Metabolomic Analysis of Exosomes Derived from Lung Cancer Cell Line H460 Treated with SH003 and Docetaxel

Exosomes released from tumor cells treated with cancer-targeting drugs reflect altered metabolic processes within the cells. Therefore, metabolites in exosomes can be used as markers to predict the therapeutic response or identify therapeutic targets. In this study, metabolite changes in exosomes were investigated by co-administration of the herbal extract SH003 and docetaxel (DTX), which exert a synergistic anti-cancer effect on lung cancer cells. Exosomes released from cells treated with SH003 and DTX were purified, and untargeted metabolic profiling was performed by liquid chromatography–tandem mass spectrometry. Analysis of altered metabolic-based pathways showed that the combined treatment synergistically increased pyrimidine metabolism compared with single-drug treatment. Additionally, xenobiotic metabolism by cytochrome P450 was specifically increased in cells treated with the combination. However, the released exosomes and increased metabolites in exosomes did not affect the anti-cancer effect of SH003 and DTX. Therefore, our study suggests that metabolite profiling can be used to evaluate the efficacy of combined treatments. Furthermore, such exosome-based metabolism may facilitate understanding the physiological endpoints of combination therapy in human biofluids.


Introduction
Exosomes are small membrane vesicles (30-150 nm) released by several cell types and are stable sources of cell-derived genetic materials that modulate multiple signaling pathways in recipient cells [1]. Cancer-derived exosomes are regarded as major mediators of tumor progression, metastasis, multidrug resistance, and immune modulation [2]. Moreover, cancer exosomes contain potential cancer-related RNAs, DNAs, and proteins and are stable in body fluids, including blood, saliva, and urine, suggesting that exosomes can be used as diagnostic and prognostic biomarkers of cancer [3]. Therefore, exosomes have been proposed to be potential therapeutic, diagnostic, and prognostic markers of several cancers [4].
Liquid chromatography-mass spectrometry (LC-MS)-based metabolomics, which is the large-scale profiling of metabolites in biofluids, cells, and tissues, is an unbiased tool for biomarker discovery. Currently, metabolomes of exosomes have gained interest in cancer research [5][6][7]. Numerous studies have reported the results of analyzing metabolomic alterations in cancer cells in response to anti-cancer drugs, which is helpful in understanding the mechanisms of anti-cancer drug actions and finding potential biomarkers for cancer diagnosis [8][9][10][11][12][13]. Interestingly, metabolites in cancer cell-derived exosomes are involved in cancer progression and multidrug resistance, suggesting that unbiased metabolomic study of cancer cell-derived exosomes is essential to identify novel cancer biomarkers for prognosis, prediction, and therapeutic responses.
washing with PBS, exosomes were dehydrated in 30%, 50%, 70%, 80%, and 95% (vol/vol) ethanol solutions for 10 min and then incubated in 100% (vol/vol) ethanol for 15 min twice. For sample drying, exosomes were transferred to a 1:2 of hexamethyldisilazane (HMDS):100% ethanol for 20 min and then a 2:1 solution of HMDS:100% ethanol for 20 min. Samples diluted with 100% HMDS were coated on glass coverslips precoated in a 10% poly-L-lysine solution in PBS for 30 min at RT overnight. For non-conductive sample analysis, a thin layer of Pt was placed in the sample-processing chambers. Images were obtained using Field Emission S-4700 scanning electron microscope (Hitachi, Japan) and analyzed under the following conditions: 10 kV accelerating voltage, 10.5 µA emission current, and 9.4 mm working distance.

Western Blotting
Exosome pellets were suspended with 1× PBS and lysed with RIPA buffer containing protease and phosphatase inhibitors. Proteins were diluted with 5× sample buffer, separated by SDS-PAGE, and transferred to a nitrocellulose membrane. The membranes were blocked with PBS containing 0.1% Tween-20 and 5% skim milk and then incubated with primary antibodies at 4 • C overnight. The blots were incubated with horseradish peroxidase-conjugated secondary antibodies and detected using an EZ-western detection kit (Dogen-Bio, Seoul, Korea). Anti-Alix, -CD9, and -GAPDH antibodies were purchased from Cell Signaling (Danvers, MA, USA).

Quantification of Exosomal Metabolites and Statistical Analysis
Data were extracted using apLCMS and characterized by xMSanalyzer software by filtering the sample at a coefficient of variation (CV) < 50%. The intensity of a metabolite was averaged, log2 transformed, normalized, and indicated by autoscaling. Hierarchical cluster analysis and principal component analysis were performed to identify differences between groups. Significant metabolites were detected by a Manhattan plot, which is indicated by −logP vs. m/z (mass-to-charge ratio), using the p-value. The pathway was identified by matching in the Kyoto Encyclopedia of Genes and Genomes (KEGG) database using MetaboAnalyst 5.0.
Statistical analysis was performed using PRISM 8.0.2 (GraphPad, San Diego, CA, USA). Differences in means between groups were analyzed by one-way ANOVA using Tukey's multiple comparisons test. p < 0.05 indicated a statistically significant difference. Results are represented as the mean ± standard deviation.

Characterization of Exosomes Released from H460 Cells Treated with Both SH003 and DTX
To isolate exosomes secreted after combination therapy by SH003 and DTX, the culture supernatant from H460 lung cancer cells was collected, purified, and analyzed for altered metabolites (Figure 1).
Statistical analysis was performed using PRISM 8.0.2 (GraphPad, San Diego, CA, USA). Differences in means between groups were analyzed by one-way ANOVA using Tukey's multiple comparisons test. p < 0.05 indicated a statistically significant difference. Results are represented as the mean ± standard deviation.

Characterization of Exosomes Released from H460 Cells Treated with Both SH003 and DTX
To isolate exosomes secreted after combination therapy by SH003 and DTX, the culture supernatant from H460 lung cancer cells was collected, purified, and analyzed for altered metabolites (Figure 1). The expression of exosomal markers Alix and CD9 [28] was enriched in the exosome fraction compared with that in the cell lysate ( Figure 2A). SEM was used to examine the morphology and diameter of exosomes, which indicated that exosomes were round vesicles and uniform with a size of approximately 100 nm ( Figure 2B). The exosome size and concentration distribution were assessed by NTA. The majority of particles had a diameter range of 145-165 nm ( Figure 2C and Table 1) and the particle concentration showed no difference between drug treatment groups ( Figure 2D). Thus, we established an exosome isolation method and found that the combinatorial treatment was not related to the number of exosomes secreted from cells. The expression of exosomal markers Alix and CD9 [28] was enriched in the exosome fraction compared with that in the cell lysate ( Figure 2A). SEM was used to examine the morphology and diameter of exosomes, which indicated that exosomes were round vesicles and uniform with a size of approximately 100 nm ( Figure 2B). The exosome size and concentration distribution were assessed by NTA. The majority of particles had a diameter range of 145-165 nm ( Figure 2C and Table 1) and the particle concentration showed no difference between drug treatment groups ( Figure 2D). Thus, we established an exosome isolation method and found that the combinatorial treatment was not related to the number of exosomes secreted from cells.
Statistical analysis was performed using PRISM 8.0.2 (GraphPad, San Diego, CA, USA). Differences in means between groups were analyzed by one-way ANOVA using Tukey's multiple comparisons test. p < 0.05 indicated a statistically significant difference. Results are represented as the mean ± standard deviation.

Characterization of Exosomes Released from H460 Cells Treated with Both SH003 and DTX
To isolate exosomes secreted after combination therapy by SH003 and DTX, the culture supernatant from H460 lung cancer cells was collected, purified, and analyzed for altered metabolites (Figure 1). The expression of exosomal markers Alix and CD9 [28] was enriched in the exosome fraction compared with that in the cell lysate ( Figure 2A). SEM was used to examine the morphology and diameter of exosomes, which indicated that exosomes were round vesicles and uniform with a size of approximately 100 nm ( Figure 2B). The exosome size and concentration distribution were assessed by NTA. The majority of particles had a diameter range of 145-165 nm ( Figure 2C and Table 1) and the particle concentration showed no difference between drug treatment groups ( Figure 2D). Thus, we established an exosome isolation method and found that the combinatorial treatment was not related to the number of exosomes secreted from cells.

Metabolome Analysis of Exosomes after Combinatorial Treatment
We identified what metabolites were influenced after combined treatment compared with SH003 or DTX alone by LC-MS/MS. The abundance levels of metabolites were transformed to the log2 value and normalized. To investigate the metabolite difference between groups, we first performed hierarchical cluster analysis (HCA) and principal component analysis (PCA) on the four groups. Each drug treatment showed a distinct clustering of metabolites ( Figure 3A,B). To classify the characteristics of these metabolites, pathways corresponding to significant metabolites (p < 0.05) were mapped from a KEGG database, and pathway enrichment and topology were analyzed using MetaboAnalyst 5.0. As a result, we identified that six pathways, including retinol metabolism, pyrimidine metabolism, lysine degradation, arginine and proline metabolism, metabolism of xenobiotics by cytochrome P450, and propanoate metabolism, have a significant impact on combination treatment ( Table 2). The results also revealed other metabolic features in response to SH003, DTX, and combined treatments ( Figure 3C). Furthermore, a Venn diagram illustrated the distribution of metabolic pathways enriched in the treatment groups ( Figure 4). We found pathways that overlapped among all treatment groups as follows: lysine degradation; porphyrin and chlorophyll metabolism; pyrimidine metabolism; steroid hormone biosynthesis; ubiquinone and other terpenoid-quinone biosynthesis; purine metabolism; glutathione metabolism; and primary bile and biosynthesis. However, three pathways, including retinol metabolism, metabolism of xenobiotics by cytochrome P450, and propanoate metabolism, were only mapped in combinatorially treated exosomes. Metabolites 2022, 12, x FOR PEER REVIEW 6 of 13

Identification of Metabolites Altered by Combinatorial Treatment
Next, we investigated functional metabolites modulated by combined treatment. O the metabolites shown in Table 2, we identified two metabolites, uridine and NNAL which were significantly increased by the combined treatment (p < 0.05). Uridine is in volved in pyrimidine metabolism, which was significantly altered in all groups ( Figure 4 and plays an important role in the synthesis of glucose, lipids, and amino acids [29]. Th level of uridine was increased by combinatorial treatment at a 7.4 ratio compared wit that in the control, SH003, and DTX groups ( Figure 5A). As shown in Figure 5B, the vo cano plot also showed upregulation of uridine by drug exposure in all treated sample compared with that in the control. Furthermore, the NNAL metabolite belonging to th metabolism of xenobiotics by cytochrome P450 was enriched in only the combined grou (Figure 4), showing a higher level of approximately 42-fold compared with that in th control ( Figure 5C). Therefore, we determined whether exosomal metabolites secreted a ter combinatorial treatment were related to cell-cell communication for the death of lun cancer cells. When only exosomes secreted from drug-treated cells were treated in cells, did not affect the cell viability ( Figure 5D). These results suggested that the combinatio of SH003 and DTX led to cell death regardless of the exosome-mediated signal transduc tion. Additionally, neither uridine nor NNAL altered the effect of the combination on ce viability, showing no significant difference compared with the combined treatment (Fig  ure 5E,F). Thus, we concluded that metabolites enriched in exosomes may be biomarker to predict the therapeutic response for cancer suppression rather than improved drug sen sitivity of combination treatment.

Identification of Metabolites Altered by Combinatorial Treatment
Next, we investigated functional metabolites modulated by combined treatment. Of the metabolites shown in Table 2, we identified two metabolites, uridine and NNAL, which were significantly increased by the combined treatment (p < 0.05). Uridine is involved in pyrimidine metabolism, which was significantly altered in all groups (Figure 4), and plays an important role in the synthesis of glucose, lipids, and amino acids [29]. The level of uridine was increased by combinatorial treatment at a 7.4 ratio compared with that in the control, SH003, and DTX groups ( Figure 5A). As shown in Figure 5B, the volcano plot also showed upregulation of uridine by drug exposure in all treated samples compared with that in the control. Furthermore, the NNAL metabolite belonging to the metabolism of xenobiotics by cytochrome P450 was enriched in only the combined group (Figure 4), showing a higher level of approximately 42-fold compared with that in the control ( Figure 5C). Therefore, we determined whether exosomal metabolites secreted after combinatorial treatment were related to cell-cell communication for the death of lung cancer cells. When only exosomes secreted from drug-treated cells were treated in cells, it did not affect the cell viability ( Figure 5D). These results suggested that the combination of SH003 and DTX led to cell death regardless of the exosome-mediated signal transduction. Additionally, neither uridine nor NNAL altered the effect of the combination on cell viability, showing no significant difference compared with the combined treatment ( Figure 5E,F). Thus, we concluded that metabolites enriched in exosomes may be biomarkers to predict the therapeutic response for cancer suppression rather than improved drug sensitivity of combination treatment.   (D) Cells were treated with 20 and 40 μg/mL exosomes or SH003 and/or DTX for 24 h. Cell viability was analyzed by WST assays. E1: exosome 20 μg/mL; E2: exosome 40 μg/mL. (E,F) Cells were pre-treated with both SH003 and DTX for 1 h and then incubated with uridine or NNAL for 24 h. * p < 0.05, ** p < 0.01, and **** p < 0.0001 relative to the control by one-way ANOVA with Tukey's post-hoc test.

Discussion
Despite remarkable progress in diagnosis and therapeutic options for NSCLC patients, new cases and deaths of NSCLC have increased for both men and women from 1998 to 2021. Recently, exosomes have been used as a diagnostic marker for cancer and as a predictive marker for therapy [3]. Exosomes released from tumor cells reflect the hallmarks of cancer, and genetic materials and other substances within exosomes provide important information about the intracellular and intercellular signaling that occurs in cancer. Therefore, profiling exosome-carried molecules will help to better understand cancer treatments. Previous studies have shown that SH003 has anti-cancer properties and the potential for combination with DTX in NSCLC treatment. In this study, we identified markers that predicted the effect of combined treatment by analyzing the metabolic changes in exosomes induced by SH003, DTX, or both in vitro. Our results showed that metabolites and metabolic pathways as biomarkers indicated the treatment outcome of SH003-DTX combination-treated H460 cells.

Discussion
Despite remarkable progress in diagnosis and therapeutic options for NSCLC patients, new cases and deaths of NSCLC have increased for both men and women from 1998 to 2021. Recently, exosomes have been used as a diagnostic marker for cancer and as a predictive marker for therapy [3]. Exosomes released from tumor cells reflect the hallmarks of cancer, and genetic materials and other substances within exosomes provide important information about the intracellular and intercellular signaling that occurs in cancer. Therefore, profiling exosome-carried molecules will help to better understand cancer treatments. Previous studies have shown that SH003 has anti-cancer properties and the potential for combination with DTX in NSCLC treatment. In this study, we identified markers that predicted the effect of combined treatment by analyzing the metabolic changes in exosomes induced by SH003, DTX, or both in vitro. Our results showed that metabolites and metabolic pathways as biomarkers indicated the treatment outcome of SH003-DTX combination-treated H460 cells.
Abnormal metabolic regulation for faster energy supply is considered as a hallmark of cancer. Therefore, to deal with complex cancer metabolism, metabolomics has been continuously increasing in cancer research [30]. Miyamoto et al. demonstrated the potential role of metabolites as cancer diagnostic markers by identifying altered metabolites in lung cancer patients by GC-TOF MS analysis [31]. In this study, the metabolites changed by SH003 and DTX were identified by LC-MS/MS. The screening results showed that "pyrimidine metabolism in cancer", especially uridine, was a metabolite changed significantly by the combined treatment of H460 cells with SH003 and DTX. Uridine is involved in the synthesis of the building block, including proteins, nucleic acids, and lipids, for tumor cell proliferation via the pyrimidine salvage pathway. In terms of these activities associated with tumorigenesis, decreasing the intracellular uridine pool can be helpful for cancer proliferation-targeted therapy and increase the efficacy of anti-cancer drugs [32]. Uridine also inhibits side effects such as neurological deficits and myelotoxicity caused by pyrimidine metabolism-targeting drugs [33,34]. We assumed that the increased uridine in exosomes was associated with the synergistic effect, but both released exosome and uridine did not affect the inhibition of lung cancer growth. These results suggest that uridine is a putative biomarker to evaluate combination efficacy. In addition, based on the known protective effect of uridine on normal tissues, it is necessary to prove the possibility that the increased uridine of exosomes is involved in the effective cancer suppression mechanism of combination therapy through alleviation of toxicity to normal tissues.
In the metabolomics data, a metabolic change in xenobiotics by cytochrome P450 (CYP) was selectively induced only by the combined treatment. Among related metabolites, NNAL was significantly enhanced by the combined treatment ( Figure 5C). NNAL is mostly known as an intermediate metabolite of nicotine, a xenobiotic, and metabolized by the CYP2A6 enzyme [35,36]. Because xenobiotics, which are recognized as foreign substances in cells, can potentially cause toxicity, they are absorbed, detoxified, and excreted by CYP enzymes to maintain a stable state within the cell [37]. However, CYP enzymes induce cell damage by inducing the biotransformation of toxic substances depending on their subtype. Therefore, the pharmacokinetics and pharmacodynamics of anti-cancer drugs are determined by the function of the CYP enzyme. Many studies have shown that NNAL is a metabolite detected as a carcinogen in lung cancer patients who smoke. However, the relationship of NNAL with the mechanism of anti-cancer drugs has not been elucidated. Our data showed that NNAL did not potentiate the cytotoxicity of the combined treatment ( Figure 5F). Therefore, similar to uridine, we concluded that NNAL may be a response biomarker for combined treatment by SH003 and DTX. However, further studies are required to show how CYP enzymes and xenobiotic mechanisms are related to an effective response to SH003 and DTX.
Although the efficacy and molecular mechanism of SH003 have been well evaluated in non-clinical and clinical studies, further studies are required to find therapeutic biomarkers. Because changes in metabolites reflect the indispensable quality of drug dynamics in cells, this study revealed metabolites in exosomes as biomarkers to evaluate the effectiveness of combined treatment. Combined treatment with SH003 and DTX did not affect exosome production, but different levels of uridine and NNAL metabolites were found in released exosomes, showing potential as biomarkers for drug-efficacy evaluation. However, it is necessary to identify the metabolic pathways regulated by SH003 and DTX in cells to use these metabolites as biomarkers and to check the level of metabolites in blood, which do not interfere with combination therapy.

Conclusions
This study suggests exosomes as an indicator to evaluate the effectiveness of SH003 and DTX combination therapy rather than the function of exosomes as an inducer of cancer cell inhibition involved in cell-cell interactions. Therefore, metabolite analysis of exosomes can explore new biomarkers to evaluate concomitant efficacy. Additionally, analysis of altered metabolites will aid in the understanding of novel molecular mechanisms that inhibit cancer cells induced by SH003 with multidrug action properties.