Abstract

Purpose

   SN-38 (7-ethyl-10-hydroxycamptothecin), the active metabolite of
   irinotecan, has been extensively studied in drug delivery systems.
   However, its impact on neural metabolism remains unclear. This study
   aims to investigate the toxic effects of SN-38 on mouse brain
   metabolism.

Methods

   Male mice were divided into an SN-38 group and a control group. The
   SN-38 group received SN-38 (20 mg/kg/day) via intraperitoneal
   injection, while the control group was given an equal volume of a blank
   solvent mixture (DMSO and saline, ratio 1:9). Gas chromatography-mass
   spectrometry (GC-MS) was employed to analyze differential metabolites
   in the cortical and hippocampal regions of the SN-38-treated mice.

Results

   SN-38 induced metabolic disturbances in the central nervous system.
   Eighteen differential metabolites were identified in the hippocampus
   and twenty-four in the cortex, with six common to both regions. KEGG
   pathway enrichment analysis revealed statistically significant
   alterations in six metabolic pathways in the hippocampus and ten in the
   cortex (P<0.05).

Conclusion

   This study is the first to demonstrate the neurotoxicity of SN-38 in
   male mice through metabolomics. Differential metabolites in the
   hippocampal and cortical regions were closely linked to purine
   metabolism, pyrimidine metabolism, amino acid metabolism, and glyceride
   metabolism, indicating disruptions in the blood-brain barrier, energy
   metabolism, and central signaling pathways.

   Keywords: SN-38 (7-ethyl-10-hydroxycamptothecin), biomarker, gas
   chromatography mass spectrometry, metabolomics, toxicity mechanism

Introduction

   Ethyl-10-hydroxycamptothecin (SN-38) is the active metabolite of
   irinotecan (camptothecin-11, also known as hydrochloride irinotecan).
   Irinotecan, a broad-spectrum cytotoxic anticancer drug, is widely used
   to treat various solid tumors and is frequently combined with other
   chemotherapeutic agents in treating cancers of the central nervous
   system (CNS).SN-38’s antitumor activity is 100 to 1000 times more
   potent than CPT-11, leading to its extensive study in various drug
   delivery and controlled release systems.[36]1 Several SN38 prodrugs,
   including DTS108, EZN-2208, and SN38 antibody-drug conjugates (ADCs)
   such as IMMU132 and IMMU130, are currently in clinical trials.[37]2
   Previously, our team developed an effective antitumor drug release
   system, the SN-38 loaded PCL/GT electrospun film, and investigated its
   cytotoxicity against gliomas.[38]3

   The central mechanism of SN-38’s anti-tumor activity involves
   inhibiting DNA topoisomerase I, resulting in DNA damage and cell death
   in rapidly proliferating normal cells, including bone marrow and
   intestinal basal cells. Studies have reported common hematologic and
   gastrointestinal side effects in cancer patients treated with
   Irinotecan.[39]4 Rarer toxicities, including articulatory disorders,
   severe systemic weakness, paralysis, and aphasia induced by irinotecan,
   have also been reported.[40]5 However, research on SN-38’s effects on
   brain tissue is limited. Thus, the aim of this experiment is to
   thoroughly evaluate the neurotoxicity of SN-38 by examining its
   metabolite levels in the hippocampus and cerebral cortex, and to
   initially explore the specific mechanisms of SN-38-induced toxicity.

   Metabolomics plays a crucial role in the realm of systems biology. By
   combining it with other omics data such as genomics, transcriptomics,
   and proteomics, a more holistic molecular profile is achieved. This
   amalgamation offers a deeper understanding of various states such as
   disease pathology, pharmacological impacts, and responses to stress,
   while revealing underlying molecular processes. The use of multi-omics
   integration is prevalent across various domains, encompassing
   fundamental medical science, clinical diagnostics, pharmaceutical
   innovation, agricultural enhancement, and research into stress
   tolerance.[41]6–8

   Building on neurotissue engineering research, our team employed an
   established metabolomics research platform to determine the overall
   impact of SN-38 on brain metabolites and metabolic pathways. This
   approach clarified the metabolic profile following SN-38 exposure and
   offered experimental evidence for clinical translation.

Materials and Methods

Chemicals and Reagents

   SN-38 was obtained from MedChemExpress, Shanghai, China. Dimethyl
   sulfoxide (DMSO) was obtained from Tianjin Yongda Chemical Reagents,
   Tianjin, China. Heptadecanoic acid (purity 98%), methanol
   (chromatographic grade), and pyridine were obtained from Shanghai
   Macklin Biochemical, 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 purchased from Hangzhou Wahaha Company,
   Hangzhou, China.

Animal Treatment

   In this research, 20 male Kunming mice, 6 weeks old, were procured from
   Jinan Pengyue Experimental Animal Breeding Co., Ltd., based in Jinan,
   China. Using male mice minimizes hormone-related variability,
   eliminating confounding factors like estrous cycles.[42]9 These mice
   were housed at a steady temperature of 22±2°C and a humidity range of
   60–70%, with a 12-hour light/dark cycle maintained. They had free
   access to standard food and water throughout the study. A week-long
   adaptation period was observed before initiating the main experiment.
   To determine the appropriate sample size, we used the Resource Equation
   Method. The mice were then evenly divided into a control group and a
   test group, each containing 10 mice. For the experimental procedure,
   SN-38 was mixed in a solution of sodium chloride and 5%DMSO to achieve
   the required concentrations. The test group (n=10) received daily
   intraperitoneal injections of SN-38 (20 mg/kg) for a week. The drug
   schedule was established based on existing research.[43]10 The control
   group, consisting of 10 mice, was administered an equivalent volume of
   a blank solvent mixture (DMSO and saline in a 1:9 ratio). This study
   was performed in accordance with the Laboratory animal—Guideline for
   ethical review of animal welfare of China (GB/T 35892–2018) and
   approved by the Ethics Committee of the Affiliated Hospital of Jining
   Medical University (Approval Number: JNMC-2022-DW-041).

Sample Collection and Preparation

   The mice were euthanized on the second day following the final dose.
   Initially, 0.1 mg/kg of Atropine sulfate is administered
   intraperitoneally, followed by anesthesia using a 10% solution of
   chloral hydrate to achieve both hypnotic and analgesic effects. Animals
   are subsequently euthanized using cervical dislocation. Subsequently,
   the cerebral cortex and hippocampus tissues were promptly removed,
   immediately cooled on ice, rinsed with phosphate-buffered saline at pH
   7.2, rapidly frozen in liquid nitrogen, and preserved at −80°C. In the
   tissue processing phase, each 50 mg of tissue was thoroughly
   homogenized in 1 mL of methanol inside a grinding tube. To this
   mixture, 50 μL of heptadecanoic acid (concentration 1 mg/mL in
   methanol) was added, and the mixture was then centrifuged at 4°C for 15
   minutes at 14,000 rpm. Following this, 80 μL of O-methylhydroxylamine
   hydrochloride (15 mg/mL dissolved in pyridine) was introduced to the
   sample. This was followed by an incubation period at 70°C for 90
   minutes, after which 100 μL of N,
   O-bis(trimethylsilyl)trifluoroacetamide (with 1% trimethylchlorosilane)
   was added, and the mixture was re-incubated at 70°C for another 60
   minutes. A 0.22-μm filter was utilized for the final purification step.
   Take 10 μL from each sample in both the control and SN-38 groups to use
   as quality control (QC).

GC-MS Analyses

   The analysis of the samples was conducted using an Agilent 7890B GC
   System coupled with a 7000C Triple Quadrupole Mass Spectrometer,
   originating from the United States. To maintain consistency and
   reliability of the instrument, a quality control sample was run after
   every ten test samples. The samples were injected in a randomized order
   and underwent separation in an HP-5MS fused silica capillary column.
   Helium was employed as the carrier gas, with a split ratio of 50:1. The
   settings for the front inlet purge flow were at 3 mL/min, while the
   overall gas flow rate was maintained at 1 mL/min. The transfer line,
   sample inlet, and ion source temperatures were set at 250°C, 280°C, and
   230°C, respectively. The GC’s temperature protocol commenced at 60°C,
   held for 4 minutes, followed by an increment of 8°C per minute up to
   300°C, where it remained for 5 minutes. The electron beam energy was
   calibrated to −70eV, and the system recorded data at a rate of 20
   spectra per second. For mass spectrometry, electrospray ionization was
   utilized, scanning a full mass-to-charge (m/z) range from 50 to 800.

Data Processing and Metabolite Identification

   The acquired data from Gas Chromatography-Mass Spectrometry (GC-MS)
   underwent analysis using the Agilent Mass Hunter software, Version
   B.07.00, from Agilent Technologies, based in California, USA. This
   process involved converting the raw data into a mass-to-charge ratio
   (m/z) format. The study categorized metabolites as Level 2,
   representing putatively annotated compounds, determined by their
   resemblance to existing commercial spectral libraries.[44]11 A
   comprehensive spectral library, encompassing all quality control
   samples, was compiled. The National Institute of Standards and
   Technology (NIST 14) GC-MS library from the United States played a
   pivotal role in identifying unknown metabolites. This included
   procedures like alignment, time correction, baseline filtering, and
   deconvolution. Metabolites with a structural identification similarity
   exceeding 80% underwent manual verification to ensure precision.
   Subsequently, a bespoke spectral library was constructed for matching
   the experimental samples, culminating in a unified data matrix
   containing peak indices (RT-m/z pairs), sample identifiers, and
   corresponding peak areas.

   Normalization of these peak areas within the data matrix was executed
   using Microsoft Excel™, from Microsoft, Redmond, WA, USA. The
   three-dimensional dataset, consisting of peak indices (RT-m/z pairs),
   sample names, and normalized peak area percentages, was imported into
   SIMCA-P 14.0 (Umetrics, Umeå, Sweden) for statistical analysis.
   Unsupervised Principal Component Analysis (PCA) and Orthogonal
   Projections to Latent Structures Discriminant Analysis (OPLS-DA) were
   used to distinguish between the SN-38-treated group and the control
   group. The validity of the model was confirmed using a permutation test
   with SIMCA-P software. A two-tailed Student’s t-test was conducted
   using SPSS 19.0 (SPSS Inc, Chicago, IL, USA). In the OPLS-DA, variables
   with Importance in Projection (VIP) values greater than 1.0 and
   p-values less than 0.05 were deemed statistically significant and
   selected for identifying differential metabolites.

Statistical Analysis

   We rigorously assessed the data distribution and homogeneity using IBM
   SPSS Statistics 19.We applied the Shapiro–Wilk test for normality,
   selecting compounds with significance (Sig) values greater than 0.05 in
   both sample groups. Levene’s Test assessed the homogeneity of variance,
   selecting compounds with a mean Sig value greater than 0.05. Compounds
   that met both criteria underwent further analysis. We used the
   Independent-Samples T Test for parametric data to identify significant
   compounds based on specific Sig values, and the Mann–Whitney U-Test for
   non-parametric data. These steps ensured that our data met the
   necessary assumptions for valid statistical testing, enhancing the
   robustness and reliability of our findings.

Metabolic Pathway Analysis

   For the pathway analysis in this study, MetaboAnalyst 6.0 (available at
   [45]http://www.metaboanalyst.ca) and the Kyoto Encyclopedia of Genes
   and Genomes (KEGG; accessible at [46]http://www.kegg.jp) were employed.
   Key discriminant metabolites were pinpointed based on having variable
   importance in projection (VIP) values exceeding 1.0 and two-tailed
   Student’s t-test p-values falling below 0.05. These selected
   metabolites underwent further examination through pathway analysis
   tools, aiming to pinpoint differential metabolites and their
   significant metabolic pathways. Pathways that exhibited a raw p-value
   less than 0.05 were deemed significant in the context of SN-38
   toxicity.

Results

GC-MS TICs

   [47]Figure 1 displays the total ion chromatograms (TICs) of quality
   control samples derived from both cortical and hippocampal tissues.
   These chromatograms reveal the successful detection of various
   metabolites throughout the run time, characterized by substantial peak
   capacities, consistent retention times, and outstanding
   reproducibility.

Figure 1.

   [48]Figure 1
   [49]Open in a new tab

   Representative gas chromatography–mass spectrometry (GC–MS) total ion
   chromatograms (TICs) from quality control (QC) samples. (A) the
   hippopotamus and (B) the the cerebral cortex.

Multivariate Analysis of Metabolomic Data

   [50]Table 1 provides a detailed description of the OPLS-DA model
   parameters, each approaching a value close to 1.0. The model’s
   robustness is evidenced by a ranking test, which confirms its
   reliability; all intersection points of the blue regression line
   (Q2-points) with the left vertical axis are negative ([51]Figure 2).
   Furthermore, the OPLS-DA analysis distinctly demonstrates a significant
   difference between the SN-38 and control groups.

Table 1.

   The Scores of the Model Parameters
                   R^2X(cum) R^2Y(cum) Q^2(cum)
   Hippocampus     0.392     0.988     0.648
   Cerebral cortex 0.407     0.994     0.828
   [52]Open in a new tab

Figure 2.

   [53]Figure 2
   [54]Open in a new tab

   OPLS-DA score plots and permutation tests. (A)Hippocampus (B) Cerebral
   cortex.

Identification of Metabolic Changes

   Metabolite identification was based on a well-established mass
   spectrometry database, initially using the Agilent Fiehn GC/MS
   Metabolomics RTL Library for profile comparison.Standard criteria for
   metabolite identification include VIP scores and p-values.Metabolites
   are marked as significantly altered between groups if they have a VIP
   score over 1 in OPLS-DA analysis and a p-value under 0.05 in the
   Student’s t-test.[55]Table 2 outlines altered metabolites in tissues
   across groups.Post-SN-38 challenge, the hippocampus showed 2 increased
   and 16 decreased metabolites compared to the control group.SN-38
   exposure in the cerebral cortex resulted in 24 altered metabolites: 3
   up-regulated and 21 down-regulated.Heatmaps are used for quality
   control and to present differential data.Clustering groups similar
   items into categories or subsets through static classification,
   highlighting similarities within each subset.[56]Figure 3 illustrates
   significant group differences.

Table 2.

   Differential Metabolites in Different Tissues After SN-38 Treatment
   Samples Metabolites HMDB ID VIP p-value Fold change
   Hippocampus N-Dodecane HMDB0031444 0.935179 0.022 0.779712
   L-Alanine HMDB0000161 1.2432 0.001 0.780999
   Beta-Alanine HMDB0000056 1.24455 0.004 0.763298
   Oxalic acid HMDB0002329 1.09233 0.019 0.781427
   O-Phosphoethanolamine HMDB0000224 1.85869 0 0.590684
   L-Tyrosine HMDB0000158 1.42048 0 0.741043
   scyllo-Inositol HMDB0006088 1.49664 0 0.726179
   Ribothymidine HMDB0000884 1.05567 0.016 1.23005
   Arachidic acid HMDB0002212 1.20825 0.019 0.747809
   Uridine HMDB0000296 1.1302 0.007 1.22134
   Inosine HMDB0000195 1.06666 0.003 0.813259
   Guanosine HMDB0000133 2.16882 0.001 0.483594
   Cholesterol HMDB0000067 1.1387 0.026 0.762898
   L-Lactic acid HMDB0000190 2.00912 0 0.591114
   Urea HMDB0000294 1.09792 0.031 0.788596
   Succinic acid HMDB0000254 2.97199 0.019 0.142377
   Oleic acid HMDB0000207 1.10526 0.024 0.722348
   2,5-Furandicarboxylic acid HMDB0004812 1.522 0 0.715994
   Cerebral cortex D-Lactic acid HMDB0001311 1.45558 0 0.646974
   L-Alanine HMDB0000161 1.25143 0.002 0.720946
   L-Valine HMDB0000883 1.61827 0 0.628527
   Glycerol HMDB0000131 1.04313 0 0.812167
   Gamma-Aminobutyric acid HMDB0000112 1.63792 0 0.637064
   Glyceric acid HMDB0000139 1.16445 0.047 0.668106
   Uracil HMDB0000300 1.44232 0 0.678135
   Serine HMDB0062263 1.42756 0 0.699546
   L-Threonine HMDB0000167 1.25731 0.001 0.724564
   L-Aspartic acid HMDB0000191 1.19529 0 0.757575
   Beta-Alanine HMDB0000056 2.28132 0 0.411832
   L-Glutamic acid HMDB0000148 1.2445 0.002 1.38101
   Taurine HMDB0000251 1.02676 0.032 0.712355
   O-Phosphoethanolamine HMDB0000224 2.67578 0 0.283213
   L-Lysine HMDB0000182 1.64567 0 0.617583
   scyllo-Inositol HMDB0006088 1.38441 0.045 0.593565
   Inosine HMDB0000195 1.24768 0 1.32621
   Glycine HMDB0000123 1.10208 0.01 0.741338
   Ethanolamine HMDB0000149 1.99027 0 0.533868
   L-Isoleucine HMDB0000172 1.60849 0 0.631164
   Hypoxanthine HMDB0000157 1.67037 0 0.640728
   Citric acid HMDB0000094 1.88007 0 0.521356
   L-Tyrosine HMDB0000158 1.38276 0.01 0.605454
   Adenosine HMDB0000050 2.80038 0 8.10382
   [57]Open in a new tab

Figure 3.

   [58]Figure 3
   [59]Open in a new tab

   Heatmap of differential metabolites in the SN-38 group compared with
   controls, featuring: (A) Hippocampus and (B) Cerebral Cortex.Each color
   in the heatmap represents metabolite changes: blue for down-regulation,
   and red for up-regulation.In the heatmap, rows represent samples and
   columns represent metabolites.

Analyses of Metabolic Pathways

   MetaboAnalyst 6.0 was used to analyze identified metabolites and assess
   metabolic alterations between the SN-38 and control groups. Metabolic
   pathways with a Raw P-value < 0.05 (see [60]Table 3 and [61]Figure 4)
   were deemed potentially disturbed. [62]Figure 5 presents a detailed
   metabolic network diagram. The Results indicated that SN-38 exposure
   affected multiple metabolic pathways, including those for various amino
   acids, propanoate, purine, pyrimidine, butanoate, glyoxylate, and
   dicarboxylate, as well as glycerolipid metabolism. Additionally,
   pathways involved in pantothenate and CoA biosynthesis and the
   biosynthesis of unsaturated fatty acids were also impacted.

Table 3.

   Pathway Analysis Performed Using MetaboAnalyst 6.0 Software
   Samples Pathway Names Raw p Impact
   Hippocampus Propanoate metabolism 0.014045 0.0
   Purine metabolism 0.019694 0.00274
   Alanine, aspartate and glutamate metabolism 0.022329 0.0
   Phenylalanine, tyrosine and tryptophan biosynthesis 0.033654 0.5
   Biosynthesis of unsaturated fatty acids 0.035813 0.0
   Pyrimidine metabolism 0.041522 0.03985
   Cerebral cortex Alanine, aspartate and glutamate metabolism 3.83E-05
   0.50722
   Glyoxylate and dicarboxylate metabolism 7.54E-05 0.25927
   Valine, leucine and isoleucine biosynthesis 0.00015938 0
   Pantothenate and CoA biosynthesis 0.00016191 0.04762
   Glycine, serine and threonine metabolism 0.0013463 0.53603
   beta-Alanine metabolism 0.0033292 0.39925
   Arginine biosynthesis 0.017698 0.11675
   Butanoate metabolism 0.020234 0.03175
   Histidine metabolism 0.022915 0
   Glycerolipid metabolism 0.022915 0.33022
   [63]Open in a new tab

   Abbreviations:SN-38, Ethyl-10-hydroxycamptothecin;TICs, total ion
   chromatograms; GC-MS, Gas Chromatography-Mass Spectrometry;CNS, central
   nervous system;ATP, Adenosine 5’-triphosphate;GABA, γ-aminobutyric
   acid;BCAAs, Branched-chain amino acids;TAG, triacylglycerol;PE,
   phosphatidylethanolamine.

Figure 4.

   [64]Figure 4
   [65]Open in a new tab

   Summary of the pathway analysis conducted with MetaboAnalyst 6.0. The
   metabolome view shows all matched pathways according to the p values
   from the pathway enrichment analysis and pathway impact values from the
   pathway topology analysis. (A)Hippocampus: a Propanoate metabolism; b
   Purine metabolism; c Alanine, aspartate and glutamate metabolism; d
   Phenylalanine, tyrosine and tryptophan biosynthesis; e Biosynthesis of
   unsaturated fatty acids; f Pyrimidine metabolism; (B) Cerebral cortex:
   c Alanine, aspartate and glutamate metabolism; g Glyoxylate and
   dicarboxylate metabolism; h Valine, leucine and isoleucine
   biosynthesis; i Pantothenate and CoA biosynthesis; j Glycine, serine
   and threonine metabolism; k beta-Alanine metabolism; l Arginine
   biosynthesis; m Butanoate metabolism; n Histidine metabolism; o
   Glycerolipid metabolism.

Figure 5.

   [66]Figure 5
   [67]Open in a new tab

   Schematic diagram showing metabolic pathways in tissues affected by
   SN-38 processing.Metabolites marked in red indicate potential
   biomarkers identified in this study.

Discussion

   Clinical evidence indicates that the toxic effects of Irinotecan and
   its metabolite SN-38 are primarily manifested in the gastrointestinal
   tract and liver, with little reported toxicity to other organs. For the
   first time, we evaluated the nervous system in mice, employing
   multivariate statistical analysis to study the neurotoxicity related to
   SN-38 and potential biomarkers. This could aid researchers in
   understanding the pathogenesis of SN-38-induced neurotoxicity. In our
   study, 36 potential biomarkers were identified, 18 in the hippocampus
   and 24 in the cortex; these will be discussed separately below.

   Adenosine not only participates in ATP formation but also serves as a
   signaling molecule in purinergic pathways. In the central nervous
   system(CNS), purinergic receptors are prevalent in neurons, glial
   cells, and vascular endothelial cells.[68]12,[69]13 This signaling is
   instrumental in maintaining the functionality and homeostasis of
   neuronal, astrocytic, and microglial cells, thereby facilitating
   essential processes such as synaptic transmission and higher cognitive
   functions[70]14–16 Recent studies highlight purinergic signaling’s role
   in modulating blood-brain barrier permeability.[71]17,[72]18 Pathway
   analysis reveals disrupted purine metabolism in the hippocampus of the
   SN-38 group, accompanied by an increased adenosine level in the cortex.
   This indicates blood-brain barrier dysfunction and energy metabolic
   disorders. Purine metabolism dysregulation is a known factor in the
   pathogenesis of neurodegenerative diseases, including Alzheimer’s and
   Parkinson’s disease,[73]19,[74]20 pointing to potential signaling
   pathway and cognitive function disruptions due to SN-38 treatment.

   Guanosine, a purine nucleoside, exhibits neuroprotective effects in
   conditions like ischemic stroke, Alzheimer’s disease, and Parkinson’s
   disease.[75]21–23 Its neuroprotection involves multiple mechanisms:
   reducing glutamatergic excitotoxicity in astrocytes,[76]24 modulating
   the adenosinergic system,[77]25 and affecting inflammatory cascades and
   oxidative stress.[78]26 This study’s finding of reduced guanosine in
   the hippocampus implies a compromised neuroprotection in the SN-38
   group.

   Amino acids and their metabolites regulate neuronal activity in the
   CNS, assisting in diagnosing various diseases.[79]27 In comparison to
   the control group, the SN-38 group exhibited a downregulation of
   L-alanine, L-tyrosine, valine, serine, L-threonine, L-aspartate,
   L-lysine, glycine, and L-isoleucine, as well as an upregulation of
   L-glutamate. The affected metabolic pathways include alanine,
   aspartate, and glutamate metabolism; arginine biosynthesis; and
   biosynthesis of phenylalanine, tyrosine, and tryptophan; as well as
   valine, leucine, and isoleucine; glycine, serine, and threonine; and
   histidine metabolism. Glutamate, an excitatory neurotransmitter, is
   crucial in cellular metabolism and implicated in neuropsychiatric
   disorders such as epilepsy, schizophrenia, and depression. Stress and
   inflammation may increase glutamate, further harming neurons. Brain
   endothelial cells exposed to high levels of glutamate activate
   N-methyl-D-aspartate receptors, disrupting the blood-brain
   barrier.[80]28 Aspartate, a non-essential amino acid, exhibits abnormal
   levels in various neurological disorders.[81]29 It affects glycolysis
   via the malate-aspartate shuttle, disrupting brain energy
   metabolism.[82]30 Aspartate also regulates nitrogen and carbon dioxide
   levels in the blood and offers neuroprotection.[83]31 Glutamate’s
   metabolism into γ-aminobutyric acid(GABA) through butyrate metabolism
   makes it the CNS’s primary inhibitory neurotransmitter. Reduced GABA
   levels are linked to increased neuronal excitability and cognitive
   impairments. Studies show reduced GABA in Alzheimer’s patients.[84]32

   Branched-chain amino acids(BCAAs), key metabolic precursors, play a
   role in neurotransmitter synthesis and amino acid metabolism in the
   brain.[85]33 Studies show reduced valine and isoleucine levels in the
   cortical tissue of SN-38-treated mice compared to controls, indicating
   an effect on BCAA biosynthesis. This suggests that SN-38 could
   negatively impact brain function by disrupting the synthesis of vital
   proteins, neurotransmitters, and metabolites. Tyrosine, a catecholamine
   precursor, influences neural transmission through its varying
   levels.[86]34 Histidine, a carnosine precursor, contributes to
   carnosine synthesis, utilizing ATP hydrolysis and β-alanine. Carnosine
   functions significantly as a buffer and antioxidant.[87]35 Animal
   research reveals β-alanine supplementation enhances brain carnosine
   levels, offering neuroprotection.[88]36 While human studies have not
   shown a significant increase in brain carnosine levels, evidence
   suggests β-alanine may enhance cognitive function and mood,[89]37
   although the exact mechanisms are still unknown. Research links
   phenylalanine, tyrosine, and tryptophan biosynthesis to Alzheimer’s
   disease.[90]38 Overall, amino acid metabolic alterations in the cortex
   and hippocampus of SN-38-treated mice could illuminate SN-38-induced
   neurotoxicity pathogenesis.

   Metabolomic and lipidomic analyses revealed abnormal glycerolipid
   metabolism in the brains of SN-38-treated mice. Glycerol, a crucial
   energy reserve, stores excess glucose as triglycerides in the body.
   Research has confirmed that diminished levels of triacylglycerol (TAG)
   correlate strongly with an increased severity of cognitive impairments,
   underscoring the critical role of lipid metabolism in brain function.
   Glycerolipids, significant as signaling molecules in the central
   nervous system, play key roles in appetite control, immune system
   regulation, and pain management.[91]39

   The observed abnormalities in triglyceride metabolism suggest a broader
   disruption that extends to glycerophospholipid and myo-inositol
   phosphate metabolism, potentially impacting these critical signaling
   pathways.The reduction in phosphatidylethanolamine (PE), a major
   component and precursor in brain phospholipid synthesis, usually
   indicates compromised myelination and could thus exacerbate cognitive
   and behavioral deficits.[92]40 The link between PE and epilepsy is
   well-documented.[93]41 Myo-inositol, prevalent in central glial cells,
   reflects glial cell activity. Studies show elevated brain myo-inositol
   levels in Alzheimer’s, amnestic mild cognitive impairment, and chronic
   alcoholism patients. Conversely, lower brain myo-inositol levels in
   fibromyalgia patients correlate with memory and attention
   deficits.[94]42 This study’s observed changes in PE and myo-inositol
   levels suggest links to brain dysfunction-related diseases, warranting
   further investigation.

   The metabolomic study results elucidate the effects of SN-38 on brain
   metabolism. However, the study has limitations; further experiments in
   animal tumor models are needed to understand the impact of SN-38 on
   brain metabolism. Based on the established metabolomics platform, the
   drug-related toxicity of SN-38 in combination with other chemotherapy
   drugs still requires experimental evaluation. This evaluation can help
   advance the clinical application of various SN-38 delivery systems. In
   our research, atropine sulfate combined with chloral hydrate was chosen
   to anesthetize rats due to its rapid induction of sleep and
   anticonvulsant properties. Despite its historical use in small animal
   anesthesia, chloral hydrate has limitations. Increasing evidence
   suggests that it does not provide sufficient analgesia and can cause
   significant respiratory depression at doses required for surgical
   anesthesia, posing potential risks to animal health and experimental
   safety.[95]43 Considering these limitations, we plan to replace chloral
   hydrate with safer anesthetics such as pentobarbital or isoflurane in
   future studies.

Conclusion

   This study employed GC-mass spectrometry to investigate the metabolomic
   profile of the nervous system in mice exposed to SN-38. Through initial
   screening of metabolites, 24 and 18 characteristic metabolites were
   identified in the cortex and hippocampus, respectively. It was found
   that SN-38 affects purine metabolism, pyrimidine metabolism, amino acid
   metabolism, and glycerolipid metabolism in the central nervous system,
   leading to blood-brain barrier disruption, abnormal energy metabolism,
   and central signaling pathway disorders. This metabolomics data
   provides new evidence for studying the neurotoxicity induced by SN-38.

Funding Statement

   This work was supported by the National Natural Science Foundation of
   China (Grant No.82302795); Natural Science Foundation of Shandong
   Province [ZR2023QH094]; Jining Key Research and Development Program
   [2022YXNS019];Nursery Research Project of Affiliated Hospital of Jining
   Medical University [MP-MS-2022-006].

Disclosure

   The authors report no conflicts of interest in this work.

References