Host genotype is associated with selective colonization of human fecal microbiota in NCG and SGM3 mice: implications for microbiota-based therapies

Abstract

Humanized mouse models are widely used to investigate host–microbiota interactions, yet the extent to which host background contributes to engraftment fidelity remains incompletely defined. In this study, we transplanted fecal microbiota from healthy human donors into NCG and SGM3 mice and characterized engraftment using 16S rRNA sequencing. Both models exhibited reduced diversity relative to donors, but their colonization trajectories diverged. NCG recipients appeared closer to donors in β-diversity space, a pattern largely associated with the expansion of a limited set of opportunistic Proteobacteria such as Escherichia–Shigella and Citrobacter. In contrast, SGM3 mice displayed modestly higher α-diversity and retained a broader set of donor-associated genera, with selective enrichment of Bacillus, yet exhibited greater predicted functional divergence, with reductions in pathways related to ABC transport and carbohydrate metabolism. Several strictly anaerobic commensals, including Faecalibacterium, failed to colonize in either genotype. Collectively, these findings suggest that host genotype is associated with selective colonization of human fecal microbiota and support the utility of integrating compositional and functional criteria when selecting experimental models for translational microbiome research.

INTRODUCTION

The gut microbiota, a complex community of approximately 3 × 10^13 microorganisms, plays a fundamental role in regulating host physiology, including nutrient metabolism, intestinal barrier integrity, and immune homeostasis [1]. Dysbiosis of the gut microbiota has been associated with a wide spectrum of diseases, ranging from inflammatory bowel disease (IBD) and metabolic syndrome to neurodegenerative and cardiovascular disorders [2]. Fecal microbiota transplantation (FMT), the transfer of stool from healthy donors into patients, has emerged as an effective therapeutic intervention, with established clinical efficacy for recurrent Clostridioides difficile infection and promising outcomes in ulcerative colitis and metabolic diseases [3,4,5,6,7]. These developments highlight the urgent need for reliable preclinical models to dissect host–microbiota interactions and to improve the translational relevance of microbiota-based therapies.

Humanized immunodeficient mice provide an indispensable platform for preclinical FMT research by enabling the engraftment of human microbiota in a controlled experimental setting [8]. However, their translational value is constrained by genotype-specific colonization resistance, a phenomenon in which host immune and physiological barriers restrict the stable colonization of exogenous taxa [8,9,10,11]. For instance, even in germ-free or immunodeficient backgrounds, human commensals such as Bacteroides and Faecalibacterium exhibit poor persistence compared with in human donors [12, 13]. This limitation has restricted efforts to model human-relevant microbial communities in mice and to link microbial alterations with disease phenotypes.

Among humanized strains, NCG mice (NOD/Shi-scid/IL-2Rγnull) are widely used because the absence of functional T, B, and NK cells minimizes immune-mediated rejection of transplanted human microbiota [14]. Nevertheless, their gut microbial composition remains relatively stable, and long-term engraftment of key human taxa, such as Bacteroides, Faecalibacterium, Clostridia, and Bifidobacterium, is limited [12, 13]. These constraints restrict the applicability of NCG mice to only a narrow range of intestinal disease models. In contrast, SGM3 mice (NSG-SGM3) are transgenic for three human cytokines: interleukin-3 (IL-3), granulocyte/macrophage colony-stimulating factor (GM-CSF), and stem cell factor (SCF). These cytokines promote the development of human-like myeloid cells, generating a cytokine-driven immune niche [15, 16]. These features suggest that SGM3 mice may provide a more permissive environment for human microbiota colonization, but systematic evaluations of their utility in FMT and host–microbiota interaction studies are lacking.

We hypothesized that transgenic cytokine expression in SGM3 mice would modulate immune–microbiota interactions and create a niche more supportive of human microbial engraftment, particularly for taxa associated with gut health, such as Bacillus. Members of the genus Bacillus include spore-forming bacteria that have been studied as candidate probiotics in humans and animals. Certain Bacillus species, such as Bacillus subtilis, Bacillus clausii, and Bacillus coagulans, have been shown to survive passage through the gastrointestinal tract and to contribute to intestinal health by supporting gut microbial balance, strengthening epithelial barrier function, and modulating host immune responses [17,18,19].

To address this knowledge gap, we conducted a direct comparison of NCG and SGM3 mice to evaluate engraftment efficiency, taxonomic composition, and stability of transplanted human fecal microbiota. By dissecting genotype-specific determinants of colonization resistance, our study provides a mechanistic foundation for enhancing the translational relevance of FMT research and informing the design of next-generation microbiota-based therapies.

MATERIALS AND METHODS

Animals

Two types of SPF-grade mice (8 weeks old, male, n=6 per group) were purchased from the Laboratory Animal Center of Xuzhou Medical University (Xuzhou, China): NCG (NOD/Shi-scid/IL-2Rγnull) and SGM3. SGM3 is based on the NSG (NOD.Cg-Prkdc^scid Il2rg^tm1Wjl^) background and transgenically expresses human IL-3, GM-CSF, and SCF, resulting in profound immunodeficiency combined with a cytokine-driven human-like myeloid niche. Mice were housed in individually ventilated cages (IVCs) under controlled conditions: temperature 22 ± 1°C, humidity 50 ± 10%, and a 12-hr light/dark cycle. They were provided sterile food and water ad libitum. All animal experiments were approved by the Ethical Guidelines of the Xuzhou Medical University Laboratory Animal Ethics Management Committee (approval number 202205A053) and conducted in accordance with the Guide for the Care and Use of Laboratory Animals.

Reagents and instruments

Key reagents included ampicillin sodium salt (Tuopu Bioengineering Co., Ltd., Shandong, China), neomycin trisulfate hydrate KGI Biotechnology Co., Ltd., Jiangsu, China), metronidazole (Weikemai Biotechnology Co., Ltd., Jiangsu, China), vancomycin (Lianshuo Biotechnology Co., Ltd., Zhejiang, China), and phosphate-buffered saline (PBS; YEASEN Biotech Co., Ltd., Shanghai, China). Instruments included a −80°C freezer (Thermo Fisher Scientific, Waltham, MA, USA), high-speed centrifuge (Eppendorf, Germany), homogenizer (IKA, Staufen, Germany), and clean bench (Purification Equipment Co., Ltd., Jiangsu, China).

Antibiotic pretreatment

Mice were pretreated with a 4-antibiotic cocktail to deplete indigenous gut microbiota: ampicillin (1 g/L), neomycin (1 g/L), metronidazole (0.75 g/L), and vancomycin (0.75 g/L) dissolved in sterile drinking water. The cocktail was replaced weekly for 2 weeks.

FMT procedure

Fecal sample collection from human donors: Fresh fecal samples were collected from 3 healthy human donors (25–35 years old, no history of gastrointestinal disease, antibiotics use, or chronic illness within 3 months) at the Affiliated Hospital of Xuzhou Medical University. Donors provided written informed consent.

Fecal suspension preparation: Each fecal sample (0.3 g) was homogenized with 3.5 mL sterile PBS in a sterile centrifuge tube. The homogenate was centrifuged at 3,000 rpm for 5 min at 4°C to remove large particles, and the supernatant (fecal suspension) was collected. The suspension was used within 20 min of preparation to ensure microbial viability.

Mouse gavage: Mice were gavaged with the fecal suspension (100 µL/10 g body weight) once daily for 3 consecutive days (D14–D16).

Sample collection

Fecal samples were collected at 2 time points: baseline (D0), before antibiotic pretreatment; and 1 week post-FMT (D23), 7 days after the last gavage. Samples were immediately frozen in liquid nitrogen and stored at −80°C until DNA extraction.

DNA extraction and sequencing

Fecal microbial DNA was extracted using a QIAamp PowerFecal Pro DNA Kit (Qiagen, Hilden, Germany) according to the manufacturer’s protocol. The V4–V5 region of the 16S rRNA gene was amplified with primers 515F (5′-GTGCCAGCMGCCGCGGTAA-3′) and 907R (5′-CCGTCAATTCMTTTRAGTTT-3′). Amplicon libraries were sequenced on the Illumina NovaSeq 6000 platform (Illumina, San Diego, CA, USA) by Shanghai Meiji Biomedical Technology Co., Ltd.

Bioinformatics and statistical analysis

Alpha diversity was calculated using the Shannon (diversity), Chao1 (richness), and Simpson (evenness) indices in QIIME 2. Differences between groups were tested via Kruskal–Wallis H test.

Beta diversity was calculated by principal coordinate analysis (PCoA) based on unweighted UniFrac distances to visualize community similarity. Analysis of similarities (ANOSIM) was used to test group differences.

Taxonomic composition was analyzed at the phylum and genus levels using MEGAN6. Relative abundances were compared via t-test (two groups) or one-way ANOVA (multiple groups).

Discriminant taxa were identified via linear discriminant analysis (LDA) effect size (LEfSe; LDA >4, p<0.05) to highlight strain-specific enriched taxa.

Statistical software

All statistical analyses were performed using GraphPad Prism 9.0 (GraphPad Software, San Diego, CA, USA) and R 4.3.0. p<0.05 was considered statistically significant.

RESULTS

Experimental design and sequencing overview

To investigate genotype-dependent colonization of human fecal microbiota in immunodeficient mice, we performed FMT into NCG and SGM3 mice following broad-spectrum antibiotic pretreatment. Fecal samples were obtained at baseline and following FMT, and bacterial community profiles were determined by 16S rRNA gene sequencing (Fig. 1A). Across all samples, a total of 1,875,561 high-quality reads were obtained, clustered into 4,417 amplicon sequence variants (ASVs). Sequencing quality assessment demonstrated sufficient depth, as rarefaction curves reached clear plateaus across all groups (Fig. 1B, 1C). Rank-abundance curves indicated that human donors exhibited the richest and most even microbial communities, SGM3 recipients displayed intermediate distributions, and NCG mice showed the steepest decline, reflecting dominance of a restricted set of taxa (Fig. 1D). These quality assessments confirmed that the dataset provided a reliable basis for diversity and taxonomic comparisons.

Fig. 1.

Study design and sequencing quality control.

(A) Experimental timeline: two weeks of antibiotic conditioning followed by three consecutive fecal microbiota transplantation (FMT) gavages (D14–D16) into NCG and SGM3 mice. Fecal sampling was performed at baseline (D0) and 1 week post-FMT (D23). (B) Rarefaction curves reached plateaus across groups, indicating adequate sequencing depth. (C) Sequencing summary: 1,875,561 high-quality reads clustered into 4,417 ASVs across all samples. (D) Rank-abundance plots show the highest richness/evenness in human donors, intermediate profiles in SGM3, and a steep drop-off in NCG, consistent with dominance of a restricted set of taxa. See Methods for library prep and bioinformatics (QIIME 2/DADA2; unweighted UniFrac).

Alpha and beta diversity revealed genotype-dependent engraftment

Analyses of within-sample diversity revealed important genotype-specific effects on engraftment (Fig. 2A–2C). Prior to FMT, both NCG and SGM3 mice displayed similarly depleted communities, with the observed richness, Shannon and Simpson indices showing no statistically significant differences between genotypes. After transplantation, α-diversity remained significantly lower in both strains than in human donors, while non-zero diversity values indicated partial establishment of donor communities. However, alpha diversity remained significantly lower than in human donors, consistent with the well-documented loss of low-abundance taxa that typically accompanies cross-species microbiota transfer [12, 13]. SGM3 recipients displayed modestly higher Sobs and Chao1 indices compared with NCG mice, although these differences did not reach statistical significance.

Fig. 2.

Alpha and beta diversity indicate genotype-dependent engraftment.

(A–C) Alpha diversity (Sobs, Chao1, Shannon/Simpson) decreased after antibiotic conditioning and remained lower than human donors post-fecal microbiota transplantation (FMT). SGM3 displayed a modest trend toward higher richness versus NCG (ns). (D) Principal coordinate analysis (PCoA) of unweighted UniFrac distances shows donor, NCG, and SGM3 separation; NCG shifts closer to donor centroids than SGM3 post-FMT. (E) Hierarchical clustering recapitulates ordination. Group differences were significant by ANOSIM/PERMANOVA. Loss of low-abundance taxa following cross-species transfer is consistent with prior reports.

Beta diversity analyses offered complementary insights into community-level structure (Fig. 2D, 2E). Principal coordinate analysis of unweighted UniFrac distances showed clear separation among human donors and the NCG and SGM3 recipients (Fig. 2D). Donor samples formed a distinct cluster, while NCG mice shifted closer to donor centroids after FMT, suggesting greater apparent similarity in overall community structure. In contrast, SGM3 samples remained more distant in ordination space. This pattern was corroborated by hierarchical clustering, where NCG samples grouped more closely with donors than SGM3 samples (Fig. 2E). Consistent with these visualizations, NMDS analysis together with ANOSIM (R=0.58, p=0.001) and PERMANOVA confirmed significant community-level differences across groups (Supplementary Fig. 1).

Shared genera and baseline comparisons with human donors

Shared-taxa analyses provided an additional dimension for evaluating engraftment fidelity by quantifying overlap with the donor community (Fig. 3A–3D). Prior to FMT, both mouse strains were compositionally distinct from human donors at the species level. In NCG mice, Citrobacter and Muribaculaceae dominated baseline communities, whereas SGM3 mice harbored higher proportions of Firmicutes, including Lactobacillus and Staphylococcus (Fig. 3A, 3B). Such baseline discrepancies are not trivial, because pre-existing communities are known to exert strong priority effects, whereby early colonizers can monopolize ecological niches and restrict the establishment of incoming taxa [20,21,22].

Fig. 3.

Differentially abundant taxa and shared-taxa analyses.

(A) Phylum/genus-level compositions for donors and recipients. (B) Baseline differences between genotypes prior to fecal microbiota transplantation (FMT) (e.g., Citrobacter/Muribaculaceae in NCG; Firmicutes including Lactobacillus/Staphylococcus in SGM3) highlight potential priority effects. (C) Donor-shared genera counts: 67 in SGM3 vs. 49 in NCG. (D) A heatmap of dominant genera shows broader donor-like bands in SGM3 and expansion of a few Proteobacteria lineages (Escherichia–Shigella, Citrobacter) in NCG. (E) Genus-level differential abundance underscores SGM3-specific enrichment of Bacillus, while Akkermansia and Lactobacillus colonized both genotypes. (F, G) LEfSe identifies group-discriminant taxa. Strict anaerobes (e.g., Faecalibacterium, Collinsella) failed to engraft in either genotype.

Quantitative overlap analyses showed that 67 of 185 donor genera were detected in SGM3 recipients, compared with 49 in NCG mice (Fig. 3C), indicating that SGM3 retained a broader set of donor-associated genera. At the level of individual samples, genus counts followed the expected hierarchy, with humans exhibiting the highest richness, SGM3 mice exhibiting intermediate richness, and NCG exhibiting the lowest richness (Fig. 3A). This pattern is consistent with previous reports that humanized mice generally exhibit reduced richness compared with donors [23, 24] but that immune-modified strains can sustain more complex communities. Heatmap clustering further illustrated these patterns, with human donors displaying a broad representation of taxa, SGM3 samples preserving wider donor-associated profiles, and NCG samples being dominated by a limited number of genera, particularly Escherichia–Shigella and Citrobacter (Fig. 3D).

Together, these results indicated that assessments of engraftment fidelity depend on the analytical perspective: although NCG samples appeared closer to donors in ordination space, this pattern was associated with dominance by a restricted set of shared taxa, whereas SGM3 mice retained a broader repertoire of donor-associated genera.

Differentially abundant taxa highlight distinct colonization signatures

Differential abundance analyses resolved distinct genus-level colonization patterns between NCG and SGM3 mice (Fig. 3A, 3E; Supplementary Fig. 2). In both recipient genotypes, Escherichia–Shigella and Citrobacter expanded to high relative abundance, despite their comparatively low prevalence in human donors (Fig. 3A, 3E). These taxa have been reported to display strong growth capacity under variable nutrient conditions and are frequently observed to expand following antibiotic perturbation in mammalian gut ecosystems [25, 26].

Among health-associated commensals, genotype-specific patterns were observed. Akkermansia, a mucin-degrading Verrucomicrobia with important roles in mucosal health and host metabolic regulation [27], and Lactobacillus, a hallmark beneficial genus frequently used in probiotic interventions [28], successfully colonized both NCG and SGM3 recipients (Fig. 3E, Supplementary Fig. 2). In contrast, Bacillus species, which are spore-forming Firmicutes with broad probiotic functions, were selectively enriched in SGM3 mice (Fig. 3E, Supplementary Fig. 2). The enrichment of Bacillus is particularly noteworthy because members of this genus can produce antimicrobial peptides, modulate epithelial barrier function, and stimulate host immune responses [29,30,31]. Several strictly anaerobic commensals that are dominant in the human gut, including Faecalibacterium, Collinsella, Dorea, and others (Fig. 3E, Supplementary Fig. 2), failed to establish in either recipient genotype, despite their presence in human donors. Prior studies have similarly reported that certain strictly anaerobic taxa fail to persist in humanized or antibiotic-conditioned mice [13, 32].

LEfSe analysis further corroborated these findings by identifying discriminant taxa that defined each group (Fig. 3F, 3G). NCG recipients were characterized by enrichment of commensals such as Lactobacillus and Akkermansia, together with taxa related to Burkholderiales and Parasutterella, indicating a microbiota shaped by both beneficial and opportunistic lineages. By contrast, SGM3 recipients were typified by indicator taxa including Citrobacter, Muribaculaceae, and Bacillus, suggesting a broader capacity to sustain diverse donor-derived microbes. Human donors, in turn, were distinguished by enrichment of hallmark health-associated taxa such as Faecalibacterium, Collinsella, and members of Ruminococcaceae and Lachnospiraceae, which were not maintained in murine hosts.

Functional predictions revealed divergent metabolic configurations

Functional inference based on PICRUSt2 indicated that genotype-specific engraftment was accompanied by distinct metabolic remodeling in recipient communities [33] (Fig. 4A, 4B). Despite harboring a broader set of donor-derived genera, SGM3 mice exhibited more extensive deviations from human donors at the functional level. Multiple modules associated with ABC-type transport systems, including permease and ATP-binding components (K01992, K01990, K06147, K02003, K02004), were consistently downregulated in SGM3 recipients. ABC transporters are central to the import of amino acids, peptides, sugars, and metal ions [34, 35], and their reduction suggests a community less reliant on active nutrient uptake [36]. Regulators of carbohydrate metabolism such as the galactose operon repressor (galR, K02529) and sucrose-6-phosphatase (K07024) were also downregulated, indicating reduced sugar catabolism. In addition, pathways linked to fermentation, nucleotide biosynthesis, and fatty-acid production (PWY-7111, PWY-7220, PWY-7663) were predicted to be depleted, consistent with a shift away from energy-intensive processes.

Fig. 4.

Predicted functional profiles reveal divergent metabolic configurations.

(A, B) PICRUSt2-based functional inference indicates more pronounced functional deviation from human donors in SGM3 communities (e.g., reduced ABC transporter components; altered carbohydrate metabolism regulators), whereas NCG retained donor-like pathways with enrichment of iron transport and sugar-catabolic enzymes linked to Proteobacteria opportunists. Results should be interpreted with caution given the predictive nature of 16S-based functional inference. FMT: fecal microbiota transplantation.

In contrast, NCG recipients showed fewer functional deviations from human donors, aligning with their closer ordination in β-diversity space. Notably, iron complex transport (K02015) was enriched in NCG mice, suggesting an increased capacity for iron acquisition characteristic of Proteobacteria such as Escherichia–Shigella and Citrobacter [37, 38]. These taxa also contributed to enrichment of carbohydrate-degrading enzymes, including 2,3-bisphosphoglycerate-dependent phosphoglycerate mutase (gpmB, K15634) and 6-phospho-β-glucosidase (bglA, K01223), consistent with an opportunistic metabolic strategy prioritizing rapid sugar utilization [39, 40].

These functional profiles were inferred from 16S rRNA gene data. They therefore represent predicted metabolic potential rather than direct measurements of gene content or activity and should be interpreted with appropriate caution. Overall, PICRUSt2 analysis showed that SGM3 communities exhibited broader taxonomic compositions with more pronounced functional divergence from human donors, whereas NCG communities displayed fewer functional deviations despite reduced taxonomic breadths (Figs. 4, 5).

Fig. 5.

Host genotype modulates the colonization pattern of human fecal microbiota after FMT in NCG and SGM3 mice

This graphical summary integrates the key findings of fecal microbiota transplantation (FMT) in two humanized mouse models. (1) Diversity features: NCG mice exhibited lower alpha diversity, but their β-diversity was closer to that of human donors, a pattern driven by the expansion of opportunistic taxa. SGM3 mice had moderately higher alpha diversity and retained a broader range of donor-derived genera. (2) Taxonomic composition: NCG recipients were dominated by Proteobacteria (e.g., Escherichia-Shigella, Citrobacter), while SGM3 mice showed selective enrichment of Bacillus. Strictly anaerobic commensals, including Faecalibacterium and Clostridia, failed to colonize in either genotype. (3) Functional profiles: NCG mice preserved donor-like functional characteristics with upregulated iron transport pathways, whereas SGM3 mice displayed significant functional divergence, featured by the downregulation of ABC transporters and carbohydrate metabolism-related pathways. Collectively, these results demonstrate that host genotype contributes to complementary engraftment patterns of human gut microbiota in murine recipients.

DISCUSSION

Our study provides a systematic comparison of microbiota engraftment in NCG and SGM3 mice, two widely used humanized models for FMT research. Although both strains showed substantial loss of diversity relative to human donors, their engraftment patterns differed in meaningful ways. NCG mice appeared closer to donors in β-diversity space, a pattern largely associated with dominance of a limited number of opportunistic lineages, whereas SGM3 mice supported a broader spectrum of donor-associated taxa, reflected by modestly higher richness, despite greater overall structural divergence from donors. These findings indicate that ordination-based similarity and taxonomic breadth represent complementary dimensions of engraftment fidelity.

Notably, SGM3 recipients displayed selective enrichment of Bacillus, a genus with documented probiotic potential, suggesting that the cytokine-modulated immune environment of SGM3 mice may create ecological conditions favorable for colonization of specific beneficial taxa. Functional predictions further revealed divergence in metabolic potential between models: SGM3 communities showed greater deviation from donor functional profiles, whereas NCG recipients retained donor-like metabolic features despite reduced taxonomic diversity. Although these functional differences were derived from 16S rRNA-based inference rather than metagenomic or metabolomic profiling and should be interpreted cautiously, the discordance between taxonomic composition and functional similarity underscores the need to integrate multi-layered analytical criteria when evaluating engraftment outcomes and selecting appropriate experimental models.

A key methodological limitation of the present study is that genus-level 16S rRNA profiling cannot discriminate true donor-derived strains from the rebound or expansion of residual murine microbiota, particularly for genera commonly shared between humans and mice. Accordingly, the greater taxonomic breadth observed in SGM3 recipients should be interpreted as increased donor-associated overlap rather than definitive donor engraftment. Future studies incorporating strain-resolved metagenomics or donor-specific tracking strategies will be required to establish microbial origin with higher confidence. Moreover, it is important to contextualize these findings within the broader landscape of experimental FMT models. Germ-free mice provide the most unambiguous system for assessing donor-derived engraftment because they lack confounding resident microbiota, whereas antibiotic-conditioned models, as used here, more closely resemble practical preclinical conditions but are inherently influenced by residual microbial communities and post-antibiotic ecological dynamics. Recognizing the complementary strengths and limitations of these model systems is essential for accurate interpretation of engraftment outcomes and for informed model selection in translational microbiome research.

Additionally, the observed colonization patterns may not reflect host genotype alone. NCG and SGM3 mice differed in their residual gut microbiota following antibiotic pretreatment, raising the possibility that pre-existing microbial communities exerted priority effects that influenced subsequent donor assembly. Moreover, antibiotic-conditioned ecological dynamics may further shape post-FMT trajectories. Thus, the engraftment outcomes described here should be interpreted within an ecological framework that integrates host immune background, residual microbiota, and post-antibiotic community restructuring.

These findings extend our previous work [41], which demonstrated that humanized mice can support partial reconstruction of donor microbiota following FMT. Whereas that study primarily focused on overall post-transplant community similarity, the present work further dissects how distinct immune genotypes are associated with differential colonization trajectories in the context of residual microbiota and antibiotic-conditioned ecological processes.

In summary, NCG and SGM3 mice capture complementary facets of human microbiota engraftment: NCG mice preserve donor-like functional attributes but are dominated by opportunistic taxa, whereas SGM3 mice support broader colonization, including selective enrichment of Bacillus and greater functional divergence. These results indicate that no single model fully recapitulates the human donor microbiota and that model selection should be guided by experimental objectives. Collectively, our study highlights the translational potential of SGM3 mice for probing probiotic colonization and immune–microbiota interactions and emphasizes the importance of ecological context in interpreting FMT outcomes.

ETHICS APPROVAL

This study protocol was reviewed and approved by the Ethical Guidelines of the Xuzhou Medical University Laboratory Animal Ethics Management Committee (approval number 202205A053).

AUTHOR CONTRIBUTIONS

RC conceived the study. RC and KY designed the experiments and supervised the study. KY, GP, and SZ performed the experiments. KY, GP, YW, and DY analyzed the data. KY and GP wrote the paper. RC reviewed and edited the paper. All authors have read and approved the article.

FUNDING

This work was supported by the Key University Science Research Project of Jiangsu Province (22KJA320005) , and the Jiangsu Provincial Key Project on Geriatric Health Research (LKZ2024009) and the Graduate Research and Innovation Projects of Jiangsu Province (KYCX23_2947) .

CONFLICT OF INTEREST

The authors declare no conflicts of interest.