Colostrum samples were collected at 3 days postpartum from lactating mothers who had given birth to term (37–41 weeks) infants at the Shanghai Tongren Hospital. Informed consent was obtained from all mothers. The study was approved by the Human Research Ethics Committee of Shanghai Tongren Hospital (Reference number 2021–086-02). Each sample was about 4 ~ 5 mL and collected on the third day postpartum. Samples were stored in sterile tubes at −80°C. We excluded mothers with diabetes, other medical conditions, or pregnancy complications. Participants were grouped by maternal prepregnancy. For the Chinese population, many studies reference the classification proposed by the China Working Group (24), which defines overweight as a BMI ≥ 24.0 kg/m² and obesity as a BMI ≥ 28.0 kg/m² (25, 26). Accordingly, we selected participants with BMI <24 kg/m² (n = 6, lean) and ≥28 kg/m² (n = 6, obese) for comparison.

Calibra Lab at DIAN Diagnostics (Hangzhou, Zhejiang, China) conducted an untargeted metabolomic analysis on the CalOmics metabolomics platform. Samples were extracted using methanol (Sigma-Aldrich, United States) at a 1:4 ratio. The mixtures were shaken for 3 min and precipitated by centrifugation at 4,000 × g for 10 min at 20°C. Four aliquots of 100 μL supernatant were transferred to sample plates, dried under nitrogen, and re-dissolved in reconstitution solutions for UPLC–MS/MS analysis. The instrument for the UPLC–MS/MS methods are ACQUITY 2D UPLC (Waters, Milford, MA, United States) plus Q Exactive (QE) hybrid Quadrupole-Orbitrap mass spectrometer (Thermo Fisher Scientific, San Jose, United States). QE mass spectrometer was operated at a mass resolution of 35,000, the scan range was 70–1,000 m/z. Statistical analysis was performed using the R language (R version 3.4.1: http://www.cran.com; RStudio version 1.4.1717: https://www.rstudio.com; mixOmics version 6.10.9: http://mixomics.org/; randomForest version 4.6–14: https://www.rdocumentation.org/packages/randomForest/versions/4.6-14). Significantly changed metabolites between cases and control groups were identified using parametric (Student’s t-test, ANOVA) or non-parametric (Wilcox’s rank test, Kruskal-Wallis) methods.

Quantification of 5′-UMP (Yuanye, China; T94168; HPLC purity ≥99%) in human milk samples was performed using a CalQuant-S LC–MS/MS system. Data acquisition was conducted using Analyst software version 1.6.2, and quantitative analysis was performed using Multiquant software. An electrospray ionization source operated in positive ion mode was used. The scan mode was multiple reaction monitoring (MRM). Chromatographic separation was achieved on a C18 reverse-phase (UPLC BEH C18, 2.1×100 mm, 1.7 μm; Waters). The mobile phase consisted of (A) water containing 0.1% formic acid (Sigma-Aldrich, United States) and 2 mM ammonium acetate (Sigma-Aldrich, United States), and (B) acetonitrile (Sigma-Aldrich, United States) containing 0.1% formic acid. A gradient elution was employed over a total run time of 4 min.

Human milk samples collected from obese and normal-weight mothers were used for analysis. For each sample, 40 μL of milk was aliquoted, and an appropriate volume of protein precipitation solution containing internal standard was added. The mixture was vortexed thoroughly and centrifuged. The resulting supernatant was collected and subjected to LC–MS/MS analysis.

All animal experiments were approved by the Medical Ethics Committee of Shanghai Tongren Hospital of China (Approval No. A2022-017-01). Zebrafish (Danio rerio) were initially obtained from the China Zebrafish Resource Center and raised in the Hongqiao International Medical Research Institute, Shanghai Jiao Tong University School of Medicine. Wildtype zebrafish (Tuebingen, strain) and transgenic zebrafish [Tg(Xla. Eef1a1:mlsEGFP)] (27) were used in this study. All animal experiments were conducted following ARRIVE guidelines and regulations. Adult zebrafish were housed in an automated zebrafish system at 28.5°C in a 14-h light: 10-h dark cycle. Healthy fertilized zebrafish embryos at 24 h post-fertilization (hpf) were transferred into E3 embryo medium (5 mM NaCl, 0.17 mM KCl, 0.33 mM CaCl₂, 0.33 mM MgSO₄, and 0.1 μg/mL methylene blue) supplemented with 200 μM 1-phenyl-2-thiourea (PTU; Sigma, United States) to inhibit melanin pigmentation. Embryos were maintained in this medium until 7 days post-fertilization (dpf). In toxicity experiments, wildtype zebrafish embryos were randomly collected in E3 fish water with different concentrations (0 μM, 100 μM, 200 μM) of 5′-UMP from 8 hpf; each group contained approximately 50 embryos. Mortality, hatchability, and heart rate were assessed at 72 hpf and 96 hpf. As no toxicity was observed at the highest tested dose (200 μM) of 5′-UMP, this concentration was selected for subsequent experiments.

To establish a zebrafish obesity model (28), 0.5 g egg yolk powder (fat% ≥ 60; Company: Nanjing Ximenuo Biotechnology Co., LTD) was diluted in a 50 mL E3 medium. A high-fat diet (HFD) was prepared daily and fed for 6 h every day from 9 a.m. to 3 p.m. for 3 days, from 4 dpf to 7 dpf. 5′-UMP was diluted to the final concentration with E3 fish water containing PTU. From 4 to 7 dpf, zebrafish larvae were treated with 5′-UMP and fed for 6 h with HFD or a control diet. Zebrafish Lipid staining was analyzed at 18 h after the last day of feeding.

On the 8 dpf, larvae were washed with PBS three times, fixed in 4% paraformaldehyde (PFA, Beyotime, P0099, China) overnight, and then washed with PBS three times. Larvae were fixed in 60% isopropanol for 30 min and stained with Oil Red O solution (0.3% Oil Red O in 60% isopropanol; Aladdin, China) for 2 h. After staining, larvae were washed alternately with PBS and observed after treatment under an Optec SZ760 stereomicroscope (Zeiss, Germany). Following the approach described in Karla Misselbeck et al. (29), the percentage of larvae exhibiting low, medium, and high lipid accumulation was calculated for each experimental group.

Nile Red staining was performed on 7 dpf zebrafish larvae (30). Zebrafish larvae were incubated with 0.5 μg/mL Nile Red solution (Aladdin, China) for 30 min in the dark, washed twice with E3 fish water, and imaged using the fluorescence microscope (Nikon, Japan).

To evaluate mitochondrial abundance in vivo, we used transgenic zebrafish larvae expressing mitochondrially targeted GFP [Tg(Xla. Eef1a1:mlsEGFP)]. Zebrafish were maintained under standard conditions and subjected to HFD feeding from 4 dpf to induce lipid accumulation. Larvae were randomly assigned to four groups: (1) control, (2) HFD, (3) HFD + 5′-UMP, and (4) 5′-UMP. At 7 dpf, larvae were anesthetized with tricaine (Sigma, USA, E10521) and imaged using a fluorescence microscope to capture GFP signals. Mitochondrial abundance was quantified based on mean fluorescence intensity using ImageJ software.

Human preadipocytes (ScienCell, United States, 7210/7220) were cultured in Preadipocyte Medium (PAM, ScienCell, United States) containing 5% fetal bovine serum (FBS), 1% penicillin–streptomycin (PS), and 1% preadipocyte growth supplement. Cells were maintained in a humidified incubator at 37°C with 5% CO₂. To induce differentiation, confluent preadipocytes were first cultured in DMEM/F12 medium (Gibco, United States) supplemented with 1 mM dexamethasone, 100 nM insulin, 1 μM rosiglitazone, 0.5 mM 3-isobutyl-1-methylxanthine (IBMX; all from Sigma, United States), and 1% PS for 4 days. Following induction, the cells were maintained in DMEM/F12 medium supplemented with 10 nM insulin and 1% PS for an additional 4 days, completing the 8-day differentiation protocol.

Cells were treated with 5′-UMP throughout the differentiation process. Control cells received an equal volume of 1% DMSO. Adipocyte differentiation efficiency was typically evaluated on Day 8 by performing Oil Red O staining (ScienCell, United States). The culture medium was refreshed every 2 days during the differentiation period.

The effect of 5′-UMP on the proliferation of human preadipocytes was evaluated by the CCK8 (Beyotime, China) assay. Briefly, cells were plated at 3.0 × 10³ density in a 100 μL volume in a 96-well plate. Then, adipocytes were treated for 0, 24, 48, and 72 h with various concentrations of 5′-UMP (0, 100, 200, and 400 μM). At each time, 10 μL of CCK-8 reagent was added to each well, followed by incubation at 37°C in a 5% CO₂ atmosphere for 1 h. Absorbance was measured at 450 nm using a microplate reader (Tecan, Swiss). Each condition was tested in 10 replicates.

Preadipocytes were treated for 48 h with various concentrations of 200 μM 5′-UMP. After that, the cells were harvested, washed, and resuspended in PBS. According to the manufacturer’s protocol, apoptotic cells were analyzed by staining cells with Annexin V-PE and 7-aminoactinomycin D (7-AAD; Vazyme, China). The cells were suspended in 1 × binding buffer at a 1.0 × 10⁶ cells/mL concentration. The cell suspension (100 μL) was then transferred to a 1.5-mL Eppendorf tube, mixed with 5 μL of Annexin V-PE and 10 μL of 7-AAD, and incubated for 20 min at RT in the dark. Within 1 h, the samples were analyzed using flow cytometry (BD, Canto II, United States). Data were analyzed using the FlowJo software.

Oil Red O staining kit (ScienCell, United States) was used to detect the lipid accumulation. Cells were fixed in 4% PFA for 15 min and stained with ORO solution (0.3% in isopropanol) for 30 min. After staining, the cells were washed 3 times and imaged using a microscope (Nikon, Japan).

Oxygen consumption rate (OCR) was assessed using a Seahorse XF24 extracellular flux analyzer (Agilent, United States). Preadipocytes were plated in Seahorse XF 24-well microplates (Agilent, United States) and fully differentiated for 8 days. Subsequently, adipocytes were treated with 200 μM 5′-UMP for 24 h. Following treatment, adipocytes were incubated in a pre-warmed assay medium (XF base medium with 10 mM glucose, 1 mM sodium pyruvate, and 2 mM glutamine) for 1 h without CO₂ before oxygen consumption analysis OCR was then assessed under basal conditions and after sequential injection of different reagents (103015–100, Agilent, United States), including 1.5-uM ATP synthase inhibitor oligomycin, 0.5 uM carbonyl cyanide 4-(trifluoromethoxy) phenylhydrazone, and 0.5-u rotenone/antimycin A. The OCR values were adjusted based on protein content.

Total RNA was isolated from cultured cells using an RNA isolation reagent (Waltham, MA, United States). RNA purity and quantification were evaluated using the NanoDrop 2000 spectrophotometer (Thermo Scientific, United States). After RNA quantification, cDNA was synthesized using 1 μg of total RNA and a cDNA synthesis kit (Takara, Japan), according to the manufacturer’s instructions. Quantitative PCR (qPCR) was performed with a StepOnePlus Real-Time PCR system (ThermoFisher, Waltham, MA, United States) using a Premier qPCR kit (Thermo Scientific, Waltham, MA, United States). The expression levels were normalized to PPIA (Peptidyl-prolylcis-trans isomerase A) and analyzed by the 2^-ΔΔCt method. PPIA: F,5’-TTCATCTGCACTGCCAAGAC-3′; R,5’-TCGAGTTGTCCACAGTCAGC-3. UCP1: F,5’-AGGTCCAAGGTGAATGCCC-3′; R,5’-TTACCACAGCGGTGATTGTTC-3′. PGC1-α: F,5’-ACCTGACACAACACGGACAG-3′; R,5’-GTCTCCATCATCCCGCAGAT-3′.

Cells were lysed in RIPA (Beyotime, China). Samples were separated by electrophoresis on 10% SDS-polyacrylamide gels and transferred to PVDF membranes. Membranes were blocked with 3% skimmed milk. After blocking, the membranes were incubated overnight at 4°C with antibodies against UCP1 (Proteintech, 23,673-1-AP, rabbit polyclonal, 1:3,000), Drp1 (Abcam, ab184247, rabbit monoclonal, 1:1,000), NRF1 (Proteintech, 12,482-1-AP, rabbit polyclonal, 1:1,500), MFN2 (Proteintech, 12,186-1-AP, rabbit polyclonal, 1:5,000), β-actin (Affinity Biosciences, AF7018, rabbit polyclonal, 1:1,000), followed by incubation with horseradish peroxidase-conjugated secondary antibodies (GeneTex, rabbit polyclonal, GTX213111-01, 1:1000) for 2 h.

To assess mitochondrial abundance and activity, differentiated adipocytes were stained using MitoTracker™ Red CMXRos (ThermoFisher, Waltham, MA, United States). Before staining, cells were washed twice with PBS to remove the residual culture medium. The adipocytes were then incubated with MitoTracker working solution (final concentration: 100 nM) diluted in serum-free PAM for 40 min at 37°C in a humidified incubator containing 5% CO₂. After staining, cells were washed gently with PBS to remove excess dye and fixed with 4% paraformaldehyde for 15 min at room temperature. Fluorescent images were captured using a Nikon fluorescence microscope (Nikon, Tokyo, Japan), and mitochondrial fluorescence intensity was quantified using ImageJ software (NIH, United States). All experimental conditions were kept constant to ensure comparability between treatment groups.

Results are presented as the mean ± standard deviation (SD). Data were analyzed using t-tests in Prism 9 (version 9.0.0), and p < 0.05 was considered statistically significant.

Demographic and clinical information for the mother and infant is presented in Table 1. As expected, women in the obesity group had a significantly higher prepregnancy BMI than the normal group (28.79 ± 0.6387 kg/m², compared with 21.83 ± 2.029 kg/m², p < 0.0001). To investigate the impact of maternal body weight on colostrum composition, untargeted metabolomic profiling was performed on colostrum samples collected at 3 days postpartum from obese (n = 6, BMI ≥ 28 kg/m²) and normal-weight (n = 6, BMI < 24 kg/m²) mothers. The results showed that 866 metabolites were identified in both groups. As shown in Figure 1A, lipids accounted for the most significant proportion (44.23%), followed by amino acids (24.25%). Notably, nucleotide metabolites represented a considerable proportion (6.93%), indicating a significant enrichment of nucleotide-related compounds in colostrum (Figure 1A).

Based on a significance threshold of p < 0.05, 79 metabolites were differentially expressed between the two groups (Figures 1B,C; Supplementary Table 1). Among them, three nucleotide-related metabolites were significantly downregulated in the colostrum of obese mothers compared to normal-weight controls: orotate (p = 0.0306), 8-hydroxyguanine (p = 0.0194), and guanine (p = 0.0306; Figure 1D). KEGG pathway enrichment analysis was then performed on the differentially expressed metabolites. Pyrimidine metabolism emerged as a significantly enriched pathway (Figure 1E). Notably, orotate (orotic acid), a key precursor in de novo pyrimidine biosynthesis, was enriched in this pathway and significantly reduced in the obese group (Figures 1B–E; Supplementary Table 2). These results suggest that orotate-related pyrimidine metabolism may be disrupted in obese mothers, play a crucial role in the pyrimidine pathway, and be associated with body weight regulation. Based on these findings, we further analyzed the metabolites of orotic acid and identified UMP as a key downstream metabolite (Figure 2A). Targeted metabolomic analysis was conducted on human milk samples from obese mothers and healthy controls (Figures 2B,C). The results revealed that the level of 5′-UMP was higher in the milk of normal-weight mothers compared to that of obese mothers. These findings suggest that 5′-UMP, as a downstream metabolite of orotic acid, may play a role in influencing offspring weight development.

To investigate the potential effects of 5′-UMP on body weight, we conducted studies using a zebrafish model. As an initial step, a toxicity test on zebrafish embryos was undertaken to study the toxicity of 5′-UMP on zebrafish. Zebrafish larvae were exposed to increasing doses (0, 100, and 200 μM) of 5′-UMP during the first 3 dpf to determine the safe concentration. Given their high sensitivity to environmental stimuli, zebrafish serve as a reliable model for early toxicity assessment (Figure 3A). The results showed no significant differences in morphological changes (Figure 3B), survival rate (Figure 3C), hatching rate (Figure 3D), or heart rate (Figure 3E) at concentrations up to 200 μM 5′-UMP. Based on these results, 200 μM was determined to be a non-toxic concentration and was subsequently used for downstream functional analyses.

The HFD-induced zebrafish model was used to validate the potential effect of 5′-UMP. Zebrafish larvae were divided into control, HFD, and 5′-UMP-treated HFD zebrafish. After 4 days of rearing, the zebrafish larvae were fixed and stained (Figure 4A). As shown in Figure 4B, Oil Red O and Nile Red staining revealed a marked increase in lipid deposition in HFD-fed larvae compared to the control group. In contrast, 5′-UMP treatment visibly reduced lipid accumulation, as evidenced by weaker Oil Red O staining and decreased Nile Red fluorescence intensity. Quantitative analysis on Day 7 further confirmed this observation (Figure 4C): HFD feeding significantly increased the number of zebrafish with high lipid content (p < 0.01), whereas 5′-UMP administration led to a shift toward lower lipid categories, with a significantly reduced proportion of high-lipid individuals (p < 0.01) and an increased proportion in the low-lipid group. These results indicate that 5′-UMP effectively attenuates lipid accumulation in vivo under HFD exposure.

Inhibiting adipogenesis and promoting the thermogenesis of adipocytes are two strategies against obesity. To evaluate the adipocyte cytotoxicity of 5′-UMP, we tested the viability using the CCK8 for preadipocytes exposed to graded concentrations of 5′-UMP (0, 100, 200, 400 μM) for 24, 48, and 72 h calculated as a percentage of control, untreated cells. The concentrations of 400 μM 5′-UMP and lower did not cause any changes in the relative cell viability (Supplementary Figure S1A). Then, we evaluated the apoptosis of human preadipocytes pretreated with 200 μM 5′-UMP for the indicated times. The flow cytometry-based apoptosis analysis exhibited no differences after 48 h treatment compared to that in the non-treated cells (Supplementary Figure S1B). These findings suggest that the selected dose of 5′-UMP was safe for preadipocytes.

To explore the cellular effects of 5′-UMP on lipid metabolism and mitochondrial function, human preadipocytes were treated with 5′-UMP during adipogenic differentiation. As shown in Figure 5A, Oil Red O staining revealed reduced intracellular lipid accumulation in 5′-UMP-treated cells compared to the control group, suggesting an inhibitory effect on adipogenesis. Mitochondrial respiration was assessed using Seahorse extracellular flux analysis. As shown in Figure 5B, 5′-UMP treatment significantly enhanced maximal OCR following FCCP stimulation (p < 0.01), indicating improved mitochondrial oxidative capacity. In line with this observation, thermogenic and mitochondrial biogenesis marker expression levels were significantly upregulated upon 5′-UMP treatment. Specifically, mRNA expression of PGC1α and UCP1 was markedly elevated (p < 0.01, p < 0.001, respectively; Figure 5C). Western blot analysis further confirmed increased UCP1 protein levels in the 5′-UMP group (Figure 5D), supporting the induction of a thermogenic gene program.

Collectively, these results demonstrate that 5′-UMP not only reduces lipid accumulation but also promotes thermogenic gene expression in human adipocytes.

To evaluate the effects of 5′-UMP on mitochondrial function, we performed MitoTracker staining in adipocytes. As shown in Figure 6A 5′-UMP treatment led to a marked increase in mitochondrial fluorescence intensity compared to the control group, suggesting enhanced mitochondrial content or activity.

Next, we utilized transgenic zebrafish larvae expressing mitochondrially targeted GFP [Tg(Xla. Eef1a1:mlsEGFP)] to evaluate the effect of 5′-UMP on mitochondrial abundance under HFD conditions. As shown in Figures 6B,C, mitochondrial fluorescence was significantly reduced in the HFD group, while 5′-UMP administration restored the fluorescence intensity to near-control levels (p < 0.01). Notably, zebrafish treated with 5′-UMP alone exhibited even higher mitochondrial fluorescence than controls (p < 0.01), indicating a mitochondrial-boosting effect of 5′-UMP in both physiological and HFD-induced conditions. We further analyzed the expression of mitochondrial biogenesis and fusion-related proteins to investigate the underlying mechanisms. Western blot analysis showed that 5′-UMP did not significantly alter the mitochondrial fission protein Drp1 expression in adipocytes (Figure 6D). In contrast, the expression of mitochondrial biogenesis regulator NRF1 and the mitochondrial fusion protein MFN2 was significantly upregulated following 5′-UMP treatment (Figure 6E), indicating enhanced mitochondrial biogenesis and fusion. Together, these results suggest that 5′-UMP enhances mitochondrial activity by promoting mitochondrial biogenesis and fusion, potentially through upregulation of NRF1 and MFN2.