Abstract SN-38 (7-ethyl-10-hydroxycamptothecin), the active metabolite of irinotecan, is a crucial anticancer agent frequently studied in drug delivery systems. Irinotecan (CPT-11) is used to treat various solid tumors but is associated with adverse effects such as nausea, vomiting, diarrhea, and steatohepatitis. However, the precise biochemical pathways underlying these side effects remain unclear. To explore SN-38’s toxic mechanisms and provide insights for clinical applications of SN-38 delivery systems, we performed untargeted metabolomics to assess metabolic changes in the lungs, heart, stomach, blood, spleen, intestine, liver, and kidneys of SN-38-exposed male mice. Mice were divided into two groups: SN-38 (20 mg/kg/day intraperitoneal) and control (blank solvent). Gas chromatography-mass spectrometry (GC-MS) identified significant metabolic disturbances in all tissues. Specifically, 24, 15, 12, 21, 35, 26, 18, and 28 differential metabolites were detected in the lungs, heart, stomach, blood, spleen, intestine, liver, and kidneys, respectively. KEGG pathway enrichment revealed significant changes in metabolic pathways across these organs, particularly in purine, pyrimidine, amino acid, and glyceric acid metabolism, implicating disruptions in protein synthesis, cellular homeostasis, energy metabolism, and antioxidant defenses. This study is the first to characterize SN-38’s multi-organ toxicity using metabolomics. Supplementary Information The online version contains supplementary material available at 10.1038/s41598-025-01753-1. Keywords: SN-38 (7-ethyl-10-hydroxycamptothecin), Toxicity mechanism, Gas chromatography mass spectrometry, Metabolomics Subject terms: Drug development, Metabolomics, Lipidomics Introduction Ethyl-10-hydroxycamptothecin (SN-38) is a highly potent anticancer agent and the active metabolite of irinotecan (CPT-11), a DNA topoisomerase I inhibitor commonly used in the treatment of various solid tumors, including colorectal, gastric, pancreatic, small cell lung, cervical, and ovarian cancers. Despite its effectiveness, irinotecan is associated with significant adverse effects, such as nausea, vomiting, severe diarrhea, and steatohepatitis^[34]1–[35]3. SN-38, which is 100 to 1,000 times more cytotoxic than irinotecan, is primarily responsible for the drug’s therapeutic activity^[36]4. However, the clinical application of SN-38 is significantly hindered by its poor solubility and chemical instability, which limit its efficient delivery for cancer treatment. To overcome these limitations and enhance SN-38’s therapeutic potential, advanced drug delivery systems, such as nanoparticles, liposomes, and micelles, have been developed. These systems improve the solubility and stability of SN-38, increasing its bioavailability and therapeutic effectiveness. Several of these drug delivery platforms are currently undergoing clinical trials^[37]5. SN-38 exerts its cytotoxic effects by inhibiting topoisomerase I, leading to DNA damage and the subsequent apoptosis of cancer cells. However, this mechanism is not unique to cancer cells and often results in collateral damage to healthy tissues, posing a significant challenge to its clinical application. Metabolomics, the comprehensive analysis of small-molecule metabolites in biological systems, has progressed considerably due to advancements in sophisticated analytical techniques and bioinformatics tools. Gas chromatography-mass spectrometry (GC-MS) is particularly advantageous for untargeted metabolomics studies due to its high separation efficiency, reproducibility, and access to extensive metabolite libraries. These features make GC-MS an ideal platform for investigating the complex metabolic changes associated with drug-induced toxicity^[38]6,[39]7. Although SN-38’s neurotoxic potential has been documented through targeted brain metabolomics^[40]8, its broader organ-specific toxicity profile remains unexplored. Here, we employ a multi-organ metabolomics approach to map systemic metabolic disruptions, revealing previously unrecognized hepatic and cardiac vulnerability hotspots.By examining metabolic changes across multiple organs, including the lungs, heart, stomach, serum, spleen, liver, intestine, and kidneys, this research seeks to elucidate the biochemical pathways involved in SN-38 toxicity. The findings are expected to provide a deeper understanding of SN-38’s toxicological effects and may help identify potential metabolic biomarkers. Ultimately, this research could contribute to developing strategies for mitigating irinotecan-induced toxicity and offer experimental support for the clinical translation of SN-38-based delivery systems. Methods Chemicals and reagents SN-38 was purchased from MedChemExpress (Shanghai, China). Dimethyl sulfoxide (DMSO) was obtained from Tianjin Yongda Chemical Reagent Co., Ltd. (Tianjin, China). Heptadecanoic acid (purity 98%), methanol (chromatographic grade), and pyridine were sourced from Macklin Biochemical Co., Ltd. (Shanghai, China). O-Methylhydroxylamine hydrochloride (purity 98%) was purchased from J&K Scientific Ltd. (Beijing, China). N, O-bis(trimethylsilyl)trifluoroacetamide (containing 1% trimethylchlorosilane) was obtained from Sigma-Aldrich (St. Louis, MO, USA). Purified water was provided by Hangzhou Wahaha Group Co., Ltd. (Hangzhou, China). Animal treatment In this study, the resource equation method was employed to determine the appropriate sample size. A total of 20 six-week-old male Kunming mice were purchased from Jinan Pengyue Experimental Animal Breeding Co., Ltd. (Jinan, China). The use of male mice helps minimize hormone-related variability, eliminating confounding factors such as the estrous cycle^[41]9. These mice were housed under controlled conditions at a stable temperature of 22 ± 2 °C, with humidity maintained between 60 and 70%, and a 12-hour light/dark cycle. Standard food and water were provided ad libitum throughout the study. A one-week acclimatization period was observed prior to the commencement of the main experiment. The mice were then randomly divided into control and test groups, with 10 mice in each group. For the experimental procedure, SN-38 was dissolved in a solution of sodium chloride and 5% DMSO to the desired concentration. The test group (n = 10) received an intraperitoneal injection of SN-38 (20 mg/kg) daily for one week. The drug administration schedule was based on existing studies^[42]10. The control group (n = 10) received an equal volume of the blank solvent mixture (DMSO and saline, 1:9 ratio). This study has been reported in accordance with the ARRIVE guidelines. This study was conducted in accordance with the “Guidelines for Ethical Review of Animal Welfare” (GB/T 35892–2018) and was approved by the Ethics Committee of the Affiliated Hospital of Jining Medical University (Approval No.: JNMC-2022-DW-041). Sample collection and preparation Two weeks later, the mice from both groups were euthanized by administering an overdose of sodium pentobarbital (80 mg/kg). Blood samples (1 mL) were taken through enucleation, and serum was separated by centrifugation at 4 °C at 4000 rpm for 10 min. The serum was immediately mixed with cold methanol (1:3 v/v) to inhibit enzymatic activity and stored at −80 °C.To prevent ongoing metabolic processes immediately after euthanasia, tissues (lung, heart, stomach, spleen, liver, intestine, kidney) were snap-frozen in liquid nitrogen within 30 s and stored at −80 °C. To prepare the serum samples, 100 µL of serum was combined with 350 µL of heptanoic acid (10 µg/mL, used as an internal standard) dissolved in methanol. Procedural blanks (methanol without biological material) were processed in parallel with each extraction batch. The mixture was then centrifuged at 14,000 rpm for 15 min at 4 °C. The resulting supernatant was evaporated under nitrogen gas at 37 °C. The dried residue was incubated with 80 µL of O-methylhydroxylamine hydrochloride dissolved in pyridine at 70 °C for 2 h. Subsequently, 100 µL of N, O-bis(trimethylsilyl)trifluoroacetamide (with 1% trimethylchlorosilane) was added, and the mixture was incubated again at 70 °C for 1 h. After a brief centrifugation at 14,000 rpm for 2 min at 4 °C, the solution was filtered through a 0.22 μm membrane and prepared for GC-MS analysis. For tissue sample preparation, 50 mg of each tissue was homogenized in methanol. The homogenate was then combined with 50 µL of heptanoic acid (10 µg/mL, internal standard) dissolved in methanol and centrifuged at 14,000 rpm for 15 min at 4 °C. The supernatant was incubated with 80 µL of O-methylhydroxylamine hydrochloride dissolved in pyridine at 70 °C for 90 min. This was followed by an additional incubation with N, O-bis(trimethylsilyl)trifluoroacetamide containing 1% trimethylchlorosilane for 60 min at the same temperature. After incubation, the samples were filtered through a 0.22 μm membrane. Quality control (QC) samples were prepared by mixing 10 µL of each sample from the control and SN-38 group, thoroughly mixed, and aliquoted into individual vials. To reduce the degradation of volatile derivatives, QC samples were sealed with nitrogen and stored at −80 °C. QC aliquots were injected every 10 experimental samples throughout the analytical sequence to monitor instrumental stability. The relative standard deviation (RSD) of QC samples should be less than 30%. Pathological examination Tissues from the lung, heart, stomach, spleen, intestine, liver, and kidney were collected and immersed in 4% paraformaldehyde for overnight fixation. After fixation, samples were dehydrated through a graded ethanol series (70% and 95% ethanol, 1 h each) and embedded in paraffin blocks. Sections of 4 μm thickness were prepared using a microtome (Leica, Germany). For deparaffinization, slides were cleared in xylene (three changes, 5 min each) and gradually hydrated through a graded ethanol series (100%, 96%, 80%, 70%, and 50%, 2 min per step). Hematoxylin and eosin (H&E, Sigma-Aldrich, USA) staining was performed to assess histopathological alterations. Whole-slide imaging was conducted using a PANNORAMIC MIDI slide scanner (3DHISTECH, Hungary), and digital sections were analyzed with Case Viewer 2.4 software (3DHISTECH, Hungary) to evaluate morphological changes across all organs. GC-MS-based metabolomics Metabolomics analysis was carried out using a 7000 C mass spectrometer connected to a 7890B gas chromatography system, featuring an HP-5MS fused silica capillary column (Agilent Technologies, USA). Prior to sample analysis, system suitability was verified by injecting methyl stearate (10 ppm in heptane) five times. The acceptance criteria required retention time RSD < 0.1 min, peak area RSD < 5%, and signal-to-noise ratio > 50:1. Samples were injected at a split ratio of 50:1, and helium served as the carrier gas with a flow rate of 1 mL/min. The injection port, transfer line, and ion source temperatures were set to 280 °C, 250 °C, and 230 °C, respectively. Electron impact ionization was employed at −70 eV, with a spectral acquisition rate of 20 spectra per second. The mass spectrometry analysis used electron spray ionization, scanning a mass-to-charge ratio (m/z) range from 50 to 800. Data processing and analysis Raw data were processed using Mass Hunter software (v.B.07.00; Agilent Technologies) with predefined parameters (peak width: 0.1 min, signal-to-noise threshold ≥ 3:1). Data matrix filtering removed features with > 20% missing values across samples. Missing values were imputed using 50% of the minimum positive value for each metabolite. Background signals from procedural and solvent blanks were subtracted. The metabolites identified in this study were classified as Level 2, representing putatively annotated compounds^[43]11.Metabolites were initially identified through the NIST 14 GC-MS library (similarity score > 80%) and validated against a custom library (“New Library”) constructed from quality control (QC) samples. All identifications underwent dual-blind manual verification. Putative annotations required spectral similarity > 80% and retention time consistency (deviation < 2%). Metabolites showing > 30% relative standard deviation (RSD) across QC replicates were excluded. Batch effects were mitigated via randomized sample sequences and QC-based LOESS normalization (MetaboAnalyst 6.0)^[44]12. Multivariate analyses (PCA/OPLS-DA) were performed in SIMCA-P 14.0 after peak area normalization. Model robustness was evaluated through 200 permutation tests (p < 0.05). Differentially expressed metabolites were selected using variable importance in projection (VIP) > 1.0 and p < 0.05 (two-tailed Student’s t-test with Benjamini-Hochberg correction). Downstream analyses included hierarchical clustering (heatmaps) and KEGG pathway enrichment (MetaboAnalyst 6.0)^[45]13–[46]16. Statistical analysis IBM SPSS Statistics 19 was used for normality (Shapiro-Wilk test) and homogeneity (Levene’s test) assessments. Only metabolites passing normality (p > 0.05) and homogeneity (p > 0.05) criteria were included in parametric statistical tests. Results GC-MS tics As illustrated in Fig. [47]1, the total ion chromatograms (TICs) of quality control (QC) samples demonstrate chromatographic stability across analytical batches, characterized by consistent retention time (RT) alignment and stable peak intensity profiles. The overlapping traces reflect sustained instrument performance during continuous analysis, with high peak capacity and baseline resolution throughout the analytical run. Fig. 1. [48]Fig. 1 [49]Open in a new tab Representative gas chromatography–mass spectrometry (GC–MS) total ion chromatograms (TICs) from the quality control (QC) samples. Multivariate analysis of metabolomics data Unsupervised PCA demonstrated tissue-specific metabolic divergence between SN-38-treated and control groups across seven organs (cumulative R2X: 0.28–0.59), with maximal separation in hepatic (R2X = 0.519) and pulmonary (R2X = 0.589) tissues (Table [50]1). Supervised OPLS-DA exhibited strong discriminative performance, showing near-complete variance explanation in cardiac (R2Y = 0.999) and splenic (R2Y = 1.000) models (Q2: 0.75–0.91; Table [51]1). Permutation testing (n = 200) confirmed model validity (all permuted Q2 intercepts < 0; p < 0.05)(Fig. [52]2), though caution is advised when interpreting cardiac/splenic signatures due to potential overfitting risks inherent to high-dimensional omics models^[53]17. Table 1. PCA and OPLS-DA parameter scores. Tissues Model type R2X(cum) R2Y(cum) Q2(cum) Lung PCA 0.589 / 0.413 OPLS-DA 0.578 0.994 0.885 Heart PCA 0.296 / −0.0822 OPLS-DA 0.298 0.999 0.86 Stomach PCA 0.368 / 0.074 OPLS-DA 0.28 0.964 0.747 Serum PCA 0.327 / 0.0754 OPLS-DA 0.346 0.998 0.896 Spleen PCA 0.436 / 0.215 OPLS-DA 0.507 1 0.88 Intestines PCA 0.485 / 0.161 OPLS-DA 0.424 0.984 0.788 Liver PCA 0.519 / 0.328 OPLS-DA 0.468 0.983 0.908 Kidney PCA 0.41 / 0.179 OPLS-DA 0.328 0.984 0.86 [54]Open in a new tab Fig. 2. [55]Fig. 2 [56]Open in a new tab OPLS-DA score plots and results from permutation tests. Identification of metabolic changes in samples Metabolite identification was conducted using a well-established mass spectrometry database, beginning with spectral comparisons from the Agilent Fiehn GC/MS Metabolomics RTL Library. The criteria for identifying metabolites included the variable importance in projection (VIP) score and p-value, with a VIP score above 1 and a p-value below 0.05 considered significant indicators of potential metabolites. Additionally, a fold-change greater than 1 signified an upward trend in metabolite levels, while a fold-change less than 1 indicated a downward trend. Table [57]2 summarizes the metabolic alterations observed across different tissues in the study groups, and the comparison of the distribution of biomarkers in each tissue are shown in Fig. [58]3.In total, 69 metabolites were identified as significantly altered in serum and tissues following SN-38 treatment. Table 2. List of metabolites that differed significantly in lung, heart, stomach, serum, spleen, intestines, liver and kidney samples from mice after SN-38 treatment. Metabolite HMDB ID Lung Heart Stomach Serum Spleen Intestine Liver Kidney VIP p-value Trend VIP p-value Trend VIP p-value Trend VIP p-value Trend VIP p-value Trend VIP p-value Trend VIP p-value Trend VIP p-value Trend L-Alanine HMDB0000161 1.43185 0 ↓ 1.5608 0.003 ↓ 1.15808 0.023 ↑ 1.68741 0 ↓ 1.66198 0.002 ↓ 1.27615 0 ↓ Ethanolamine HMDB0000149 1.42849 0 ↓ 1.26673 0 ↓ 1.4888 0.007 ↓ 1.33237 0.003 ↓ 1.18623 0 ↓ Glycine HMDB0000123 1.40574 0 ↓ 2.03319 0 ↓ 1.62385 0 ↓ 1.41217 0.002 ↓ 1.3884 0 ↓ Uracil HMDB0000300 1.25803 0 ↓ 1.52355 0.001 ↓ 1.1417 0 ↓ 1.4811 0 ↓ 1.43853 0.001 ↑ Serine HMDB0062263 1.00729 0 ↓ 1.32922 0 ↓ 1.80309 0.001 ↑ 1.39477 0 ↓ 1.44171 0 ↓ 1.48091 0.005 ↓ L-Threonine HMDB0000167 1.16436 0 ↓ 1.28401 0.022 ↑ 1.6348 0 ↓ 1.51569 0 ↓ 1.49606 0.004 ↓ L-5-Oxoproline HMDB0000267 1.21667 0 ↓ 1.6697 0 ↑ 1.44595 0 ↓ 1.36197 0.008 ↓ 1.24731 0 ↓ L-Aspartic acid HMDB0000191 1.59829 0 ↓ 2.03692 0.001 ↑ 1.667 0 ↓ 1.1527 0.012 ↓ L-Hydroxyproline HMDB0000725 1.42872 0 ↓ Taurine HMDB0000251 1.64232 0 ↓ 1.54467 0.043 ↓ 1.56647 0.001 ↓ Erythritol HMDB0002994 1.06689 0.001 ↓ Myristic acid HMDB0000806 1.12475 0 ↓ 1.29796 0 ↓ 1.01626 0.002 ↓ 1.2056 0.002 ↓ Myo-Inositol HMDB0000211 1.04457 0 ↓ 1.03526 0.001 ↑ Lactic Acid HMDB0000190 1.50075 0 ↓ 2.48085 0 ↓ 1.28865 0.027 ↓ 1.82329 0 ↓ 1.49016 0.009 ↓ 1.81304 0 ↓ Glyceric acid HMDB0000139 1.25863 0 ↓ 1.06429 0.012 ↓ Malic acid HMDB0000744 2.02242 0 ↓ 2.29639 0 ↓ 1.99384 0 ↓ 1.65963 0 ↓ 1.6327 0.001 ↓ 1.58496 0.001 ↓ Phosphorylethanolamine HMDB0000224 2.03988 0 ↓ 1.20574 0.004 ↓ 2.13797 0.001 ↓ 2.0118 0 ↓ 2.50558 0 ↓ L-Lysine HMDB0000182 1.5953 0 ↓ 1.22089 0.016 ↓ 1.84087 0 ↓ 1.59551 0.024 ↓ 1.0835 0 ↓ 1.81673 0 ↓ L-Tyrosine HMDB0000158 1.12706 0 ↓ 2.05449 0 ↓ 1.58114 0 ↓ 1.52078 0.004 ↓ 1.12152 0 ↓ 1.06791 0.004 ↓ 11-Octadecenoic acid HMDB0003231 1.19638 0 ↓ Arachidic acid HMDB0002212 1.0138 0 ↓ 1.01931 0.002 ↓ 1.0305 0.047 ↓ MG(0:0/18:1(9Z)/0:0) HMDB0011537 1.15517 0.04 ↓ MG(0:0/18:0/0:0) HMDB0011535 1.13729 0 ↓ 1.43048 0 ↓ L-Valine HMDB0000883 1.55952 0.004 ↑ 1.07133 0.034 ↓ 1.59536 0 ↓ 1.69261 0.002 ↓ L-Proline HMDB0000162 2.03295 0 ↑ 1.25163 0.001 ↓ 1.69751 0 ↓ Pipecolic acid HMDB0000070 1.63311 0.002 ↑ L-Glutamic acid HMDB0000148 1.63346 0.004 ↓ 1.44133 0.009 ↑ 1.69197 0 ↓ 1.52771 0 ↓ 1.00108 0.003 ↓ 1.12096 0.003 ↓ Petroselinic acid HMDB0002080 1.15756 0.023 ↓ 1.4087 0 ↓ 1.15674 0.006 ↑ Succinic acid HMDB0000254 2.12961 0.001 ↓ 1.81357 0.001 ↓ 2.12777 0 ↓ Phenylalanine HMDB0000159 1.38987 0.001 ↓ 1.55429 0.005 ↓ 1.08807 0.002 ↓ L-Isoleucine HMDB0000172 1.53587 0 ↓ 1.74659 0.002 ↓ 1.11152 0.001 ↓ L-Methionine HMDB0000696 1.70484 0 ↓ 1.3893 0.001 ↓ 1.45341 0 ↓ 1.59407 0.004 ↓ L-Ornithine HMDB0000214 1.56836 0.004 ↓ 1.56418 0 ↓ L-Tryptophan HMDB0000929 1.55396 0.001 ↓ 2.11505 0.001 ↓ Cholesterol HMDB0000067 1.4858 0.004 ↓ 1.2265 0 ↓ 1.27767 0.001 ↑ 5-Aminovaleric acid HMDB0003355 1.67392 0.009 ↓ Dodecanoic acid HMDB0000638 1.12705 0.043 ↓ Stearic acid HMDB0000827 1.18693 0.001 ↓ 1.25653 0 ↓ 1.23325 0.001 ↑ Urea HMDB0000294 1.44196 0.001 ↓ L-Arabitol HMDB0001851 1.43152 0.005 ↓ 1.28594 0.006 ↓ L-Leucine HMDB0000687 1.53919 0 ↓ 1.67731 0.003 ↓ 1.02754 0.003 ↓ D-Mannitol HMDB0000765 1.13177 0.001 ↑ 1.25664 0.001 ↓ Asparagine HMDB0000168 1.94572 0 ↓ 1.4482 0.002 ↓ D-Glucose HMDB0000122 2.3219 0 ↓ Gluconic acid HMDB0000625 1.05046 0.004 ↓ Oxalic acid HMDB0002329 1.21659 0.001 ↑ Ribitol HMDB0000508 1.0579 0.009 ↓ Hexadecane HMDB0033792 1.58377 0 ↑ Docosahexaenoic acid HMDB0002183 1.41852 0.006 ↓ 1.01199 0.004 ↓ 1.32795 0.001 ↑ 1.33549 0.001 ↓ 2-Aminooctanoic acid HMDB0000991 1.89931 0 ↓ D-Arabitol HMDB0000568 1.2075 0.003 ↓ 2-Phenylbutyric acid HMDB00329 1.12875 0.012 ↓ Glycerol HMDB0000131 1.30896 0 ↓ 1.1409 0.001 ↓ 1.30749 0 ↓ 1.08352 0 ↓ 1.46021 0 ↓ Niacinamide HMDB0001406 1.00698 0 ↓ Hypoxanthine HMDB0000157 1.17903 0.005 ↓ Adenine HMDB0000034 1.15503 0.033 ↑ MG(16:0/0:0/0:0) HMDB0011564 1.01955 0.005 ↓ 1.49898 0 ↓ Inosine HMDB0000195 1.32579 0.002 ↑ Adenosine HMDB0000050 1.91973 0.001 ↑ Citric acid HMDB0000094 1.64524 0.001 ↓ Beta-Alanine HMDB0000056 1.6621 0 ↓ 4-Hydroxybenzyl alcohol HMDB0011724 1.26064 0.003 ↓ Palmitic acid HMDB0000220 1.19475 0.001 ↓ 1.0434 0 ↑ Ribothymidine HMDB0000884 1.2569 0.003 ↓ Uridine HMDB0000296 1.32668 0.005 ↓ N-Dodecane HMDB0031444 1.22568 0.038 ↓ 1.42558 0.001 ↑ Arachidonic acid HMDB0001043 1.12277 0.005 ↓ 1.55127 0 ↑ L-Galactose HMDB0033704 1.33935 0.012 ↓ muco-Inositol HMDB0062138 1.54508 0.009 ↓ [59]Open in a new tab Fig. 3. Fig. 3 [60]Open in a new tab Venn diagram of the metabolite distribution in the lung, heart, stomach, spleen, intestines, liver, and kidney between the control and SN-38 groups. Note: the numbers in the figure represent the same metabolites among different matrices. In lung tissue, 24 metabolites exhibited a downregulated trend, including amino acids, lactate, fatty acids, and others. In the heart, 15 metabolites were identified, with 4 showing upregulation (adenine, creatinine, adenosine) and 11 showing downregulation. In the stomach, SN-38 treatment led to alterations in 12 metabolites, with 8 upregulated (including alanine, valine, proline, serine, aspartic acid, glutamic acid, and threonine) and 4 downregulated (such as ethanolamine and lactate). In the liver, 12 metabolites were upregulated, including palmitic acid, arachidonic acid, stearic acid, cholesterol, and uracil, while 6 metabolites, such as amino acids and galactose, were downregulated. Additionally, 21, 35, 26, and 18 metabolites were downregulated in the serum, spleen, intestine, and kidney, respectively. Clustering analysis of the differential metabolites between the SN-38 and control groups further supported these findings (Fig. [61]4). Fig. 4. [62]Fig. 4 [63]Open in a new tab Heatmap of differential metabolites in the SN-38 group compared to controls. Each color in the heatmap represents metabolite changes: blue indicates down-regulation, and red indicates up-regulation. In the heatmap, columns represent samples and rows represent metabolites. Metabolic pathway analysis To further evaluate the metabolic changes between the SN-38 and control groups, MetaboAnalyst 6.0 ([64]http://www.metaboanalyst.ca) and the KEGG database ([65]http://www.kegg.jp) were employed to identify the metabolic pathways impacted by SN-38 treatment. Pathways with a raw p-value of less than 0.05 and an impact value greater than 0 were considered potentially disrupted (see Table [66]3; Fig. [67]5). The analysis revealed that SN-38 exposure significantly affected several metabolic pathways, including those related to various amino acids, glutathione, purine, pyrimidine, pyruvate, glyoxylate and dicarboxylate metabolism, as well as glycerolipid, glycerophospholipid, and sphingolipid metabolism. The detailed pathway analysis results are presented in Table [68]3, with a summary in Fig. [69]5, and a comprehensive metabolic network map in Fig. 7. Table 3. Metabolic pathways related to SN-38 intervention mechanisms. Metabolic Pathway Lung Heart Stomach Serum Spleen Intestine Liver Kidney Raw p FDR Impact Raw p FDR Impact Raw p FDR Impact Raw p FDR Impact Raw p FDR Impact Raw p FDR Impact Raw p FDR Impact Raw p FDR Impact Glycine, Serine and Threonine Metabolism 0.0088792 0.35053 0.25139 0.037047 0.36597 0.48837 0.023345 0.2312 0.22838 0.0088792 0.11839 0.51302 0.034891 0.27912 0.51302 0.014834 0.16953 0.51302 Purine Metabolism 0.0030174 0.080463 0.04324 Glyoxylate and Dicarboxylate Metabolism 0.0074842 0.35053 0.1217 0.00013756 0.011005 0.14815 0.020803 0.2312 0.04233 3.6323E-5 0.0029058 0.1799 0.004246 0.056614 0.14815 0.012552 0.16736 0.14815 7.4154E-04 0.019774 0.18519 Glutathione Metabolism 0.0017575 0.070299 0.11548 0.0051119 0.08179 0.10839 0.020856 0.18539 0.11548 4.3771E-04 0.017508 0.11548 Alanine, Aspartate and Glutamate Metabolism 0.025758 0.36597 0.19712 2.7399E-05 0.0021919 0.42068 0.0051119 0.08179 0.19712 0.0025701 0.045573 0.42068 4.3771E-04 0.017508 0.19712 Sphingolipid Metabolism 0.033099 0.36597 0.00563 Citrate Cycle (TCA Cycle) 0.0019024 0.076095 0.16723 0.029494 0.23595 0.07685 Arginine and Proline Metabolism 0.02601 0.2312 0.01744 0.010418 0.11907 0.18139 Phenylalanine, Tyrosine and Tryptophan Biosynthesis 0.0026459 0.045573 1 0.0013941 0.052733 1 0.028533 0.80117 0.5 1.0636E-03 0.021271 1 Pyruvate Metabolism 0.034982 0.39979 0.0283 0.034982 0.23321 0.0283 0.049 0.392 0.0283 0.038313 0.27864 0.0283 Glycerophospholipid Metabolism 0.041171 0.36597 0.03747 Glycerolipid Metabolism 0.017527 0.35053 0.33022 0.019254 0.17114 0.33022 Arginine Biosynthesis 0.0040973 0.085769 0.11675 0.013501 0.13501 0.17766 0.0028483 0.045573 0.11675 0.014845 0.16965 0.17766 Primary Bile Acid Biosynthesis 0.02037 0.16296 0.07802 Beta-Alanine Metabolism 0.0093852 0.10726 0.39925 Pantothenate and CoA Biosynthesis 6.8742E-04 0.027497 0.04762 Phenylalanine Metabolism 0.018293 0.18293 0.35714 0.0098689 0.1579 0.35714 0.0075883 0.10118 0.35714 Nicotinate and nicotinamide metabolism 0.015457 0.35053 0.1943 Cysteine and methionine metabolism 0.035051 0.36597 0.1263 Galactose metabolism 0.04704 0.28948 0.03499 Pyrimidine metabolism 0.049536 0.30159 0.09246 [70]Open in a new tab Fig. 5. [71]Fig. 5 [72]Open in a new tab Summary of the pathway analysis conducted using MetaboAnalyst 6.0:a Glycine, Serine and Threonine Metabolism; b Purine Metabolism; c Glyoxylate and Dicarboxylate Metabolism; d Glutathione Metabolism; e Alanine, Aspartate and Glutamate Metabolism; f Sphingolipid Metabolism; g Citrate Cycle (TCA Cycle); h Arginine and Proline Metabolism; i Phenylalanine, Tyrosine and Tryptophan Biosynthesis; j Pyruvate Metabolism; k Glycerophospholipid Metabolism; l Glycerolipid Metabolism; m Arginine Biosynthesis; n Primary Bile Acid Biosynthesis; o Beta-Alanine Metabolism; p Pantothenate and CoA Biosynthesis; q Phenylalanine Metabolism; r Nicotinate and nicotinamide metabolism; s Cysteine and methionine metabolism; t Galactose metabolism; u Pyrimidine metabolism. Pathological examination Microscopic examination revealed preserved tissue architecture across all organs: alveolar integrity in the lung, orderly myocardial fiber alignment in the heart, intact gastric mucosa, normal splenic pulp distribution, unaltered intestinal villi, regular hepatic trabeculae, and structurally sound renal glomeruli and tubules. Quantitative assessment of characteristic regions demonstrated no significant histopathological differences between SN-38-treated and control groups (p > 0.05), indicating no overt organ damage (Fig. [73]6). Fig. 6. [74]Fig. 6 [75]Open in a new tab Tissues stained with hematoxylin and eosin include the lungs, heart, stomach, spleen, intestines, liver and kidney. Discussion Clinical evidence highlights that the toxicity of irinotecan and its active metabolite, SN-38, predominantly impacts the gastrointestinal tract—manifesting as nausea, vomiting, severe diarrhea, and mucositis—as well as the liver, particularly in the form of steatohepatitis. However, there are also reports of drug-induced toxicity in other organs, including pulmonary fibrosis and lung inflammation^[76]18,[77]19, as well as sinus bradycardia in patients with small round cell carcinoma^[78]20. In this study, we employed multivariate statistical analysis to assess the effects of SN-38 on major organs and tissues in mice. The novelty of this research lies in its investigation of SN-38’s pharmacological mechanisms from a metabolomics perspective, offering a comprehensive evaluation of the metabolic disruptions triggered by SN-38 exposure in mice using GC-MS analysis. Our findings demonstrate that SN-38 induces significant metabolic disturbances in the blood and various tissues. Specifically, 24 metabolites were identified in the lungs, 15 in the heart, 12 in the stomach, 21 in the blood, 35 in the spleen, 26 in the intestine, 18 in the liver, and 28 in the kidney. KEGG pathway enrichment analysis revealed statistically significant alterations in 5 metabolic pathways in the lungs, 8 in the heart, 5 in the stomach, 10 in the blood, 10 in the spleen, 5 in the intestine, 1 in the liver, and 9 in the kidney. To elucidate the distinct pharmacological mechanisms resulting from these varied metabolic changes across different organs, the following sections provide a detailed discussion of these findings. (Fig. [79]7) Fig. 7. [80]Fig. 7 [81]Open in a new tab Schematic diagram illustrating metabolic pathways in tissues affected by SN-38 processing. Metabolites marked in red represent potential biomarkers identified in this study.Directional arrows (▲ for upregulation, ▼ for downregulation) were introduced to unambiguously distinguish metabolite expression trends; A tissue-specific color-coding system was applied to annotate the organ origin of metabolic perturbations.This figure includes data adapted from the KEGG database (© Kanehisa Laboratories). The KEGG pathway map is used with permission from the Kanehisa Laboratory^[82]14–[83]16. The reduction in ethanolamine and phosphoethanolamine levels suggests a disruption in phospholipid metabolism, which is essential for maintaining the integrity of cell membranes. Glycerol, a crucial component in triglyceride synthesis, also plays a pivotal role in the formation of phospholipids, the primary lipid constituents of pulmonary surfactants. Disruptions in glycerol metabolism may impair the production of phosphatidylcholine, which could result in diminished surfactant synthesis and functionality. This decrease could increase alveolar surface tension, potentially leading to alveolar collapse and reduced lung compliance^[84]21. The depletion of these metabolites in the lungs may compromise cell membrane stability, making lung tissues more vulnerable to injury. Furthermore, the observed decrease in taurine levels suggests a weakened antioxidant defense in lung tissue, potentially leading to elevated oxidative stress and subsequent lung damage^[85]22. Inositol, a critical nutrient known for promoting cell survival and proliferation, is widely used in the treatment of conditions such as respiratory distress syndrome (RDS) and Alzheimer’s disease^[86]23. Additionally, proline and hydroxyproline, key components of collagen, play vital roles in lung tissue repair. The reduction in hydroxyproline levels indicates a hindered repair process, which could further aggravate lung injury^[87]24. Adenosine plays a crucial role in regulating coronary blood flow and protecting the heart from ischemic damage. Its accumulation is directly implicated in the pathophysiology of atrial fibrillation and neurocardiogenic syncope. Additionally, adenosine is fundamental in the adaptive responses to pulmonary hypertension and heart failure, with its primary effects including the reduction of heart rate, coronary vasodilation, and lowering of blood pressure^[88]25. Under stress conditions, adenosine levels rise, acting as both a signaling molecule and a cardioprotective agent, indicating a compensatory response to SN-38-induced cardiotoxicity^[89]26. In this study, the observed increase in the adenosine breakdown product, inosine, supports the hypothesis of enhanced purine metabolism. The concurrent elevation of adenine and inosine suggests an imbalance in purine metabolism, which may lead to altered nucleotide synthesis and increased oxidative stress, exacerbating cardiac injury. A disruption in glutathione metabolism is a significant contributor to the pathogenesis of myocardial injury. The reduction of key metabolites, such as glutamate and glycine, causes dysregulation in glutathione metabolism, contributing to various cardiac issues, including myocardial injury, ferroptosis, metabolic disturbances, impaired cardiac function, and cardiac remodeling^[90]27,[91]28. Previous research has shown that irinotecan significantly impacts oxidative stress parameters in cardiac tissue, with increased lipid peroxidation (elevated TBARS levels), reduced activity of antioxidant enzymes (SOD, CAT, GPx), and decreased glutathione (GSH) levels^[92]29. These findings underscore the critical role of SN-38-induced oxidative stress in the development of cardiotoxicity. Proline is vital for protein synthesis and cell growth, as well as for osmoregulation, protein stability, cellular bioenergetics, and antimicrobial or antifungal activity^[93]30. Elevated levels of proline may indicate an attempt by the stomach to repair damage caused by SN-38. However, if these repair processes become dysregulated, chronic upregulation of such pathways might lead to pathological changes^[94]31. Abnormal levels of L-alanine and L-glutamate suggest disruptions in amino acid metabolism, which plays a crucial role in energy production and neurotransmitter synthesis. These disruptions could exacerbate oxidative stress, further damaging the gastric mucosa^[95]32. Glutamate stimulates gastric acid secretion, and chronic or excessive glutamate levels may result in hyperacidity, causing gastric mucosal damage and potentially leading to gastric ulcers. Previous studies have shown that NMDA receptor agonists, which interact with glutamate receptors, increase the incidence of gastric ulcers. Excessive glutamate release or receptor dysfunction may also contribute to gastrointestinal disturbances, such as diarrhea or constipation^[96]33. Glycine is recognized as a key protective factor against gastrointestinal diseases, particularly in the context of chemically induced colitis and inflammatory bowel disease (IBD). Research suggests that glycine may have protective effects against irinotecan-induced mucositis. Elevated glycine levels in patients with active IBD further highlight its role in alleviating colitis and reducing intestinal inflammation^[97]34. However, the observed decline in glycine levels in the intestines following SN-38 treatment could worsen intestinal inflammation, emphasizing the need for glycine regulation in managing inflammation. Aspartate is associated with chronic visceral pain and hypersensitivity in patients with colonic inflammation. The RGD (arginine-glycine-aspartate) peptide motif exhibits anti-inflammatory properties in the intestine, helping to maintain gut barrier integrity by upregulating genes such as trefoil factor-3 and mucin^[98]30. Consequently, aspartate may serve as a biomarker for the severity of inflammatory bowel disease. Additionally, stearic acid has been shown to reduce inflammation caused by bile duct ligation (BDL) by inhibiting inflammatory cell recruitment and NF-κB activation. Previous studies have confirmed that stearic acid may function as a protective metabolite against irinotecan-induced colitis^[99]34. The elevation of palmitic acid, arachidonic acid, stearic acid, and docosahexaenoic acid (DHA) in the liver of the SN-38 treatment group suggests significant disruptions in lipid metabolism. Arachidonic acid, in particular, is a ω−6 polyunsaturated fatty acid that is abundant in cell membrane phospholipids and lipid droplets. Its activation of macrophages, specifically Kupffer cells, triggers the release of pro-inflammatory cytokines, such as tumor necrosis factor-α (TNF-α) and interleukin-6 (IL-6)^[100]35. Metabolites of arachidonic acid, produced through the lipoxygenase (LOX) and cyclooxygenase (COX) pathways, including prostaglandins PGE2 and PGD2, further exert pro-inflammatory effects^[101]36. Additionally, the COX pathway influences the peroxisome proliferator-activated receptor gamma (PPARγ), which is implicated in the development of non-alcoholic fatty liver disease (NAFLD). PPARγ is closely associated with insulin resistance and hepatic steatosis, contributing to the pathogenesis of NAFLD^[102]37. Palmitic acid, the most prevalent saturated fatty acid, is strongly linked to the development of fatty liver disease. Elevated palmitic acid levels can lead to excessive fat accumulation within hepatocytes, thereby promoting fatty liver formation^[103]38. Pyrimidine nucleotides, which are essential for DNA synthesis, also play a critical role in cellular signaling and energy metabolism, maintaining cellular homeostasis. Alterations in pyrimidine metabolites, such as thymidine, uracil, and deoxycytidine, have been closely associated with disruptions in pyrimidine metabolism. Studies suggest that such disruptions may contribute to liver damage, resulting in significant hepatocellular dysfunction, including increased oxidative stress, impaired bile acid synthesis, and dysregulation of the glutathione (GSH) biosynthesis pathway^[104]39. Serine is a crucial source of N5,N10-methylene tetrahydrofolate (MTHF), and its deficiency can lead to MTHF depletion, impairing DNA synthesis and cell division. Additionally, serine acts as an intermediate in the tricarboxylic acid (TCA) cycle, playing a vital role in energy metabolism. A deficiency in serine can result in energy metabolism disorders, thereby disrupting cellular functions^[105]40. L-serine also reduces the secretion of pro-inflammatory cytokines, including IL-1, IL-17, interferon-γ, and TNF-α, while inhibiting neutrophil-mediated inflammation in macrophages^[106]41. The downregulation of serine suggests impaired respiratory function^[107]42, and as a key amino acid involved in protein and phospholipid synthesis, its deficiency may further compromise lung tissue repair processes. A reduction in serum taurine levels may exacerbate systemic oxidative stress and inflammatory responses, which are common in chemotherapy-induced systemic toxicity. Taurine is known for its anti-inflammatory properties, as it inhibits the production of various inflammatory mediators and plays a protective role in intestinal mucosal cells, helping to mitigate damage caused by inflammation and oxidative stress. Studies indicate that decreased serum taurine levels may amplify inflammatory responses, potentially worsening the symptoms and tissue damage associated with inflammatory bowel disease (IBD)^[108]43. Notably, while SN-38 administration induced marked metabolic disturbances across all eight organs examined, subsequent histopathological evaluation demonstrated no gross morphological alterations in these tissues. This divergence implies that SN-38 toxicity primarily manifests as subcellular or molecular pathway dysregulation, rather than immediate structural compromise. Specifically, the observed metabolomic shifts may represent early adaptive metabolic reprogramming or compensatory biochemical adjustments to cellular stress prior to irreversible injury. Conventional histological methods, which detect toxicity only at advanced stages (e.g., necrosis or inflammatory infiltration), lack the sensitivity to identify such subclinical perturbations. In contrast, untargeted GC–MS metabolomics captures subtle metabolic imbalances predictive of incipient organ damage^[109]44. These findings highlight the predictive potential of metabolomic profiling as a sensitive preclinical screening tool, offering mechanistic insights into subclinical toxicity that precedes histologically detectable lesions.Nonetheless, the absence of physiological and functional validation remains a limitation. Future studies will be directed toward integrating organ-specific functional assays and phenotypic evaluations to confirm the biological consequences of these metabolic changes and further clarify the toxicological profile of SN-38. Conclusion A comprehensive analysis of the multi-organ toxicity induced by SN-38 reveals a complex interaction of metabolic disruptions that significantly impair organ function and systemic homeostasis. The lungs, heart, gastrointestinal tract, liver, and blood are the most affected, with purine, pyrimidine, amino acid, and glycerolipid metabolism playing central roles in the toxic effects. These metabolic disturbances suggest impairments in protein synthesis, cellular homeostasis, energy metabolism, and antioxidant defense mechanisms. This integrative metabolomics approach provides critical insights into the mechanisms underlying SN-38 toxicity and highlights potential biomarkers for monitoring and mitigating its adverse effects. Electronic supplementary material Below is the link to the electronic supplementary material. [110]Supplementary Material 1^ (211.4KB, pdf) [111]Supplementary Material 2^ (169.4KB, pdf) Author contributions X.Z., C.C., and G.H. designed and planned the research. X.Z. and Y.H. wrote the main manuscript text. J.L. and B.K. conducted the formal analysis. X.Z., Y.H., and J.L. carried out the investigation. X.Z. handled data curation and prepared all figures. Y.H. was responsible for software implementation. X.Z., C.C., and G.H. contributed to the review and editing of the manuscript. C.C. and G.H. supervised the project and managed its administration. All authors reviewed the manuscript and approved the final submitted version. Funding This research was funded by the National Natural Science Foundation of China (Grant No.82302795); Natural Science Foundation of Shandong Province [ZR2023QH094]; Jining Key Research and Development Program [2024YXNS126]; Beijing Huikang Charity Foundation [BHCF20240620]. Data availability The untargeted metabolomics data generated in this study have been deposited in the OMIX repository of the China National Center for Bioinformation ([112]https://ngdc.cncb.ac.cn/omix) under accession number OMIX009615. The dataset will be made publicly available without restriction upon publication of this article.Additional information will be made available upon request (hanguangkui1997@163.com). Declarations Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References