Tirzepatide, a dual GLP-1 and GIP receptor agonist, promotes bone loss in obese mice via gut microbial-related metabolites

Abstract

Background

As a novel dual glucose-dependent insulinotropic polypeptide (GIP) and glucagon-like peptide 1 (GLP-1) receptor agonist, Tirzepatide (TZP) is a recently approved medication for treating type 2 diabetes mellitus (T2DM) and obesity; however, the effect of TZP in bone remodeling remains unclear.

Methods

1. The effect of Tirzepatide on osteoblasts and osteoclasts was observed by inducing differentiation of bone marrow mesenchymal cells (BMSCs) in vitro. 2. Db/db mice were used as a pathological model to investigate the role of TZP on bone metabolism. After TZP intervention, the feces in the intestinal tract of mice were collected for 16s rRNA gene sequencing to select the candidate gut microbiota most related to bone mass, and the effects of gut microbiota on bone metabolism were verified through subsequent microbiota supplementation experiments. 3. Metabolomics was used to analyze the difference of fecal metabolites between mice with the candidate microbiota supplement and those without, and the effect of candidate metabolites on bone metabolism was verified by the in vitro intervention of differential metabolites in BMSCs induction differentiation experiments.

Results

We found that TZP intervention resulted in a significant decrease in bone mass accrual in vivo. TZP was not indispensable to the differentiation of osteoblasts and osteoclasts in vitro. Bone and fat homeostasis were modulated by gut microbiota. We further demonstrated that the biodiversity of the gut microbiota in db/db mice was strikingly altered after TZP treatment. Lachnospiraceae, a key pro-osteogenic component of gut microbiota was significantly reduced. As a main metabolite of Lachnospiraceae, evodiamine played a role in suppressing osteoclastogenesis in vitro. Based on this, the transplantation of the Lachnospiraceae effectively ameliorated bone loss that was seen in db/db mice due to TZP treatment.

Conclusion

TZP administration leads to bone loss in the context of diabetes and obesity, and targeting the composition of gut microbiota may provide a potential way to protect bone health in type 2 diabetic patients treating with TZP.

The translational potential of this article

This study indicates that TZP has a negative impact on bone mass, suggesting that clinical attention should be paid to the risk of further decline in bone mass after Tirzepatide treatment, and it is necessary to follow up on their bone metabolism. Additionally, the gut microbiota plays an important role in bone metabolism regulation, and supplementing with certain probiotics may have a preventive effect on bone mass reduction associated with TZP treatment. Our research provides a reference for the prevention and treatment of drug-related osteoporosis in patients with T2DM in the future.

Graphical abstract

Fig. 8. Graphic abstract of the effect of gut microbial-related metabolites on bone loss due to Tirzepatide. Tirzepatide administration leads to bone loss in the context of diabetes and obesity. Transplantation of the Lachnospiraceae effectively ameliorated bone loss through a metametabolite, evodiamine, which suppresses osteoclastogenesis.

1. Introduction

Diabetes is a very prevalent chronic disease affecting nearly 500 million people worldwide, in which 90 % of patients with type 2 diabetes mellitus (T2DM) [1]. Diabetes and its complications impose a heavy economic burden on society and individuals. It is worth noting that T2DM is a complex disease that is often accompanied by obesity, fatty liver, osteoporosis, and other metabolic diseases, which brings enormous challenges to the treatment of diabetes[[2], [3], [4]]. Obesity is one of the important risk factors of T2DM. On the one hand, excessive accumulation of fat in the body leads to insulin resistance, which increases the risk of T2DM. On the other hand, a number of factors, such as insulin treatment, can in turn cause weight gain, the two form a vicious circle [5,6]. Therefore, an integrated therapy that takes account of both glycemic control and weight loss may be a very beneficial method to treat T2DM. However, previous drugs often only focused on the single effect of either hypoglycemia or weight loss, with certain limitations. For example, weight-loss drugs such as orlistat have finite efficacy and more side effects [7]. Although metformin has been shown to improve insulin resistance and weight control, its weak curative power and side effects, such as gastrointestinal reactions, limit its clinical use. In recent years, a series of selective glucagon-like peptide 1 receptor (GLP-1R) agonists developed based on incretin, such as liraglutide and semaglutide, have shown certain advantages in lowering blood glucose and reducing body weight, but rebound after withdrawal is a common problem [8]. Therefore, the development of safe and effective drugs for the treatment of T2DM and obesity remains an important area of current medical research.

Osteoporosis is a common skeletal complication in patients with T2DM. A large body of evidence suggests that diabetes negatively affects bone strength and increases fracture risk [9,10], and that a longer duration of diabetes with poorer glycemic control is associated with higher fracture risk [11]. In addition, antidiabetic drugs used in the long-term treatment of diabetes also affect bone metabolism [12]. For example, the insulin sensitizer rosiglitazone is usually believed to increase the risk of osteoporosis and fractures[[13], [14], [15], [16]]. Canagliflozin, an inhibitor of the sodium-glucose cotransporter 2 (SGLT2i), has been shown to deteriorate femoral microstructural properties in murine models [17]. Conversely, dapagliflozin, another SGLT2i, has been demonstrated to be devoid of any impact on indices of bone formation and resorption, as well as fracture rates, among diabetic patients with mild renal dysfunction [18]. Metformin, sulfonylureas, and dipeptidyl peptidase-4 (DPP4) inhibitors are predominantly viewed as exerting a neutral influence on bone mass [19]. The influence of antidiabetic medications on bone density is intricately linked to the particular drug in question. Body mass is also a significant determinant of bone metabolic processes, and there is a growing body of evidence suggesting that individuals who are overweight or obese are at an increased risk of experiencing fractures [20,21]. In obesity, there is an accumulation of white adipose tissue, which can stimulate the release of inflammatory cytokines, leading to the degradation of bone microarchitecture. Additionally, this condition can drive the differentiation of bone marrow mesenchymal stem cells (BMSCs) towards adipogenesis rather than osteogenesis, consequently decreasing the osteoblast population and leading to an overabundance of bone marrow adipocytes. This shift can upset the balance of bone turnover [22]. Therefore, it is necessary to carefully assess the impact of drugs on bone metabolism, especially new drugs in the treatment of T2DM and obesity.

Tirzepatide (TZP) is the world's first marketed GLP-1/GIP dual receptor agonist, officially approved by the US Food and Drug Administration for the treatment of diabetes in May 2022. TZP is composed of 39 amino acids including GLP-1 and GIP homologous fragments, simultaneously modify part of the noncoding amino acid residues to make it easier to bind to plasma albumin, protect it from rapid recognition and degradation by DPP4, and extend its half-life to 116.7 h [23]. The efficacy and safety of TZP in people with T2DM have been extensively studied in the global phase 3 SURPASS program[[24], [25], [26], [27], [28]]. Among them, the SURPASS-2 study compared once-weekly subcutaneous injection of TZP versus semaglutide, which is currently the strongest glucose-lowering GLP-1R agonist, and the results showed that the HbA1c was decreased by 0.15 % (P = 0.02), 0.39 % (P < 0.001) and 0.45 % (P < 0.001) in TZP 5 mg, 10 mg and 15 mg groups versus semaglutide 1 mg group, respectively[25]. Moreover, SURMOUNT-1 [29] and SURMOUNT-CN [30] programs demonstrated the efficacy and safety of TZP for weight loss in obese (non-diabetic) patients. Based on its remarkable efficacy and good tolerance, TZP tends to have broad application prospects in the field of diabetes and obesity treatment. However, as a new type of hypoglycemic drug, the impact of TZP on bone metabolism is currently unknown. Both GLP-1 and GIP receptors are expressed in bone tissue. Although some animal experiments hint at the potential for GLP-1 and GIP to stimulate osteogenesis and suppress osteoclastic bone resorption[[31], [32], [33]], the population studies have shown conflicting results, with a meta-analysis showing that the use of GLP-1 receptor agonists is not associated with fracture risk, but in subgroup analyses, exenatide is shown to increase fracture risk [34]. Since GIP alone cannot reduce blood glucose [35], there is no single patent medicine available for clinical treatment. Therefore, the specific effects of TZP on bone metabolism in individuals with diabetes and/or obesity remain to be elucidated.

The intestinal flora is also called the "Second genome", and its dynamic equilibrium is crucial for maintaining good health [36]. Once the homeostasis is broken, the disturbance of the intestinal flora will cause a variety of body dysfunctions, such as impaired intestinal barrier, intestinal inflammation, and immune function imbalance, thus will induce a variety of diseases, including obesity, diabetes, osteoporosis, etc [[37], [38], [39], [40]]. In recent years, it has been found that intestinal flora plays a very important role in bone mass regulation and the repair of bone diseases such as osteoporosis. The concept of “gut–bone axis” has also attracted wide attention. Intestinal bacteria can enhance the function of osteoblasts, improve bone homeostasis, balance bone metabolism, and promote bone development by promoting the absorption of nutrients (such as calcium, phosphate, etc.) that are conducive to bone development, and increase the production of serotonin and vitamin D related to bone development [41,42]. At present, as beneficial gut metabolites that have been widely reported, short-chain fatty acids can enhance bone mass by inducing metabolic reprogramming of osteoclasts and down-regulating osteoclast gene expression [43]. Bone mass changes in mice lacking gut microbiome were also reversed by gut recolonization [44]. In short, intestinal flora plays an important role in regulating bone health[41,42,[45], [46], [47]].

In this study, we selected the db/db mouse model with obesity and diabetes characteristics as our research object to verify the hypoglycemic and weight-loss efficacy of TZP. Meanwhile, we will evaluated the potential effects of TZP on bone metabolism and the role of intestinal flora in it, and further revealed the mechanisms by which TZP affects bone metabolism. This study aims to provide a reference for the prevention and treatment of osteoporosis in patients with T2DM and obesity.

2. Materials and methods

2.1. Experimental animals

Wild-type C57BL/6J female mice (3–4 weeks old) were obtained from the Experimental Animal Center of Xiamen University. All animal experimental research complies with the animal ethics regulations of Xiamen University (Ethics number: XMULAC20190084). Db/db female mice (6–7 weeks old) were purchased from Hangzhou Ziyuan Company. All animal experimental research complied with the animal ethics regulations of Xiamen University (Ethics number: XMULAC20220132). After 1 week of acclimatization, 8 db/db mice in the experimental group were subcutaneously injected with 70 nmol/L TZP for 1 month, and mice in the corresponding control group received a PBS injection for 1 month. During the experiment, the body weight, food intake, and glucose level of mice were monitored every 3 days.

2.2. Osteoblast culture and alkaline phosphatase (ALP) analysis

Primary BMSCs were isolated from 21-day-old mice. Cells were cultured in α-MEM medium (Gibco) containing 10 % Fetal Bovine Serum (FBS), 2 mM l-glutamine, 1 % penicillin/streptomycin, 1 % HEPES, and 1 % non-essential amino acids, and differentiated with ascorbic acid and β-glycerophosphate. For ALP activity, osteoblasts were incubated with a tenfold diluted alamar blue solution, washed, and incubated with a solution containing 6.5 mM Na₂CO₃, 18.5 mM NaHCO₃, 2 mM MgCl_2, and phosphatase substrate (Sigma–Aldrich). ALP activity was measured by a spectrophotometer (Thermo).

2.3. Alizarin red (AR) staining of osteoblast differentiation

Firstly, 2 % alizarin red solution was adjusted to pH 4.2. On the 21st day after the osteogenic differentiation of cells, the induction medium was discarded. Cells were fixed with 4 % PFA for 30 min, washed twice with 1 × PBS buffer, and subsequently washed three times with 70 % ethanol. Then, cells were stained with Alizarin Red.

2.4. Osteoclast culture and tartrate-resistant phosphatase (Trap) staining

Murine BMM cells were flushed from the femur and tibia of mice and then cultured in petri dishes in α-MEM medium with 10 % FBS and 20 ng ml⁻¹ M-CSF. The osteoclast precursors were then differentiated into osteoclasts in the presence of RANKL (50 ng ml⁻¹; R&D) and M-CSF for 5–6 d for TRAP staining. After discarding the culture medium, cells were washed twice and fixed with prepared 4 % paraformaldehyde (PFA) for 20 min. Subsequently, the prepared Trap dye solution was added and placed in a 37 °C incubator for 30 min. The cells were observed every 10 min. When the color turns pink, the Trap dye was removed and washed twice with PBS. Finally, the 96-well plate was placed under a microscope for observation and counting the number of osteoclasts in each well.

2.5. μCT analysis

The Micro-CT instrument platform model used in this experiment is Bruker SkyScan 1272. The mouse femur was fixed in a matching tube and placed in the Micro-CT instrument for fixation. The specific scanning parameters were set: voltage 60 kV, resolution 10 mm, pixel size 2016 Scan after × 1344, and AI value 0.25 mm. The main analysis parameters of trabecular bone were BV/TV, Tb. Th, Tb. N, Tb.Sp. The main analysis parameter of cortical bone was Ct.Th.

2.6. Histology and histomorphometry

In brief, the processed vertebral samples of mice were fixed in 4 % PFA for histomorphometry, and undecalcified sections of the lumbar vertebrae were stained using Von Kossa, Toluidine blue staining, and TRAP as previously described [48].

2.7. ELISA analysis

Serum from db/db mice was obtained by centrifuging plasma at 12,000 rpm for 15 min at 4 °C, then the serum was stored at −80 °C. All ELISA assays, including CTX1 (Elabscience, E-EL-M3023) and P1NP (Cloud-clone, CEA957Mu), were run according to the manufacturer's instructions. It was noted that the serum needs to be thawed slowly in a 4 °C refrigerator in advance to avoid creating bubbles during the experiment.

2.8. Lachnospiraceae administration

The bacterial taxa termed “Lachnospiraceae” in this study were a mixture of 17 strains that belong to the family Lachnospiraceae as previously reported [49,50] unless otherwise noted. These bacteria were cultured in an anaerobic chamber in BHI broth supplemented with 5 % fetal bovine serum, 0.01 % L-cysteine, and 1 % corn starch. The recipient mice were treated with Abx for 5 days, and then subjected to oral gavage with BHI containing 2.6 × 10⁸ Lachnospiraceae (all 17 strains were used in equal abundance) colony-forming units (CFU) with 3 times weekly for 6 weeks, and the control mice received an equal amount of colony-free BHI.

2.9. Fecal DNA extraction, 16S rRNA gene sequencing, and data analysis

After a 6-week TZP intervention, the fecal samples were collected into 1.5 mL sterile cryopreservation tubes and stored at −80 °C. Fecal DNA samples were amplified by polymerase chain reaction (PCR) using bar-coded primer pairs targeting the V3–V4 region of the 16S rRNA gene. PCR amplicons were sequenced using the Illumina sequencer. α-diversity was conducted to assess the complexity of species diversity of each sample using QIIME2 software. β-diversity was also analyzed using QIIME2 software by the PCoA analysis to examine the diversity in samples between the TZP and control groups. First, the reads were against a reference collection, GreenGenes database, May 2013 version, the closed-reference OTU table was built using QIIME. And the resulting OTU table was normalized by normalize_by_copy_number.py. The statistical difference analysis was determined using ANOVA. The results were visualized using a custom R script based on ggplot2. The raw data of 16S rRNA gene sequencing have been deposited to the NCBI Sequence Read Archive (https://submit.ncbi.nlm.nih.gov/subs/sra/) under the accession numbers SUB15388544.

2.10. Untargeted metabolomics

The fecal samples were placed in EP tubes and 300 μL of 80 % methanol solution was added. Samples were put into liquid nitrogen for quick freezing, melted on ice for 5 min, then vortexed for 30 s, and sonicated for 6 min. The mixture was centrifuged at a speed of 5000 rpm at 4 °C for 1 min. After centrifugation, the supernatant was transferred to another centrifuge tube and freeze-dried into powder. Then, each sample was dissolved in an appropriate volume of 10 % methanol and analyzed with an LC-MS instrument. Samples of equal volume were taken from each experimental sample and could be used as QC samples after mixing. The blank sample was 53 % methanol solution. Then, the 96-well plates were sealed for the high-performance liquid chromatography/mass spectrometry analysis using a Q-Exactive high-resolution tandem mass spectrometer (Thermo Fisher Scientific, Waltham, MA, USA). Using CD 3.3 software, the data (.raw) was retrieved and compared with mzVault, mzCloud, and Masslist to obtain the identification and relative quantitative results of metabolites. The data processing part was based on the Linux operating system (CentOS version 6.6) and the software R and Python. LIPIDMaps63, KEGG64, and HMDB65 were used to analyze the metabolites identified above. The metabolites were annotated by matching their molecular weights (p-value<0.05, VIP>1, and fold change (FC)≥2 or FC ≤ 0.5) and were subjected to the metabolite set enrichment analysis based on the KEGG database.

2.11. Statistical analysis

All data were statistically analyzed using GraphPad Prism 9.0. Data were expressed as mean ± standard deviation. A t-test was used for differences between mice in two experimental groups and one-way analysis of variance (ANOVA) was used for comparisons of mice in three or more experimental groups. Mice that died during the experiment were removed. p < 0.05 was considered statistically significant.

3. Results

3.1. TZP lowers blood glucose and bone density in db/db mice

We monitored and compared the changes in body weight of the mice, and an oral glucose tolerance test (OGTT) was performed on the 30th day. The experimental results showed that the body weight in the TZP group was significantly lower than that of the control group (Fig. 1A); there was no significant difference in intergroup comparisons of blood glucose between the TZP group and the control group at baseline; however, blood glucose in the mice was significantly reduced after the TZP intervention (Fig. 1B); and the results of the OGTT assay indicated that the fasting glucose in the TZP group and the glucose values at all time points after glucose loading were lower than those of the control group (Fig. 1C and D, p < 0.01).

Fig. 1.

Effects of TZP on blood glucose and bone density in db/db mice. (A) Weight change of db/db mice between the control group and TZP-treated group 42 days after treatment (N = 8–9 per group). (B) Random blood glucose changes between control and TZP-treated mice 42 days after treatment. (C) Fasting blood glucose change after 6-week TZP intervention in control and experimental mice. (D) The OGTT test in the control mice and TZP-treated mice. The db/db mice were fasted for 10–12 (overnight) hours before the experiment. For OGTT, fasted mice were orally gavaged with 20 % (weight/volume) glucose solution. Glucose levels were measured using a glucometer at 0, 15, 30, 60, 90, and 120 min. (E) 3D representation of femur Micro-CT trabecular and cortical bone after 3D reconstruction in db/db mice after 6 weeks of TZP intervention (scale bar: 100 μm) n = 8. (F) Bone trabecular bone volume and other parameters of bone mineral density after 6 weeks of drug administration.Grey dots represent control group, blue squares represent tirzepatide group. Statistical plots of relevant bone density parameters, from left to right and top to bottom, relative bone volume (BV/TV), cortical bone thickness (Ct. Th), trabecular bone thickness (Tb. Th), trabecular bone number (Tb. N), and trabecular bone separation (Tb. Sp). ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ns p ≥ 0.05. n = 8.

To test the effect of TZP on bone mass in db/db mice, we performed Micro-CT scanning of femoral bone samples to analyze the data on bone mass after continuous TZP injection (70 nmol/kg/day) for 6 weeks. The db/db mice showed a significant decrease in femoral trabecular bone volume, relative bone volume (BV/TV), and trabecular bone thickness (Tb. Th), compared with the control group (p < 0.05), while there was no significant difference in cortical bone thickness (Ct.Th) between the two groups (Fig. 1E and F).

3.2. TZP exhibits negligible effects on the formation of osteoblasts and osteoclasts from BMSCs

To examine the effect of TZP on osteoblast differentiation, we treated the osteoblast cell line MC3T3-E1 in vitro with different concentrations of TZP (10 nM and 100 nM). In addition to that, BMSCs from 3-week-old wild-type mice were isolated to detect osteoblast mineralization nodules using alizarin red (AR) staining after 7 days of induction in vitro. It was found that the TZP intervention did not show a very significant promotion or inhibition effect on either the osteogenic differentiation index, alkaline phosphatase (ALP) level, or the osteogenic mineralization index in this concentration range (Fig. 2A and B).

Fig. 2.

Effects of TZP on the formation of osteoblasts and osteoclasts from BMSCs. (A) Graphs of alkaline phosphatase (ALP) level detection and quantitative statistics of the osteoblast cell line MC3T3-E1 after 5 days of osteogenic induction culture. (B) Graphs of osteoblast mineralization nodule staining of BMSCs after 7 days of osteogenic induction culture. NC is the blank control group, DIFF is the osteogenic induction and differentiation group, PM is the positive drug (Purmorphamine) control group, and DMSO is the solvent control group. The drug concentrations are 0 nM, 10 nM, and 100 nM, respectively. (C) Osteoblast cell line MC3T3-E1 was cultured for 5 days in osteogenic induction culture. The alkaline phosphatase (ALP) level as determined after 5 days of osteogenic induction. (D) Mineralized nodule staining (alizarin red S method) of osteoblasts from bone marrow mesenchymal stromal cells of 3-week-old wild-type mice after 9 days of osteogenic induction. The drug concentrations are 0.5 μM, 1 μM, 3 μM, and 5 μM, respectively. (E) Anti-tartrate acid phosphatase (TRAP) staining of osteoblasts from bone marrow mesenchymal stromal cells of 3-week-old wild-type mice after 4 days of osteoblastic induction, and quantitative osteoblastic cell counts. (F) Microscopy Magnified cell images (4×) were taken.

To further determine the effective concentration range for the safety of TZP on the differentiation of BMSCs, we selected the concentration range with four gradient ranges of 0.5 μM, 1 μM, 3 μM, and 5 μM, and we found that in the large range of concentration gradient ranges, there was no significant difference in the osteoblast differentiation index (ALP level), as well as the osteogenic mineralization index (AR staining) (Fig. 2C). However, TZP had a significant inhibitory effect on the mineralization of BMSCs at a concentration of 5 μM (Fig. 2D).

We also tested the effect of TZP on osteoclasts. Firstly, BMSCs isolated from 3-week-old wild-type mice were induced to osteoclasts by RANKL stimulation, and subsequently the cells were treated with four concentration ranges of 0.5 μM, 1 μM, 3 μM, and 5 μM. The results indicated that there was no effect on the staining of anti-tartaric acid phosphatase (TRAP) as well as the number of osteoclasts in this concentration range, suggesting that TZP has no significant promotion or inhibition effect on the differentiation of BMSCs. This suggests that TZP has no significant promotion or inhibition effect on osteoclast differentiation (Fig. 2E and F).

3.3. TZP administration alters the composition of gut microbiota in db/db mice

Over the last decade, numerous investigations have demonstrated that gut microbiota has a great impact on the regulation of bone homeostasis, which brought a new interdisciplinary field known as ‘osteomicrobiology’. To determine whether the gut microbiota could reverse the side effects such as that caused by TZP on bone mineral density, blood sugar reduction and weight loss, we collected fecal samples from female db/db mice and age-matched control ones treated with TZP for 6 weeks and performed high-throughput gene sequencing for 16S ribosomal RNA (rRNA) using fecal bacterial DNA. Principal coordinate analysis (PCoA) showed that the gut microbiome of the TZP-treated mice was significantly different from that of the control mice (Fig. 3A and B). In terms of relative abundance, among the top 20 key microbial operational taxonomic units (OTUs) at the generic level, the abundance decrease of the family of Lachnospiraceae was observed in TZP-treated mice when compared with controls (Fig. 3C and D). This pattern was recapitulated in Linear discriminant analysis effect size (LEfSe) analysis (Fig. 3E).

Fig. 3.

Effects of TZP administration on the composition of gut microbiota in db/db mice. (A) α-diversity analysis between the control group and experimental group with Observe, Chao1, ACE, Shannon, Simpson, J, according to intestinal flora. (B) Principal coordinate analysis (PCoA) plot based on 16S rDNA sequencing of feces samples from the control and experimental mice. (C) Bar chart showing the relative abundance of the top 20 microbiota at the genus level of feces between the two groups. (D) Differential bacterial genera between the two groups, including up-regulated or down-regulated ones. (E) LEfSe analysis showed that the significantly different abundances of bacterial taxa in the TZP-treated mice relative to the control mice. Ctrl: n = 4, Tirzepatide: n = 7.

3.4. Lachnospiraceae supplementation restores TZP-induced bone mass loss

Next, we investigated whether bacteria in the family Lachnospiraceae increased bone mass in experimental mice. After 6 weeks of Lachnospiraceae supplementation, the bone mass data by micro-CT scanning of the femur showed that Lachnospiraceae administration had a significant protective effect on the bone mass in TZP-treated mice compared with that of TZP-treated mice without Lachnospiraceae supplementation (p < 0.05). In TZP-treated mice, Tb.Th was significantly higher (p < 0.05) in the Lachnospiraceae supplementation group than that of mice treated without Lachnospiraceae supplementation (p ≥ 0.05), but there was no significant difference in cortical bone thickness between the two groups (p ≥ 0.05) (Fig. 4A–C).

Fig. 4.

Effects of Lachnospiraceae supplementation on TZP-induced bone mass lost. (A) 3D image of reconstruction of the trabecular bone after 6 weeks of TZP injection in db/db mice with Lachnospiraceae supplementation (scale bar: 100 μm). (B) 3D image of Micro-CT cortical bone 3D reconstruction of the femur after 6 weeks with Lachnospiraceae supplementation and TZP injection in db/db mice (scale bar: 100 μm). (C) Statistical graphs of bone density parameters related to trabecular bone volume in db/db mice after 6 weeks of dosing: relative bone volume (BV/TV), n = 10, 5, 10, 4, respectively, cortical bone thickness (Ct. Th), n = 10, 5, 10,6, respectively, trabecular bone thickness (Tb. Th), n = 10, 5, 10, 6, respectively, trabecular bone number (Tb. N), n = 12, 5, 11, 6, respectively, and trabecular separation (Tb. Sp), n = 12, 5, 12, 6, respectively. (D) Serological assays in db/db mice injected with TZP with or without Lachnospiraceae administration. ∗p < 0.05, ∗∗p < 0.001, ns p ≥ 0.05.

We further performed an enzyme-linked immunosorbent assay to detect serum bone metabolism indexes: serum expression levels of collagen type 1 cross-linked carboxy-terminal peptide (CTX-1) and procollagen type 1 amino-terminal propeptide (P1NP). The results showed that the serum P1NP level of TZP-intervened mice was significantly higher than that of the control group. The serum P1NP level in TZP-intervened mice with Lachnospiraceae was significantly higher than that of the mice without Lachnospiraceae (p < 0.05). In the control group, either Lachnospiraceae supplementation or not had no significant effect on P1NP levels, whereas no significant differences in serum CTX-1 levels were found between these groups (Fig. 4D).

3.5. Lachnospiraceae restores bone morphology following TZP treatment

Subsequently, the effect of Lachnospiraceae on bone mass loss of the femur caused by TZP was monitored in db/db mice. The femur was histologically sectioned to analyze the femur bone mass and osteoclasts. H&E staining results showed that Lachnospiraceae supplementation effectively increased the number and thickness of trabeculae of TZP-treated mice (Fig. 5A); in TRAP staining, a decrease in osteoclasts could also be found in the Tirzepatide + Lachno group compared to TZP-treated mice alone (Fig. 5B).

Fig. 5.

Effects of Lachnospiraceae on bone morphology following TZP treatment. (A) H&E staining of femurs in db/db mice injected with TZP and Lachnospiraceae administration for 6 weeks; (B) TRAP staining of femurs after db/db mice were injected with TZP and Lachnospiraceae administration for 6 weeks; (C) Von kossa staining of the vertebrae after 6 weeks of injection of TZP and Lachnospiraceae in db/db mice; (D) Toluidine blue staining of vertebrae of db/db mice after injection of TZP and Lachnospiraceae administration for 6 weeks.

We also analyzed histological sections of the vertebrae. Von kossa staining showed that the Lachnospiraceae administration effectively enhanced bone mass and mineralization in the TZP-treated mice model (Fig. 5C); an increase in osteoblasts in the Toluidine blue staining was also detected in the Tirzepatide + Lachno group compared to the TZP-treated model mice (Fig. 5D).

3.6. TZP administration alters the composition of gut microbiota metabolites in db/db mice

It has been well-studied that gut microbes regulate host metabolism and immunity mainly by their outstanding ability to produce numerous and diverse metabolites from the fermentation of dietary polysaccharides. To uncover which metabolites were the key factors during Lachnospiraceae-initiated the ameliorative period on TZP-induced bone loss, we performed an unbiased metabolomics on fecal samples from TZP-treated mice and Lachnospiraceae-replenished-TZP-treated mice. Anionic and cationic ion chromatogram metabolomic PCoA plots showed distinct metabolite profiles between these two groups (Fig. 6A). At the individual metabolite levels, a volcano plot under positive polarity mode revealed that a total of 79 and 38 metabolites were down- and up-regulated, respectively, in TZP (T) group fecal samples when compared with TZP + Lachno (T + L) group (Fig. 6B). At the same time, there were a total of 73 and 30 metabolites that were down- and upregulated on the volcano plot under negative polarity mode when compared TZP-treated mice with control mice (Fig. 6B). In particular, both the volcano plot and lollipop chart revealed that levels of Evodiamine were significantly higher in TZP-treated mice than in the control mice (Fig. 6B and C). Interestingly, the mechanism by which Evodiamine inhibits RANKL-induced osteoclastogenesis has been demonstrated to be through NF-κB and calcium signaling pathways [51]. In addition, the Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis revealed that among those enriched pathways, biosynthesis of Evodiamine was notably altered in drug-treated mice relative to the control mice (Fig. 6D).

Fig. 6.

Effects of TZP administration on the composition of gut microbiota metabolites in db/db mice. (A) PCoA analysis of fecal anionic and cationic metabolites between control and experimental groups. (B) Volcano plot showing the differential clustering patterns of metabolites, which included anionic and cationic metabolites in the two groups. (C) Lollipop chart of the top 30 most obviously changed metabolites between the control and experimental groups. (D) KEGG metabolic pathway map enriched for metabolites showed that there were significant differences in metabolic pathways between the T and T + L groups. T: Tirzepatide group (n = 6); T + L: Tirzepatide + Lachno group (n = 6).

3.7. Metabolite Evodiamine suppresses osteoclastogenesis

To examine the effect of Evodiamine on osteoclast differentiation, osteoclasts differentiated from BMSCs in osteoclast-induced culture for 5 days were stained using anti-tartrate acid phosphatase (TRAP) staining and the number of osteoclasts was quantified. We found that Evodiamine inhibited the differentiation of osteoclasts (Fig. 7A and B).

Fig. 7.

Effects of metabolite Evodiamine on osteoclastogenesis. (A) Tartrate-resistant acid phosphatase (TRAP) staining of osteoclasts and quantitative statistics of osteoclast numbers in bone marrow mesenchymal cells of 3-week-old wild-type mice after 5 days of osteoclast induction culture. (B) Magnified cell images were taken under a microscope (4×). ∗∗p < 0.001.

4. Discussion

Compared to non-T2DM patients, patients with T2DM tend to have lower bone turnover status, suggesting that they are more prone to fractures even at the same levels of bone density. Concerning the treatment of chronic diseases, T2DM patients usually need long-term hypoglycemic medication maintenance. TZP, a novel hypoglycemic and weight-loss drug, its safety on bone metabolism cannot be ignored. Our study found that bone mass in db/db mice decreased after TZP treatment, which was related to intestinal flora dysregulation, especially the decrease of bacterial members in the Lachnospiraceae family. Simultaneous supplementation of Lachnospiraceae during TZP treatment prevented bone mass loss in db/db mice caused by TZP. Metabolomics study revealed that Lachnospiraceae regulate bone metabolism through a metabolite, Evodiamine, which inhibited osteoclastogenesis.

The mechanism of osteoporosis and the risk of fracture in T2DM patients is complex [52]. At the same time, there is a complex interaction between diabetes, obesity, and osteoporosis. Both obesity and weight loss can negatively affect bone metabolism in patients with T2DM, further complicating the pathogenesis of osteoporosis in T2DM patients. Moderately obese adults have higher bone density and a lower incidence of osteoporosis than normal-weight adults [53]. However, the so-called moderate obesity is difficult to define. Although weight loss is considered an important measure to "reverse" diabetes, the effect of weight loss on BMD is generally considered to be negative.

In recent years, it has been found that incretin, the star drug of hypoglycemic and weight loss, is also involved in bone remodeling regulation, in which GIP can inhibit postprandial bone absorption, and GLP-1 also indirectly mediates acute and short-term inhibition of bone resorption through hyperphysiological concentrations of insulin [54]. However, to the best of our knowledge, the clinical studies of the effect of existing incretin therapies on fracture risk are not based on fracture as the primary outcome, and generally for small studies of GLP-1 single-channel agonists. These data need to be interpreted with caution, because fractures generally do not represent adverse events of primary concern. Shorter treatment durations may lead to more hypoglycemia-related fractures than fragility fractures. GIP single-channel agonists, which are considered to have a poor hypoglycemic effect by previous studies [55], so there is a lack of GIP treatment for clinical use. Of the currently marketed preparations of incretin, TZP has the strongest effect on weight loss. We were interested in the combined impact of the TZP intervention on the bone metabolism of already vulnerable diabetic individuals with such dramatic weight loss.

Db/db mice show severe hyperglycemia, obesity, and low bone mass [56], which is more likely to result in changes in bone mass after intervention. After 4 weeks of TZP intervention, Micro-CT results showed that femoral trabecular bone mass was significantly reduced in db/db mice compared with the control group. The reason why cortical bone isn't affected is due to structural differences. Trabecular bone and cortical bone exhibit significant differences in structure, metabolism, and blood supply, leading to their distinct sensitivity to drugs. Trabecular bone has a porous, honeycomb-like structure and is highly vascularized, allowing drugs to rapidly reach its surface via the bone marrow vasculature. This facilitates drug penetration and interaction with osteoblasts and osteoclasts on the bone surface. To explore the reasons for this finding, we first carried out TZP intervention in vitro, and the results showed that the concentration of 5.0 μM TZP could increase apoptosis of osteoblasts, considering that it was associated with excessively high drug concentrations, as this dose was already much higher than the dose (70 nmol/kg/day) used in our db/db mice study. However, TZP with a concentration range of 0.5–3.0 μM showed no significant effect in promoting or inhibiting the differentiation of both osteoblasts and osteoclasts. Therefore, we can exclude the possibility that TZP directly inhibits osteoblast and osteoclast differentiation and causes the decrease in bone mass in db/db mice.

The intestinal flora is an important link in regulating bone health[41,42,[45], [46], [47]]. Since TZP is a hormone derived from the gut, the gastrointestinal reactions of patients are extremely common, which may lead to an altered intestinal microflora. Therefore, it is reasonable to speculate that the gut microbiota mediated the decrease in BMD after TZP treatment. We collected the feces of db/db mice after TZP intervention to perform microbial 16s rRNA sequencing and found that the diversity of intestinal microflora in the Tirzepatide group was significantly reduced and the species composition was significantly altered compared to the solvent control group. We further analyzed the differences in intestinal flora between the TZP group and control group and identified the top 10 intestinal flora with significant abundance differences related to bone metabolism, especially genera in the family of Lachnospiraceae. Lachnospiraceae are cultured in the anaerobic chamber, participate in the metabolism of various carbohydrates, produce beneficial metabolites such as short-chain fatty acids, and provide energy for the host.

To test whether the decreased abundance of Lachnospiraceae is related to the Tirzepatide-mediated loss of femoral bone mass, we further carried out Lachnospiraceae supplement experiment, in parallel with the administration of Tirzepatide treatment. We found that the Lachnospiraceae-supplement can prevent the reduction of trabecular bone mass caused by Tirzepatide intervention. Therefore, we considered that the intestinal Lachnospiraceae are one of the major flora that mediates the decrease in bone mass in db/db mice after Tirzepatide treatment. It has been shown that Lachnospiraceae play an important role in regulating many diseases, such as obesity, enteritis, radiation, T2DM, and osteoporosis [57]. However, the mechanism of Lachnospiraceae in osteoporosis is still lacking.

It is well known that the intestinal microflora regulate bone metabolism mainly through relative metabolites, and the effects of short-chain fatty acids on osteoblasts and osteoclasts have been widely reported [43,58]. However, other metabolites have not been fully studied. Therefore, we tried to explore whether there are other metabolites involved in the mechanism of Tirzepatide influencing bone metabolism. We compared the fecal metabolites between the Lachnospiraceae-supplement group and the control group in Tirzepatide treatment and found significant metabolite differences, among which the expression difference of Evodiamine was particularly obvious. Therefore, we conducted an in vitro intervention experiment of Evodiamine in inducing differentiation of BMSCs and found that although Evodiamine did not effectively enhance the differentiation of osteoblasts, it had a significant inhibitory effect on osteoclast differentiation. Previous studies have clearly shown that Evodiamine inhibits RANKL-induced osteoclastogenesis through NF-κB and calcium signaling pathways [51]. While our current study focused on evodiamine's direct effects on osteoclast differentiation, we acknowledge that its metabolites may also contribute to the observed biological activities. In the future, we will expand our study to include: potential major metabolic pathways of evodiamine based on published pharmacokinetic studies; reported bioactive metabolites and their possible cellular targets; and comparative analysis of parent compound versus metabolites in bone metabolism regulation. This addition will provide more comprehensive mechanistic insights into evodiamine's anti-osteoclastic effects.

To our knowledge, there is a lack of research on the effects of Tirzepatide treatment on bone metabolism so far. Our results suggest that: first, in individuals of T2DM with poor bone mass or osteoporosis, we should pay attention to the risk of further decline in bone mass after Tirzepatide treatment, and it is necessary to follow up on their bone metabolism; second, by supplementing some probiotics, it may prevent the decline in bone mass related to Tirzepatide treatment. To some extent, our study provides a reference for the prevention and treatment of drug-related osteoporosis in patients with T2DM in the future. However, our study also had shortcomings: due to the limitations of our study design, we could not distinguish the effect of GLP-1/GIP dual-receptor activation and weight loss on bone mass in this experiment. In addition, although we recorded the dietary amounts of mice, a sufficient evaluation of nutritional composition was lacking. We look forward to conducting further research to investigate this question.

However, there was inevitable limitation in our study. Evodiamine is one of the main active indoloquinazoline alkaloids of the herbal medicine Evodia rutaecarpa, which has a variety of pharmacological actions, including anti-inflammatory [59], anti-obesity [60], hypotensive, and vasodilatory effects [61]. Based on our current data, we don't have enough evidence to prove that Evodiamine is a gut microbiota-derived metabolite. However, we hypothesize that Evodiamine may be metabolized by gut microbiota because of its chemical structure and host enzymatic hydrolysis. Evodiamine is an indole alkaloid, and its parent nucleus structure (indole ring) is highly similar to tryptophan metabolites (such as indole-3-acetic acid), which are well-studied gut microbiota-derived metabolites [62]. Besides, several studies demonstrate the microbiota-mediated modification of medicinal herbs and their constituents, especially alkaloids found in many medicinal herbs [63]. In one study, certain gut bacterial species can metabolize alkaloids, such as berberine and sanguinarine, into biologically active metabolites with enhanced pharmacological effects [64]. Berberine has a very similar chemical structure to Evodiamine. Ru Feng et al. reported that the gut microbiota converts Berberine into its absorbable form, dihydroberberine, by the bacterial nitroreductases [65]. According to the information above, we speculate that the gut microbiota also metabolizes Evodiamine. Since we lack direct data on Evodiamine's origin, we have carefully used the term “microbiota-related metabolite” in our manuscript to refer to Evodiamine. It would be an interesting topic about whether and how the gut microbiota metabolizes Evodiamine in future studies.

5. Conclusions

In summary, TZP administration leads to bone loss and a significant reduction in Lachnospiraceae in the context of diabetes and obesity, while supplementation with Lachnospiraceae ameliorates TZP-induced bone loss. Our studies suggest that targeting the composition of the gut microbiota may be a potential approach to protect bone health in T2DM patients treated with TZP.

Ethical approval statement

The animal study was reviewed and approved by The Laboratory Animal Management and Ethics Committee of Xiamen University.

Author contributions

N.C., B.H.S, M.D.Z.,and X.M.L. produced the figures and wrote the manuscript. M.D.Z. performed the mouse experiments. J.L.H. analyzed the 16s rRNA gene sequencing and untargeted metabolomics data. H.G., R.X. and Y.J.L. conceptualized the project and revised the manuscript. All the authors have agreed to the submission of this manuscript for publication.

Funding

This study was supported by the National Key R&D Program of China (No. 2024YFC3407000-2024YFC3407004), the National Natural Science Foundation of China (No. 82270918, No. 82372362, and No. 32370052), the Natural Science Foundation of Fujian Province (2022J06003), and the project of high-level talent teams by Quanzhou Science and Technology Bureau (2024CT002).

Declaration of competing interest

The authors declare no conflict of interest.

Appendix B. Supplementary data

The following are the Supplementary data to this article: