable of hydrolyzing the acetyl groups in WSCA liberating acetate as a result.

Unlike conventional cellulose acetate, which is resistant to fermentation by the microbiota of ruminants and the human gastrointestinal tract, water-soluble cellulose acetate (WSCA) has been shown to increase acetate concentrations in human feces and rat cecal contents8,9,10. Therefore, WSCA may act as a prebiotic to modulate host nutritional physiology by increasing the acetate concentration in the intestinal tract of monogastric animals. However, the effects of WSCA on various diseases remain unclear. In this study, we aimed to investigate the effects of WSCA in treating glucose intolerance and body weight control in a mouse model of diabetes mellitus.

Results

WSCA suppressed weight gain, especially in db/db mice

Figure 1a shows the protocol of the current experiment. The mice on the control diet gained weight, whereas those fed the WSCA diet did not. Significant differences in body weight were observed between the groups, with more pronounced differences in the db/db mice (Fig. 1b). Visual comparison of body size at 16 weeks showed that WSCA intake suppressed a tendency towards obesity (Fig. 1c). To assess whether the anti-obesity effect of WSCA was due to reduced food intake, we monitored dietary consumption. Figure 1d shows the weekly changes in oral food intake. The amount of food consumed during the acclimatization period was attributed the value of 1.0. (Fig. 1d). The total amount of food consumed per mouse during the 8-week experiment period is shown in Fig. 1e. No evident difference in food intake was observed in the db/m group; however, the db/db WSCA group consumed significantly less food than the db/db control group.

Fig. 1: Study design and changes in body weight and dietary intake.
Fig. 1: Study design and changes in body weight and dietary intake.
Full size image

a Schema of the research. b Changes in body weight. c Body size appearance at 16 weeks of age. d Changes in oral intake during the observation period. e Daily food intake during the observation period. Error bars represent SEM. *p < 0.05, **p < 0.01, ***p < 0.001, as determined using one-way analysis of variance (ANOVA).

WSCA normalized blood glucose levels

The db/db control group showed a trend toward elevation of random blood glucose levels, whereas the db/db WSCA group showed normalization of blood glucose levels (Fig. 2a). In the glucose tolerance test, the WSCA group showed a lower blood glucose level at 120 min (Fig. 2b) and lower area under the curves (AUCs) in the oral glucose tolerance test (OGTT) than the control group (Fig. 2c) in db/m and db/db mice. Insulin levels (Fig. 2d), fasting insulin levels (Fig. 2e), and homeostatic model assessment for insulin resistance (HOMA-IR) (Fig. 2f) were measured simultaneously in the glucose tolerance test. The db/db WSCA group maintained additional insulin secretory capacity and normalized blood glucose levels. In addition, insulin resistance tended to improve in the db/db WSCA group, although the difference was not significant.

Fig. 2: Blood glucose and serum insulin levels.
Fig. 2: Blood glucose and serum insulin levels.
Full size image

a Causal blood glucose levels (n = 7). b Glucose tolerance test (blood glucose level). c Area under the curve analysis in the oral glucose tolerance test. d Glucose tolerance test (serum insulin level). e Fasting serum insulin levels. f HOMA-IR; (fasting blood glucose) × (fasting insulin)/22.5. Error bars represent SEM. *p < 0.05, **p < 0.01, ***p < 0.001, as determined using one-way ANOVA.

WSCA improved liver injury and lipid metabolism

Blood chemistry tests showed that the db/db-WSCA group had significantly lower levels of AST (58.8 IU/L vs. 76.5 IU/L), total cholesterol (113.2 mg/dL vs. 169.6 mg/dL), and glycated albumin (3.0% vs. 7.8%) compared to the db/db-control group, indicating an improvement in liver function and lipid metabolism (Table 1).

Table 1 Serum biochemistry results in db/m and db/db mice
Full size table

WSCA improved fatty liver and maintained muscle mass or grip strength

The macroscopic appearance and microscopic findings of the liver showed apparent differences in the size and percentage of fat droplets between the db/db control and db/db WSCA groups (Fig. 3a). Liver weight was significantly lower in the db/db WSCA group than in the db/db control group (Fig. 3b). NAS score, which indicates the histological severity of fatty liver11, was significantly lower in the db/db WSCA group than in the db/db control group (Fig. 3c). No differences in epididymal or brown fat content were observed (Fig. 3d, e). While the db/m WSCA and db/db WSCA groups showed reduction in body weight as shown in Fig. 1b, the plantaris or soleus muscle weight or grip strength did not decrease following WSCA administration (Fig. 3f–h).

Fig. 3: Fatty liver, weight of adipose tissue and skeletal muscle, and grip strength of mice.
Fig. 3: Fatty liver, weight of adipose tissue and skeletal muscle, and grip strength of mice.
Full size image

a Representative image of the macroscopic and histological findings in the liver. b Relative liver weight of 16-week-old mice (n = 7). c NAS score in hepatic histological appearance (n = 7). d Relative epididymal fat weight in 16-week-old mice (n = 7). e Relative brown fat weight of 16-week-old mice (n = 7). f Relative plantaris muscle weight of 16-week-old mice (n = 7). g Relative soleus muscle weight of 16-week-old mice (n = 7). h Relative grip strength of 16-week-old mice (n = 7). Error bars represent SEM. *p < 0.05, **p < 0.01, ***p < 0.001, as determined using one-way ANOVA.

WSCA increased the concentration of acetate in stool, cecum contents, and sera

Eight weeks after administration, the concentration of acetate significantly increased in the feces, cecum contents, and blood of db/m and db/db mice in the WSCA group compared with that in the control group (Fig. 4a, d, j). The concentration of propionate significantly increased in the feces of db/m mice in the WSCA group and showed a non-significant but increasing trend in db/db mice in the WSCA group (Fig. 4b). Regarding the concentration of propionate and butyrate in cecum contents, there was a rising trend in db/m and db/db mice in the WSCA groups (Fig. 4e, f).

Fig. 4: Acetate, propionate, and butyrate concentrations in the cecum, cecum contents, and sera at 8 weeks after the administration.
Fig. 4: Acetate, propionate, and butyrate concentrations in the cecum, cecum contents, and sera at 8 weeks after the administration.
Full size image

a–c Acetate, propionate, and butyrate concentrations in feces. (n = 7). d–f Acetate, propionate, and butyrate concentrations in the cecum. (n = 7). g–i Concentrations of serum acetate, propionate, and butyrate (n = 7). Error bars represent SEM. *p < 0.05, **p < 0.01, ***p < 0.001, as determined using one-way ANOVA.

Expression of Gpr41 and Gpr43 in the colon did not differ between the db/db control group and the db/db WSCA group

The relative mRNA expression of Gpr41 and Gpr43 in the colonic mucosa (normalized to the expression of actin) did not significantly differ between the db/db control and db/db WSCA groups (Fig. 5a, b).

Fig. 5: Expression of Gpr41 and Gpr43 in the colon, and serum GLP-1 and PYY levels at 8 weeks after the administration.
Fig. 5: Expression of Gpr41 and Gpr43 in the colon, and serum GLP-1 and PYY levels at 8 weeks after the administration.
Full size image

a, b Relative mRNA expression of Gpr41 and Gpr43 in the colon normalized to that of Gapdh (n = 7). c, d Serum concentrations of GLP-1 and PYY (n = 7). Error bars represent SEM. *p < 0.05, **p < 0.01, ***p < 0.001, as determined using one-way ANOVA.

Glucagon-like peptide-1 (GLP-1) and peptide YY (PYY) levels in the sera increased after WSCA administration

Despite unchanged receptor expression, serum GLP-1 and PYY levels in db/m and db/db mice in the WSCA group were significantly higher than those in the control group after eight weeks of administration. (Fig. 5c, d).

Discussion

This is the first study to show that WSCA improves obesity and glucose intolerance in a mouse model of diabetes mellitus. The db/db mice develop diabetes due to overeating caused by their abnormal leptin receptors12. WSCA suppressed overeating and normalized body weight in db/db mice. In addition, glucose intolerance, dyslipidemia, and fatty liver observed in db/db mice improved following WSCA intake. The importance of regulating appetite in the treatment of diabetes and obesity caused by overeating has been attracting attention.

One of the most researched topics in this context is acetate role – a major SCFA. It has been shown that acetate in the large intestine is absorbed into the bloodstream, crosses the blood-brain barrier, and is directly involved in appetite regulation in the central nervous system, mainly in the hypothalamus13. It is also known that acetate in the large intestine promotes the secretion of anorectic hormones such as PYY and GLP-1 from large intestinal L cells and leptin from adipocytes via GPR41 and GPR43 expressed in the intestinal epithelium14. It has been well known that GPRs are expressed in various tissues and influence many important metabolic mechanisms that maintain energy homeostasis. In particular, GPR41 and GPR43 can be activated by SCFAs and have been targeted in therapeutic strategies for the treatment of metabolic disorders including15,16.

In this investigation, although no significant changes were observed in colonic GPR41 and GPR43 expression, WSCA administration increased serum GLP-1 and PYY levels. This suggests that mechanisms beyond transcriptional regulation of SCFA receptors may be involved. SCFAs, particularly acetate and propionate, may directly stimulate L-cells via receptor-independent pathways or act through pre-existing GPR41/43 proteins at the post-transcriptional level. Previous studies support SCFA-induced GLP-1 secretion without changes in Gpr41/43 expression17,18. Alternatively, SCFAs may activate the gut–brain axis via vagal afferents, indirectly enhancing GLP-1 secretion13,19. Epigenetic or metabolic modulation of enteroendocrine function has also been proposed20. These findings suggest that WSCA may promote incretin secretion through multiple pathways, including direct effects on L-cells and neural mechanisms.

As such, it is known that acetate regulates appetite not only via central nervous system but also through gastrointestinal hormones. In the present study, it was found that WSCA increased the concentration of acetate in the intestinal tract and increased serum GLP-1 and PYY. In addition, since an increase in serum acetate concentration was also observed following WSCA intake, it is expected that the central nervous system’s appetite control mechanism was also implicated, resulting in the normalization of the amount of food intake in db/db mice.

In type 2 diabetes, insulin resistance and chronic inflammation are known to accelerate the loss of skeletal muscle mass and strength, and decreased grip strength is considered a key risk marker for sarcopenia and frailty21,22. The observed preservation of muscle mass and grip strength in WSCA-treated mice suggests that WSCA may help prevent diabetes-associated muscle dysfunction23. Improved hepatic steatosis and insulin sensitivity, along with increased levels of SCFAs such as acetate, may contribute to these beneficial effects by enhancing muscle metabolism and exerting anti-inflammatory actions.

The WSCA examined in this study has the potential to be used as a food additive given its physical properties7. Although the involved enzyme systems have not been fully elucidated, the metabolism of WSCA in the digestive systems of ruminants and rodents was investigated. Watabe et al. reported through that WSCA is metabolized by ruminal bacteria cultures in vitro to generate acetate while increasing the proportion of the genus Prevotella in cultures10. Genda et al. reported that acetate, succinate, and propionate are increased in the cecum of rats fed with WSCA while increasing the proportion of Bacteroides xylanisolvens belonging to the order Bacteroidales in gut microbiome9. Strobel reported that Prevotella ruminicola belonging to the Bacteroidales order is capable of metabolizing succinate to propionate24. Taken together, the evidence by Genda and Strobel suggests that WSCA is metabolized to propionate in the gut of rats by bacteria such as those belonging to the order Bacteroidales via a pathway involving cellulose, glucose, phosphoenol pyruvate (PEP), and succinate. In other words, propionate generation from WSCA in the gut of rats suggests that WSCA is subject to enzymatic hydrolysis in the gut to regenerate cellulose. Although we have not analyzed the intestinal microbiota involved in this study, we confirmed the increase in propionate as well as acetate in the feces in db/m mice, and our results were consistent with the aforementioned hypothesis based on Genda’s and Strobel’s findings. The results of this study suggest that acetate is mainly involved, but propionate may also be involved.

WSCA metabolism was also investigated from human stool cultures. Yamada et al. reported that WSCA is metabolized to acetate and propionate in human stool cultures while increasing Bacteroides uniformis, and that Bacteroides uniformis can grow in pure cultures supplemented with WSCA8. Reports on the relationship between WSCA and the intestinal microbiota have been increasingly published. Rodent trials also suggested that WSCA might be beneficial to gut immunity25, cholesterolhemia9, non-alcoholic steatohepatitis (NASH)26, and moderate bodyweight gain9, in line with our findings.

One limitation of this study is that we were unable to assess the postprandial dynamics of GLP-1 and PYY levels during the oral glucose tolerance test (OGTT), which may have provided further insights into the metabolic responses. Another important limitation is the absence of gut microbiota analysis. Although WSCA administration led to increased levels of short-chain fatty acids, we could not directly determine whether these changes were associated with alterations in microbial composition. Future studies incorporating 16S rRNA sequencing or metagenomic profiling are warranted to clarify the relationship between WSCA, gut microbiota, and SCFA production.

In conclusion, WSCA contributes to host health by increasing the concentration of acetate in the large intestine in diabetic mice. Previous systematic search has shown that SCFAs such as butyrate, acetate, and propionate contribute to the improvements in blood glucose levels and dyslipidemia in diabetic mice models; thus, their application for the treatment of diabetes is expected3. In the future, it is expected that the safety of WSCA for humans will be verified and its potential application for the prevention and treatment of diabetes will be explored.

Methods

Animals

Seven-week-old male C57BLS/J lar- m + /+Lepr db (heterozygous, db/m) and C57BLKS/J lar- +Lepr db/+Lepr db (homozygous, db/db) mice were obtained from Shimizu Laboratory Supplies (Kyoto, Japan). The mice were maintained at 40–70% relative humidity, 18–24 °C, and a 12-h light/dark cycle. They were allowed free access to water and food for 1 week during the acclimation period. Experiments were conducted in accordance with the guidelines of the National Institute of Health for the use of animals in research. The Kyoto Prefectural University of Medicine Animal Care Committee approved all experimental protocols (permission numbers M2023-41).

For 8 weeks, the mice were fed either a control diet (control diet; AIN-93G, 361.2 kcal/100 g, fat kcal 7.0% soybean oil, cellulose 5%/w; Oriental Bio Service, Kyoto, Japan) or a modified AIN-93G rodent diet in which 1% cellulose was replaced with WSCA (Table 2). The WSCA used in this experiment was provided by Daicel Corporation (Osaka, Japan), and its properties are shown in Supplementary Fig. 1. WSCA with degree of acetylation of 21% and weight average molar mass of 32,600 g mol-1) was prepared from ordinary cellulose acetate (L-70, Daicel Corporation, degree of acetylation of 82%) by acid-catalyzed partial hydrolysis at 90 °C in acetic acid/water/sulfuric acid system followed by neutralization of sulfuric acid, precipitation and washing with methanol, and drying in accordance with the method reported by Ukita et al. (US 10703825 B2). Oriental Bio Service mixed WSCA with AIN-93G to create a solid rodent food upon our request. Twenty-eight db/m and db/db mice were assigned to four groups: (1) db/m control, (2) db/m WSCA, (3) db/db control, and (4) db/db WSCA. The cages used were 320 mm long,

Table 2 Composition of the test diets (%)
Full size table

The amount of food intake in each group was measured at the same time each week and the food was replaced with fresh food for 9 weeks. At 15 weeks of age, all the mice were sacrificed under isoflurane anesthesia after 16 h of overnight fasting (Fig. 1a).

Analytical procedure for blood glucose level estimation and OGTT

Blood was drawn from the tail vein and random blood glucose levels were measured using a glucometer (Glutest Mint; Sanwa Kagaku Kenkyusho, Nagoya, Japan) at the same time each week under ad libitum feeding conditions. The OGTT was performed after 16 h of fasting, and 1 g glucose per kilogram of body weight was administered orally by needle feeding at 15 weeks of age. Blood glucose levels were measured at baseline and at 30, 60, and 120 min after administration. The AUC in the OGTT was analyzed.

Measurement of grip strength

The grip strength of 15-week-old mice was measured using a strain gauge (GPM-100; Melquest, Toyama, Japan). Grip strength measurements were taken 10 consecutive times at 1-min intervals over 2 days, and the investigators were blinded to the experimental conditions.

Biochemistry

Blood samples were taken from fasted mice and serum samples were collected after centrifugation at 900 g for 15 min at 17 °C. The level of AST and alanine aminotransferase were measured using a standardization support method of the Japanese Society for Clinical Chemistry27. The levels of total cholesterol28, triglycerides29, and non-esterified fatty acids30 were measured using the enzymatic method of FUJIFILM Wako Pure Chemical Corporation (Osaka, Japan). Low-density lipoprotein cholesterol31 levels were measured directly by Sekisui Medical (Tokyo, Japan). Glycated albumin level was measured using the enzymatic and BCG methods by Asahi Kasei Pharma (Tokyo, Japan) and FUJIFILM Wako Pure Chemical Corporation (Osaka, Japan).

Liver histology

The livers of the sacrificed mice were fixed in 10% buffered formaldehyde and embedded in paraffin. Sections were prepared and stained with hematoxylin and eosin (HE). To evaluate fatty liver, microscopic observations were performed at 40× and 200× and evaluated according to the NAS score11.

Measurement of organ weights

The livers, epididymal fat, brown fat, and plantaris and soleus muscles of the mice were removed immediately after they were sacrificed. The weight of each organ per unit of body weight during dissection was evaluated.

Measurement of SCFA levels in the feces and cecum contents

A portion of the fecal or cecal content was collected, weighed, and suspended in perchloric acid (0.5 mL) to remove the protein. After centrifugation at 10,000 × g for 5 min at 4 °C, the resulting supernatant was filtered through a cellulose acetate membrane filter with a pore size of 0.45 μm. Organic acid content was analyzed using ion-exclusion high-performance liquid chromatography. The amounts of organic acid substances per unit weight of feces and cecum were measured32.

Measurement of SCFA levels in sera

Serum contents were immediately added to 5-sulfosalicylic acid, followed by vortexing for 1 min. The samples were centrifuged at 15,000 × g for 15 min and the supernatant was collected. The supernatant was mixed with 2-ethylbutyric acid (internal control), hydrochloric acid, and diethyl ether and vortexed for 1 min. The samples were centrifuged at 3000 × g for 5 min, and the SCFA-containing ether layers were collected and pooled for gas chromatography-mass spectrometry analysis using a GCMS-QP2010 Ultra instrument (Shimadzu). The VF-WAXms (30 m × 0.25 mm internal diameter × 1 μm; Agilent Technologies) was used for chromatographic separation. Helium (0.92 mL/min) was used as the carrier gas. The mass spectrometer was set to scan mode from m/z 40 − 130 and to the selected ion monitoring mode at m/z 60 (retention time: 9.6) for acetate, m/z 74 (retention time: 10.7) for propionate, and m/z 60 (retention time: 12.5) for n-butyrate. The concentration of SCFAs in each sample was determined using an external standard calibration curve over a specified concentration range32.

Gene expression analysis in murine colons

Total RNA was isolated using the acid guanidinium phenol-chloroform method with TRIzol reagent (Thermo Fisher Scientific USA), according to the manufacturer’s instructions as previously described33. The resultant cDNA was used for quantitative reverse transcription polymerase chain reaction (qRT-PCR) using specific primers: GPCR41, 5′-TGGCTTTTCTTTTCCGTCTACCT-3′ and antisense 5′-AAGACCACCAGGGCCATCA-3′; GPCR43, 5′-GATGTGGTACTGCCCGTACGA-3′ and antisense 5′-GACTGCCATGGGAACGAAAA-3′; actin, 5′-TATCCACCTTCCAGCAGATGT-3′ and antisense 5′-AGCTCAGTAACAGTCCGCCTA-3′. PCR was performed using PowerUp SYBR Green PCR master mix and QuantStudio 6 Pro real-time PCR system (Thermo Fisher Scientific). The PCR conditions included 40 cycles at 95 °C for 15 s and primer annealing at 60 °C for 1 min, with a subsequent melting curve analysis in which the temperature was increased from 60 °C to 95 °C. Gene expression levels were calculated from the quantitative reverse transcription-PCR data and normalized to actin expression.

Measurement of serum GLP-1 and PYY levels

Serum samples were collected from the sacrificed mice and processed as described in the “Biochemistry” section. Serum GLP-1 and PYY levels were quantified using a GLP-1 enzyme-linked immunosorbent assay (ELISA) kit (FUJIFILM Wako Pure Chemical Corporation) and a PYY ELISA kit (FUJIFILM Wako Pure Chemical Corporation).

Statistical analysis

Analysis of variance (ANOVA) was performed to assess the trend of the mean stratified according to normally distributed continuous variables. All analyses were performed using JMP PRO version 14.0.0 (SAS Institute Japan Ltd). Differences were considered statistically significant at p < 0.05.