Synthesis and use of ginsenoside compound k derivatives

Abstract

Nontoxic anticancer compounds are derivatives of 20-O-β-(D-glucopyranosyl)-20(S)-protopanoxadiol (CK) having at least one glucose hydroxyl group substituted with at least one acetal group. The derivatives enhance binding to mitochondrial membrane proteins and are more cytotoxic to cancer cells than CK. The derivatives can be acetals of unsubstituted or substituted aromatic groups, such as acetals formed from substituted 1-(dimethoxymethyl)-benzene. One or more CK derivative (CKD) anticancer compounds can be used as active moieties of anticancer drugs.

Description

Technical Field

[0001] This invention belongs to the field of medicine. Specifically, this invention relates to the synthesis and use of a ginsenoside compound K derivative. Background Technology

[0002] Lung cancer is one of the most common types of cancer and a leading cause of cancer-related deaths worldwide. Non-small cell lung cancer (NSCLC) accounts for approximately 85% of diagnosed lung cancers and has poor treatment efficacy. Ginseng has been used for thousands of years in Asian countries as a traditional medicine to replenish vital energy, with no side effects. The main components of ginseng are ginsenosides, which have a steroidal nucleus containing four trans-rings with modified side chains at C-20 and sugar chains at different positions. Compound K(CK)(20-O-β-(D-glucopyranosyl)-20(S)-protopanaxadiol) is an intestinal bacterial metabolite of ginsenosides Rb1, Rb2, and Rc. Over the decades, numerous studies have described the anticancer activity of CK in various cancer cell lines, including lung cancer, liver cancer, breast cancer, and colorectal cancer. CK inhibits the proliferation of lung cancer cells, including NCI-H460, A549, and H1299, by suppressing the expression of the downstream gene GLUT1 and glucose metabolism regulated by hypoxia-inducible factor-1α (HIF-1α). CK induces apoptosis and autophagy in A549 and H1975 cells through the AMPK-mTOR and JNK pathways. CK exhibits synergistic effects with other therapeutic agents. For example, the efficacy of cisplatin is improved when co-treated with CK in lung cancer, and this effect is p53-dependent. Combined therapy with CK and gamma-ray radiation is effective in both cell culture and NCI-H460 tumor xenograft models. However, the molecular targets of CK in lung cancer remain unclear.

[0003] Different chemical proteomics approaches have been developed for target identification, including activity-based protein analysis (ABPP), thermal proteomics analysis (TPP), and functional identification of targets through expression proteomics (FITExP). TPP is based on ligand binding that induces the thermostability of target proteins. Based on this phenomenon, a protein blot pattern called Cellular Thermal Drift Detection (CETSA) has been designed to validate drug-target interactions. Savitski et al. extended CETSA in a whole-proteomics manner using liquid chromatography-tandem mass spectrometry (LC-MS / MS), enabling simultaneous monitoring of changes in thousands of proteins induced by drug treatment. Zubarev's team proposed a simplified version of TPP called Integrated Solubility Change in the Whole Proteome (PISA), which measures the difference in Sm (the integral of the area under the unwinding curve for each protein) between different samples. Conversely, ProTargetMiner (PTM) is a proteomics characterization database of anticancer compounds (>50 drugs) based on FITExP. Similar to FITExP, PTM monitors changes in protein expression in cells after prolonged (typically 48 hours) incubation with the drug at LC50 concentrations. Both methods are based on the phenomenon that protein targets exhibit strong and specific regulation upon drug stimulation. To obtain more specific target information, PTM incorporates orthogonal projection into potential structure discriminant analysis (OPLS-DA) for comparison with other anticancer drugs in the database (http: / / proteargetminer.genexplain.com).

[0004] Mitochondria contribute to the production of ATP, macromolecules, apoptosis, and oxidative stress. They are involved in key steps of cancer invasiveness, sustained tumor growth, and cancer progression. The mitochondrial membrane is composed of various phospholipids such as cardiolipin (CL) and phosphatidylethanolamine (PE). Cardiolipin (CL) is unique to mitochondria and is involved in mitochondrial dynamics, participating in the oxidative phosphorylation system and maintaining mitochondrial cristae organization. Dysregulation of CL metabolism is essential in anticancer therapies for different types of cancer. Maintaining the composition, transport, and membrane distribution of these phospholipids is important for mitochondrial homeostasis. A family of proteins containing the related evolutionary and lymphoid interest (PRELI) domain is responsible for the lipid transfer system in mitochondria. This family is conserved in all eukaryotic life from the yeast system, highlighting its importance in cell viability by maintaining mitochondrial lipid homeostasis and playing a crucial role in malignant cancers. Knockdown of PRELID1 inhibits breast cancer cell growth and is associated with apoptosis in liver cancer. There is no direct relationship between PRELID3b and cancer. However, knockdown of PRELID3b in HeLa cells inhibits the expression of PRELID1 and mitochondrial activator protein-like GTPase (OPA1), both of which play important roles in mitochondrial function.

[0005] The mechanisms (multiple mechanisms) of the anticancer activity of ginsenosides remain unclear. One reason is the complex structure of ginsenosides, which presents challenges in their conversion into derivatives. Although CK has been shown to exert anticancer activity in lung cancer cells, its in vivo effects are not efficient, mainly due to its low bioavailability. Modification of CK has not resulted in enhanced anticancer activity. Therefore, modifications to CK to achieve higher activity are desired. Summary of the Invention

[0006] Embodiments of the present invention relate to an anticancer compound, which is a derivative of 20-O-β-(D-glucopyranosyl)-20(S)-protopanaxadiol (CK), wherein at least one glucose hydroxyl group has been converted into at least one acetal group. The acetal group provides a hydrophobic moiety that enhances the binding between the anticancer compound and mitochondrial intermembrane proteins capable of assembling a lipid transfer complex with TRIAP1 in mitochondria. One CK derivative (CKD), CKD-4, exhibits higher cytotoxicity than CK.

[0007] In the implementation scheme, CKD can be an acetal derivative having the following structure:

[0008]

[0009] R is an unsubstituted, monosubstituted, or disubstituted aromatic group. For example, but not limited to, R can be -C6H. x X 5-x Where x is 3 to 5, and X is independently selected from methoxy, nitro, trifluoromethyl, or trifluoromethoxy, wherein the X substituent can be in any of the para, meta, or ortho positions of the acetal. R can be p-methoxyphenyl, phenyl, o-methoxyphenyl, m-methoxyphenyl, p-nitrophenyl, p-trifluoromethyl, or p-trifluoromethoxyphenyl. Other acetals containing CKD can form acetals of 2,2-propane or 2,2-fluoromethane with two adjacent glucose hydroxyl groups.

[0010] In the implementation plan, one or more of the aforementioned anticancer compounds may be included in the anticancer drug. These CKD compounds can be used as alternatives to currently used cancer drugs or in combination with currently used cancer drugs.

[0011] Other embodiments relate to methods for preparing anticancer compounds from CK, wherein CK is combined with an aldehyde, acetal, ketone, or ketal in a reaction mixture to form CKD. The CKD is then separated by conventional separation methods, which may include crystallization, distillation, precipitation, or chromatography. Reagents may be combined in a solvent, and a catalyst may be included in the reaction mixture. Attached Figure Description

[0012] Figure 1 The chemical structures of CK and five CK derivatives (CKD) according to the implementation scheme are shown.

[0013] Figure 2A The structural equation for the synthesis of CKD-2 according to one embodiment is shown, wherein a is 2,2-dimethoxypropane:acetone (1:5) and p-toluenesulfonic acid.

[0014] Figure 2B The structural equation for the synthesis of CKD-3 is shown, where b is 2,2-dimethoxypropane and p-toluenesulfonic acid.

[0015] Figure 2C The structural equations for the synthesis of CKD-4 and CKD-5 according to one embodiment are shown, where c is a boron trifluoride diethyl ether complex in anhydrous dichloromethane.

[0016] Figure 3 Bar graphs showing CK and CKD-4 uptake in A549 cells after treatment with 10 μM CK or CKD-4 for 4 or 24 hours and extracted with 70% MeOH (n = 3 biologically independent replicates).

[0017] Figure 4 The image shows flow cytometry plots of A549 cells collected and stained with annexin V-FITC and propidium iodide after 24 hours of treatment with 60 μM CK or 20 μM CKD-4.

[0018] Figure 5 The image shows bright-field micrographs of MS 1 cells treated with control and CK or CKD-4, seeded in a matrix gel-coated 96-well plate, wherein the cells were treated with various concentrations of CK or CKD-4 for 4 hours.

[0019] Figure 6 Images showing lung cancer-derived organoids after treatment with DMSO, carboplatin, cisplatin, paclitaxel, and CKD-4 at concentrations of 5, 10, 20, 35, and 50 μM.

[0020] Figure 7 The use of CellTiterGlo™ showed CKD-4 inhibition at 72 hours. Figure 7 Bar graphs of organoid proliferation derived from patients, where DMSO-treated cells were considered to have 100% viability (control), plotted as mean ± SD.

[0021] Figure 8AThe graph shows the changes in tumor volume in vivo due to cytotoxicity when treated with 5 mg / kg CK and 5 mg / kg CKD-4, where data are mean ± SEM.**, P ≤ 0.01, *, P ≤ 0.05.

[0022] Figure 8B The graph shows the changes in body weight and volume of mice treated with 5 mg / kg CK and 5 mg / kg CKD-4.

[0023] Figure 9A A volcano plot showing the PISA analysis of intracellular CK.

[0024] Figure 9B A volcano plot showing the PISA analysis of intracellular CKD-4.

[0025] Figure 10 The graph shows the OPLS-DA analysis of CK relative to 9 drugs and controls in A549 cells, with the top 20 upregulated proteins marked in red, the top 20 downregulated proteins marked in blue, and RELID3b marked in green.

[0026] Figure 11 The graph shows the OPLS-DA analysis of CKD-4 in A549 cells relative to 9 drugs and controls, with the top 20 upregulated proteins marked in red, the top 20 downregulated proteins marked in blue, and RELD3b marked in green.

[0027] Figure 12A This is a scatter plot showing the fold change in protein expression of CKD-4 cells relative to DMSO-treated cells (vs. CK cells relative to DMSO-treated cells) in PTM.

[0028] Figure 12B This is a scatter plot showing the fold change in protein expression of CK relative to DMSO-treated cells (vs. methotrexate relative to DMSO-treated cells) in PTM.

[0029] Figure 12C This is a scatter plot showing the fold change in protein expression of CKD-4 relative to DMSO-treated cells (vs. methotrexate relative to DMSO-treated cells) in PTM.

[0030] Figure 13A This is a graph showing the cumulative gradation of fold changes analyzed by PISA and PTM in the presence of CK.

[0031] Figure 13B This is a graph showing the cumulative gradation of fold changes analyzed by PISA and PTM in the presence of CKD-4.

[0032] Figure 14ACellular thermal drift of PRELID3b is shown in A549 cells when treated with CK or CKD-4. A549 cells were treated with 50 μM CK or 15 μM CKD-4 for one hour, during which eight aliquots of lysate were heated to the indicated temperature.

[0033] Figure 14B The thermal drift of PRELID1 in A549 cells was observed when treated with CK or CKD-4. A549 cells were treated with 50 μM CK or 15 μM CKD-4 for one hour, during which eight aliquots of lysate were heated to a specified temperature.

[0034] Figure 15A Bar graphs of CK and CKD-4-induced apoptosis quantified by mitochondrial central phospholipids in A549 cells after 24 hours of treatment with CK or CKD-4, as mean ± SD, **, P ≤ 0.01, *, P ≤ 0.05.

[0035] Figure 15B Western blot analysis of cytochrome c cytosol fractions after treatment with 50 μM CK and 15 μM CKD-4 for 24 and 48 hours is shown, with COX IV indicating no mitochondrial contamination.

[0036] Figure 15C Bar graphs of A549 cell viability after treatment with 100 μM CK or 12.5 μM CKD-4 for 24 hours in DMEM or DMEM supplemented with phosphatidylglycerol (PG), as mean ± SD, ***P≤0.001.

[0037] Figure 15D Western blot analysis of cytochrome c cytosol fractions in culture medium, with or without PG, after treatment with 50 μM CK and 15 μM CKD-4 for 24 hours, with COX IV indicating no mitochondrial contamination.

[0038] Figure 16A A computer-generated image showing the docking posture of CK and PRELID3b, where the protein surface represents the electrostatic potential surface, with red to blue inferring the change in potential from negative to positive and the interaction between CK and surrounding residues.

[0039] Figure 16B A computer-generated image showing the interaction of amino acid residues at the CK-PRELID3B binding site.

[0040] Figure 16C This is the rmsd distribution of CK during a 50-nanosecond MD simulation.

[0041] Figure 16DA computer-generated image showing the docking posture of CKD-4 and PRELID3b, where the protein surface represents the electrostatic potential surface, with red to blue inferring the change in potential from negative to positive and the interaction between CKD-4 and surrounding residues.

[0042] Figure 16E A computer-generated image showing the interaction of amino acid residues at the CKD-4 binding site with PRELID3b.

[0043] Figure 16F This is the rmsd distribution of CKD-4 during a 50-nanosecond MD simulation.

[0044] Figure 17A The binding curves of biolayer interferometry using different concentrations of his-labeled PRELID3b-TRIAP1-MBP and CK are shown.

[0045] Figure 17B This plot shows the Kd calculated by steady-state analysis, assuming the kinetics are a 1:1 ratio of his label PRELID3B-TRIAP1-MBP to CK.

[0046] Figure 17C The binding curves of biolayer interferometry using different concentrations of his-labeled PRELID3b-TRIAP1-MBP and CKD-4 are shown.

[0047] Figure 17D The Kd plot is shown, calculated by steady-state analysis, assuming the kinetics are a 1:1 ratio of his label PRELID3b-TRIAP1-MBP to CKD-4.

[0048] Figure 17E The binding curves of biolayer interferometry using different concentrations of his-tagged MBP and CK are shown.

[0049] Figure 17F The binding curves of biolayer interferometry using different concentrations of his-tagged MBP and CKD-4 are shown. Detailed Implementation

[0050] In one embodiment, a derivative of ginsenoside 20-O-β-(D-glucopyranosyl)-20(S)-protopanaxadiol (CKD) is prepared, which improves the activity of CK in anticancer compositions. The derivative has a different number of protected glucose hydroxyl groups. According to one embodiment, a derivative CKD-4 exhibits higher cellular uptake than CK in A549 cells and shows significantly greater cytotoxicity than CK in vitro and in vivo. The molecular target of CK and CKD-4 is PRELID3b, indicating a significant relationship between PRELID3b and cancer. The anticancer activity of ginsenosides increases with decreasing sugar content, with ginsenosides having one sugar moiety, such as Rh2 and CK, exhibiting the strongest cytotoxicity, and the glucose moiety in CK is essential for its interaction with its molecular target. Both CK and CKD-4 interact with PRELID3b and appear to bind to a lipid-binding pocket of the protein. CKD-4 exhibits stronger anticancer activity than CK both in vitro and in vivo. CK and CKD-4 share similar mechanisms because both compounds bind to the mitochondrial intermembranous proteins PRELID3b and PRELID1. Based on these interactions, both compounds appear to disrupt normal phospholipid transport, leading to decreased mitochondrial cytochrome c (CL) levels. Treatment with these anticancer compounds promotes cytochrome c release and mitochondrial division, resulting in mitochondrial dysfunction and apoptosis.

[0051] According to embodiments of the present invention, novel anticancer compounds are formed by modifying natural products and studying their novel modes of action (MOAs). New and potent CK derivatives, such as, but not limited to, CKD-4, have shown stronger cytotoxicity and higher inhibition rates in animal models than their derived natural products, and exhibit relatively low toxicity. Using proteomics methods combined with whole-proteome integral solubility alteration (PISA) and ProTargetMiner (PTM), the molecular target PRELID3b of CK, CKD-4, and other CK derivatives present in lung cancer cell lines was identified. PRELID3b is a family of proteins that regulate lipid metabolism in mitochondria, and emerging evidence suggests their importance in tumor progression. Novel anticancer therapies involve novel therapeutic strategies targeting PRELID3b according to embodiments of the present invention.

[0052] 20-O-β-(D-glucopyranosyl)-20(S)-protopanaxadiol (CK) derivatives are produced by reacting with one or more hydroxyl groups of the monosaccharide unit of CK:

[0053]

[0054] Derivatization can occur at any single hydroxyl group or at any number of hydroxyl groups. Multiple hydroxyl groups can be on adjacent carbons or separated by chains of two or more carbons. In one embodiment, two hydroxyl groups form an acetal or ketal upon derivatization. The derivative provides a molecular moiety that provides an additional hydrophobic portion of the molecule, which provides a favorable balance between hydrogen bonding and van der Waals interactions with the protein moiety targeted by the anticancer agent.

[0055] In one embodiment, the anticancer compound is an aromatic acetal derivative of CK having the following structure:

[0056]

[0057] X is H, R1 is independently H, F, or methyl, and R2 is independently F, methyl, or an unsubstituted, monosubstituted, or disubstituted aromatic group. The substituents of the monosubstituted or disubstituted phenyl or other aromatic groups can be, but are not limited to, methoxy, nitro, trifluoromethyl, or trifluoromethoxy, wherein one or more of these substituents may be present at the ortho, meta, or para position relative to the sugar ring. A non-exhaustive list of CKDs includes CKD-2 (2,2-dimethyl), CKD-4 (p-methoxyphenyl), CKD-5 (phenyl), CKD-6 (o-methoxyphenyl), CKD-7 (m-methoxyphenyl), CKD-8 (p-nitrophenyl), CKD-9 (p-trifluoromethylphenyl), CKD-10 (p-trifluoromethoxyphenyl), and CKD-11 (2,2-difluoro).

[0058] In other embodiments, CKD-x can be a compound with the following structure:

[0059]

[0060] The substituent X can be F or other halogens, for example:

[0061]

[0062] Both PRELID3b and PRELID1 belong to the same family of lipid transfer complexes that assemble with TRIAP1 in mitochondria. Specifically, PRELID3b / TRIAP1 carries PS from the endoplasmic reticulum (ER) to the inner mitochondrial membrane (IM) for PE synthesis, and PRELID1 / TRIAP1 transports PA across the intermembrane space (IMS) for CL synthesis. Therefore, the interaction / conjugation of CK or CKD-4 with these two complexes appears to induce altered accumulation of CL and PE. Although these two complexes specifically transport precursors of PE or CL, the absence of PRELID1 or PRELID3b in HeLa cells affects the accumulation of both CL and PE. A decrease in CL levels occurred in A549 cells after treatment with CK or CKD-4. This decrease in CL triggered the release of cytochrome c from the mitochondria, leading to apoptosis. During CK or CKD-4-induced apoptosis, the release of cytochrome c from the mitochondria was restored upon replenishment of CL levels by supplementation with its precursor phosphatidylglycerol (PG). Furthermore, the inhibitory effect of CK or CKD-4 on cell viability was also suppressed upon PG supplementation.

[0063] The cytotoxicity of CK was measured using the MTT assay for various cancer cell lines, as shown in Table 1 below. CK was cytotoxic and non-selective at submolar levels against lung cancer (NCI-H460), cervical cancer (HeLa), liver cancer (MHCC97-L and HepG2), and colon cancer (SW480).

[0064] Table 1. Cytotoxicity of CK at 72 hours in various cancer cell lines (IC50) 50 (μM).

[0065]

[0066] like Figure 1 As shown, five CK derivatives were synthesized by protecting different numbers of hydroxyl groups. CKD-1 was synthesized by acetylation of all six hydroxyl groups. The synthesis of derivatives CKD-2, CKD-3, CKD-4, and CKD-5 is shown in [the figure]. Figures 2A to 2C CKD-1 is highly hydrophobic and exhibits poor solubility even in dimethyl sulfoxide (DMSO). Four other derivatives were synthesized by protecting the hydroxyl group via cycloacetal formation. The cytotoxicity of these four compounds was measured by MTT assay in NCI-H460 cells and is shown in Table 2 below. Of the five newly synthesized compounds, CKD-4 was the most potent, with an IC50 concentration of 100%. 50 The value is 16.5 μM. CKD-2's IC 50The values ​​were almost identical to CK, while CKD-3 lost its anticancer activity. Furthermore, the cytotoxicity of CKD-4 was further validated in other lung cancer cell lines and normal lung fibroblasts (CCD-19Lu). CKD-4 was three times more effective than CK, although no difference was shown between the two in cancer cell lines and normal lung fibroblasts, as shown in Table 3 below.

[0067] Table 2. Cytotoxicity of CK and its derivatives in NCI-H460 cells treated with CK and CKD-x derivatives for 48 hours (IC50). 50 (μM).

[0068]

[0069] Table 3. Cytotoxicity of CK and CKD-4 at 24, 48, and 72 h in different lung cancer cell lines and normal lung fibroblasts (IC50). 50 (μM).

[0070]

[0071] Cellular uptake of CK and CKD-4 was monitored at 4 and 24 hours, and the formation of CK lysed from CKD-4 was investigated. Cellular uptake in A549 cells treated with 10 μM CK or CKD-4 for 24 hours was extracted and quantified by LC-MS / MS. Figure 3 The results showed that CKD-4 uptake by cells was higher than that of CK at both 4 and 24 hours. To test the stability of CKD-4 in cells, the levels of CK in cells treated with CKD-4 were also measured. Figure 3 As shown, only negligible CK levels were found (accounting for 1.33% of the total CKD-4 uptake after 4 hours and 1.05% of the total CKD-4 uptake after 24 hours). These data indicate that CKD-4 is stable after 24 hours, and that CKD-4 uptake is higher than CK, which may be one of the reasons why CKD-4 is more cytotoxic than CK.

[0072] The effects of CK and CKD-4 on apoptosis and cell death were demonstrated by flow cytometry analysis of A549 cells treated with 60 μM CK and 20 μM CKD-4 for 48 hours. Figure 4 This indicates a significant increase in the percentage of apoptotic cells for CK and CKD-4. Angiogenesis is another hallmark of cancer, playing a crucial role in metastasis and tumor growth. Figure 5The results showed that tube formation was inhibited by 40 μM CK compared to untreated MS 1 cells, with a more pronounced inhibitory effect when treated with 15 μM CKD-4. Cancer stem cells, due to their self-renewal capacity, have become a major cause of tumor recurrence and poor prognosis. The efficacy of CKD-4 in organoids derived from lung cancer patients was demonstrated using the CellTiterGlo assay, with three existing-technology clinical anticancer drugs—carboplatin, cisplatin, and paclitaxel—serving as positive controls. The doses of these three drugs were sufficient to trigger inhibition of 30% cell viability. Figure 6 and Figure 7 This indicates that treatment with 20 μM CKD-4 resulted in less than 30% inhibition of cell viability. Figure 7 The description indicates that CKD-4 significantly inhibits the growth of patient-derived organoids in a dose-dependent manner.

[0073] like Figure 8A As shown, treatment of nude mice carrying NCI-H460 xenografts with 5 mg / kg CK and CKD-4 via intravenous injection every two days showed significant tumor volume inhibition in the CKD-4 group, while almost no difference was observed between the CK group and the solvent control group. Figure 8B As shown, no significant change in body weight was observed throughout the treatment. Therefore, CKD-4 is more effective than CK both in vitro and in vivo.

[0074] Two label-free chemical proteomics methods, Proteome Integrated Stability Change Assay (PISA) and ProtargetMiner (PTM), are alternative strategies for studying CK and CKD-4. PISA is a turbocharged approach developed for thermal proteomics analysis (TPP), based on the principle of observing ligand-induced protein stability upon temperature elevation. PTM is used to compare proteome markers of CK and CKD-4 with the PTM database to reveal their modes of action. Figure 9A and 9B Provides an overview of PISA results in cells with CK and CKD-4, where the data is displayed as a volcano plot, with the X-axis representing the log2ΔS of the drug relative to DMSO treatment. m And the Y-axis is the log10p value. Figure 10 and 11 In this study, the PTM results of CK and CKD-4 in cells were compared with those of nine other anticancer drugs in A549 cells from the database, and the results were plotted for OPLS-DA analysis.

[0075] To investigate whether CKD-4, after modification, shares the same mechanism as CK, protein expression distribution maps (drug-to-DMSO treatment ratio) in PTM were compared. Figure 12AAs shown, the expression distribution maps of CKD-4 and CK are significantly correlated, where R 2 It equals 0.6877. Conversely, as shown below: Figure 12B and 12C As shown, CK and CKD-4 did not show a significant association with methotrexate because both have low R... 2 value.

[0076] Based on the absolute value of log2ΔSm of the drug relative to DMSO treatment, the most likely target candidate proteins are graded in PISA, and the PTM grading is based on the fold change of the drug relative to DMSO treatment. The grades from both experiments are summed to provide a list of candidate drug targets. Figure 13A and 13B As indicated, PRELID3B and JUN stood out as the first and second candidates. Therefore, the study focused on PRELID3B. Interestingly, PRELID1 was stabilized by CK and CKD-4 in PISA, although this protein was not identified in PTM assays. The stability of PRELID3B and PRELID1 was confirmed by Western blotting using cell thermal drift assays, which showed the thermal stability of both proteins to CK and CKD-4 upon treatment in A549 cells, as... Figure 14A and 14B As indicated.

[0077] PRELID3b and PRELID1 are PRELI proteins. These proteins, along with TRIAP1, are located in the intermembrane space responsible for regulating phospholipid metabolism. Therefore, cardiolipin probes are used to measure mitochondrial cardiolipin levels after drug treatment. In A549 cells, cardiolipin levels were significantly reduced after 24 hours of treatment with CK or CKD-4, such as... Figure 15A As shown. Decreased cardiolipin levels in PRELI or TRIAP1-depleted cells appear to accelerate cytochrome c release, making cells more susceptible to apoptosis. Consistent with this phenomenon, cytochrome c levels in the cytoplasm significantly increased after 24 hours of treatment with CK and CKD-4, as shown. Figure 15B As indicated, cardiolipin is synthesized through a series of steps, in which phosphatidylglycerol (PG) serves as a precursor in the final step catalyzed by cardiolipin synthase. To demonstrate that a decrease in cardiolipin contributes to apoptosis, PG was artificially provided in the culture medium to restore CL levels during treatment with CK and CKD-4. NBB assays showed that, in the presence of PG supplementation, cell viability was almost doubled after 24 hours of treatment with 100 μM CK and 12.5 μM CKD-4, respectively. Figure 15C As indicated, CKD-4 releases reduced levels of cytochrome c, although this was not observed in CK, as... Figure 15D As indicated.

[0078] Computational docking and molecular dynamics simulations provide models of the interaction between CK or CKD-4 and PRELID3b. Based on the chemical similarity between CK and phosphatidylserine lipids, the lipid binding cavity of PRELID3b is assumed to be a potential binding bag for CK and CKD-4. For both compounds, 50-ns MD simulations were performed on the top posture predicted by docking for stability assessment, which is an efficient method for identifying the correct binding posture and improving the docking prediction. The calculation results are summarized in Table 4 below, and in... Figures 16A to 16F The illustration in the diagram is as follows.

[0079] Table 4. Docking results of CK and CKD-4 with PRELID3b

[0080]

[0081]

[0082] The CK molecules bound to the PRELID3b dimer are quite stable, as supported by the ligand rsmd analysis plot. The average rmsd values ​​of the two CK molecules during the last 10 nanoseconds were... and It is below the threshold that is generally acceptable for stable binding. The PRELID3B / CKD-4 complex also showed up in Figure 16A and 16B The average RMSD values ​​for the last 10 nanoseconds of the two CKD-4 molecules are as follows: and Regarding the stable postures of CK and CKD-4, AutoDock estimates the binding energies to be -6.55 kcal / mol and -7.52 kcal / mol, respectively. In terms of dissociation constant, CKD-4 has a 5-fold higher affinity than CK.

[0083] Within the lipid-binding pocket, the orientation of both CK and CKD-4 is similar to that of PS lipids, such as... Figure 16C and 16FThe illustration shows the hydrophilic glucose head embedded in a deep cavity, while the hydrophobic sterane portion is attached to a hydrophobic wall formed by residues Val33, Leu55, Thr57, Trp59, Thr76, and Val78. The main difference between CK and CKD-4 lies in the position of the glucose head and the olefin chain. In CK, the glucose head is close to Glu80 and forms hydrogen bonds with the carboxyl oxygen atom of Glu80 and the phenolic group of Tyr112. The olefin chain is embedded under the α3 helix. However, in the case of CKD-4, the glucose unit protected by p-methoxybenzyl (PMB), rather than the olefin chain, is embedded under the α3 helix. The PMB group is surrounded by residues Phe10, Lys27, Leu110, Tyr112, Leu124, Gln126, and α3 helical residues including Gly156, Arg157, and Met160. The hydroxyl groups on the glucose ring form hydrogen bonds with the side chains of Tyr26 and Asn30. The olefin chain is attached to a hydrophobic bag formed by residues Val33, Thr57, Trp59, and Val78.

[0084] Clearly, the large PMB group creates more hydrophobic contact between CKD-4 and PRELD3b, leading to an enhanced binding affinity for CKD-4. This hypothesis is supported by the breakdown of the binding energies of the two compounds. The AutoDock scoring function estimates the binding energy by combining van der Waals interactions, hydrogen bonds, electrostatic interactions, desolvation effects, and ligand torsion energies. By comparing these terms, the combination of van der Waals forces and hydrogen bonds (ΔEvdw+HB) is the major contributor to the difference in binding energy (3.26 kcal / mol). Only two hydrogen bonds were found between CKD-4 and the protein, in contrast to the four hydrogen bonds between CK and the protein. Therefore, the van der Waals interactions between CKD-4 and PRELD3b outweigh the loss of hydrogen bonds, resulting in an increase in the total binding energy.

[0085] To validate the docking analysis, the interaction between PRELID3b and CK or CKD-4 was measured using biolayer interferometry (BLI). The recombinant protein was purified by co-expression of PRELID3b with a hexahistidine tag and TRIAP1, a p53-regulated protein conjugated with a maltose-binding protein (MBP). The reported crystal structure of PRELID3b is complexed with TRIAP1, and they form an exchange dimer. PRELID3b is inherently unstable in the absence of TRIAP1 and is readily degraded by mitochondrial proteases. The maltose-binding protein (MBP) is an essential component for improving the solubility of the entire protein complex.

[0086] The sensor used was coated with NiNTA, allowing proteins with a hexahistine tag to be immobilized on it. After protein immobilization, the sensor was incubated with different concentrations of the compound in ascending order. The stoichiometric ratio of CK or CKD-4 to PRELD3b was 1:1, but the kinetic patterns differed for both cases. CK interacted with PRELD3b at a rapid on / off rate, while the rate was slower with CKD-4. This assay using PRELD3b and CKD-4 yielded a Kd of 5.3 μM, which is almost 7 times smaller than the Kd between CK and PRELD3b (55.7 μM). Figures 17A to 17D As shown. To demonstrate that the interaction is not attributable to the presence of MBP, the interaction between MBP and CK or CKD-4 was examined. The results show that no dependent response occurred for CK or CKD-4, as... Figure 17E and 17F The results are consistent with the docking analysis, which shows that the interaction between CKD-4 and PRELID3b is stronger than that between CK.

[0087] Materials and methods

[0088] Synthesis of CKD-2

[0089] Under argon atmosphere, p-toluenesulfonic acid (13.9 mg, 0.081 mmol) was added to a mixture of acetone and 2,2-dimethoxypropane (5:1, v / v, 20 mL) containing CK (50.3 mg, 0.081 mmol). The reaction mixture was stirred at room temperature for 1.5 h, and the progress was monitored by TLC. Triethylamine (1 mL) was added to quench the reaction. The mixture was concentrated, diluted with dichloromethane, and washed several times with water. The organic fraction was dried over anhydrous MgSO4, then filtered and concentrated under reduced pressure. Purification by rapid column chromatography gave CKD-2 (3.8 mg, 0.0057 mmol, 7% yield). 1 H NMR (400MHz, CDCl3), δppm 5.29(m,1H),4.57(d,J=7.7Hz,1H),3.84(m,2H),3.65(m,3H),3.39(t,J=4.4Hz,1H),3.21(t,J=4.4Hz,2H),1.68-0.78(m,30H). 13C NMR (600MHz, CDCl3), δppm 124.45,99.75,97.48,84.49,78.86,74.55,73.84,72.83,70.59,67.17,62.26,55.79,51.70,51.42,49.77,47.95,39.75,38.91,38.87,37.06,35.41,34.71,30.61,30.44,29.70,29.08,28.01,27.38,26.66,25.71,22.27,21.62,19.13,18.25,17.75,16.94,16.12,15.71,15.37. MS(ESI): For C 37 H 62 The calculated m / z value for O8 is 634.4445 [M+Na]. + The measured value is 685.30.

[0090] Synthesis of CKD-3

[0091] Under argon atmosphere, p-toluenesulfonic acid (13.6 mg, 0.079 mmol) was added to a solution of 2,2-dimethoxypropane (20 mL) and CK (49.3 mg, 0.079 mmol). The mixture was stirred overnight at room temperature, and the reaction was terminated by adding triethylamine (1 mL). Dichloromethane was added to dilute the mixture, and the mixture was washed with water. The mixture was dried over anhydrous MgSO4, filtered, and concentrated under reduced pressure. Rapid column chromatography purification finally yielded CKD-3 (11.3 mg, 0.016 mmol, 20.3%). 1 HNMR (400MHz, CDCl3), δppm 5.15 (m, 1H), 4.91 (d, J = 7.6Hz, 1H), 4.06 (m, 1H), 3.95 (m, 1H), 3.63 (m, 1H), 1.70-0.78 (m, 36H). 13CNMR (400MHz, CDCl3), δppm 125.30,109.63,108.87,94.87,79.28,78.79,78.53,78.12,76.72,76.36 ,70.87,67.61,55.93,50.77,50.33,48.49,48.37,39.71,39.13,38.97,3 7.28, 35.90, 34.79, 32.73, 29.72, 28.71, 28.04, 27.40, 26.86, 26.68, 25.78, 25.53, 25.44, 24.71, 22.09, 18.28, 17.74, 16.35, 16.26, 15.45, 15.35. HRMS(ESI): For C 42 H 70 The calculated m / z value for O8 is 702.5071 [M+Na]. + The measured value is 725.4977.

[0092] Synthesis of CKD-4

[0093] Add 5 mL of p-methoxybenzaldehyde and 100 μL of the boron trifluoride diethyl ether complex to a round-bottom flask and suspend in methanol (80 mL). Stir the mixture and heat under reflux overnight. Quench the reaction with triethylamine (10 mL). Extract the product with diethyl ether and wash with K₂CO₃ solution. Dry the mixture with anhydrous MgSO₄, then filter and concentrate under reduced pressure. Purify by rapid column chromatography to give 1-(dimethoxymethyl)-4-methoxybenzene.

[0094] Under argon atmosphere, 1-(dimethoxymethyl)-4-methoxybenzene (2 mL) and boron trifluoride diethyl ether complex (6 μL) were added to a suspension of CK (50 mg, 0.08 mmol) in anhydrous dichloromethane (5 mL). The resulting mixture was stirred overnight at room temperature, and the reaction progress was monitored by TLC until all CK was consumed. Triethylamine (1 mL) was added to terminate the reaction. The product was extracted with dichloromethane and washed with K2CO3 solution. The mixture was dried over anhydrous MgSO4, filtered, and concentrated under reduced pressure. Rapid column chromatography purification finally yielded CKD-4 (19.3 mg, 0.026 mmol, 32.5% yield). 1H NMR (400MHz, CDCl3), δppm 7.43(d,J=8.7Hz,2H),6.89(d,J=8.8Hz,2H),5.49(s,1H),5.12(m,1H),4.64 (d,J=7.7Hz,1H),4.26(m,1H),3.57(m,2H),3.44(m,2H),1.80-0.77(m,24H). 13 C10 NMR (600MHz, CDCl3), δppm 160.18, 131.63, 129.54, 127.60, 124.47, 113.64, 101.70, 97.40, 84.51, 80.16, 78.85, 74.43, 73.54, 70.57, 68.76, 66.19, 55.80, 55.30, 51.65, 51.44, 49.79, 48.00, 39.47, 38.89, 37.04, 35.39, 34.71, 30.58, 30.36. HRMS (ESI): for C10 44 H 68 The calculated m / z value for O9 is 740.4863 [M+Na]. + The measured value is 763.4765.

[0095] Synthesis of CKD-5

[0096] Methanol (60 mL), benzaldehyde (5 mL), and boron trifluoride diethyl ether complex (2 mL) were mixed, stirred, and heated under reflux overnight. Triethylamine (10 mL) was added to quench the reaction. The product was extracted with diethyl ether and then washed with K₂CO₃ solution. The mixture was dried over anhydrous MgSO₄, filtered, and concentrated under reduced pressure. Rapid column chromatography was used to purify the final product to (dimethoxymethyl)benzene.

[0097] Under argon atmosphere, a suspension of CK (31.6 mg, 0.05 mmol) in anhydrous dichloromethane (5 mL) was added to (dimethoxymethyl)benzene (1 mL) and a boron trifluoride diethyl ether complex (6 μL). The resulting mixture was stirred overnight at room temperature, and the reaction was monitored by TLC. The reaction was terminated with triethylamine (1 mL). The product was extracted with dichloromethane and washed with K₂CO₃ solution. The mixture was dried over anhydrous MgSO₄, filtered, and concentrated under reduced pressure. Rapid column chromatography was used to purify the final product to CKD-5. 1¹H NMR (400MHz, CDCl₃), δppm 7.50–7.47 (m, 2H), 7.36–7.34 (m, 3H), 5.53 (s, 1H), 5.10 (t, J = 6.7Hz, 1H), 4.64 (d, J = 7.7Hz, 1H), 4.27–4.25 (m, 1H), 3.81–3.77 (m, 2H), 3.59–3.54 (m, 2H). MS (ESI): for C 43 H 66 The calculated m / z value for O8 is 710.4758 [M+Na]. + The measured value is 733.40.

[0098] All patents, patent applications, provisional applications, and publications cited herein, including all figures and tables, are incorporated in their entirety by reference to the extent that they do not contradict the explicit teachings of this specification.

[0099] It should be understood that the embodiments and implementations described herein are for illustrative purposes only, and various modifications or variations thereof will be suggested to those skilled in the art, and such modifications or variations will be included within the spirit and scope of this application and the appended claims. Furthermore, any element or limitation of any invention or implementation thereof disclosed herein may be combined with any and / or all other elements or limitations disclosed herein (alone or in any combination) or any other invention or implementation thereof, and all such combinations are considered within the scope of the invention but are not limited thereto.

[0100] References

[0101] Wong AS, Che CM, Leung KW. Recent advances in ginseng as cancertherapeutics: a functional and mechanistic overview. Nat Prod Rep. 32, 256-72 (2015).

[0102] C.Gridelli et al.,Non-small-cell lung cancer.Nature Reviews DiseasePrimers 1,15009(2015).

[0103] L.Zhou,Z.K.Li,C.Y.Li,Y.Q.Liang,F.Yang,Anticancer propertiesandpharmaceutical applications of ginsenoside compound K:A review.ChemBiolDrug Des 99,286-300(2022).

[0104] B.H.Han et al.,Degradation of ginseng saponins under mild acidicconditions.44,146-149(1982).

[0105] D.-H.J.J.o.g.r.Kim,Gut microbiota-mediated pharmacokinetics ofginsengsaponins.42,255-263(2018).

[0106] L.Chen et al.,Ginsenoside compound K sensitizes human colon cancercells toTRAIL-induced apoptosis via autophagy-dependent and-independentDR5upregulation.Cell Death Dis 7,e2334(2016).

[0107] K.Zhang,Y.Li,Effects of ginsenoside compound K combined withcisplatin onthe proliferation,apoptosis and epithelial mesenchymal transitionin MCF-7 cellsof human breast cancer.Pharm Biol 54,561-568(2016).

[0108] Z.Z.Zheng et al.,Compound K-induced apoptosis of humanhepatocellularcarcinoma MHCC97-H cells in vitro.Oncol Rep 32,325-331(2014).

[0109] C.Li et al.,Ginsenoside metabolite compound K induces apoptosisandautophagy in non-small cell lung cancer cells via AMPK-mTOR andJNKpathways.Biochem Cell Biol 97,406-414(2019).

[0110] H.F.Chen et al.,Ginsenoside compound K inhibits growth of lung cancercellsvia HIF-1α-mediated glucose metabolism.Cell Mol Biol(Noisy-le-grand)65,48-52(2019).

[0111] Y.Li et al.,Ginsenoside metabolite compound K enhances the efficacyofcisplatin in lung cancer cells.J Thorac Dis 7,400-406(2015).

[0112] S.Chae et al.,Effect of compound K,a metabolite of ginseng saponin,combinedwith gamma-ray radiation in human lung cancer cells in vitro and invivo.J AgricFood Chem 57,5777-5782(2009).

[0113] S.Ziegler,V.Pries,C.Hedberg,H.Waldmann,Target identification forsmallbioactive molecules:finding the needle in the haystack.Angew Chem Int EdEngl52,2744-2792(2013).

[0114] M.M.Savitski et al.,Tracking cancer drugs in living cells by thermalprofiling ofthe proteome.Science 346,1255784(2014).

[0115] A.Chernobrovkin,C.Marin-Vicente,N.Visa,R.A.Zubarev,FunctionalIdentification of Target by Expression Proteomics(FITExP)revealsproteintargets and highlights mechanisms of action of small moleculedrugs.Sci Rep 5,11176(2015).

[0116] R.Jafari et al.,The cellular thermal shift assay for evaluating drugtargetinteractions in cells.Nat Protoc 9,2100-2122(2014).

[0117] M.Gaetani et al.,Proteome Integral Solubility Alteration:A High-ThroughputProteomics Assay for Target Deconvolution.J Proteome Res 18,4027-4037(2019).

[0118] A.A.Saei et al.,ProTargetMiner as a proteome signature library ofanticancermolecules for functional discovery.Nat Commun 10,5715(2019).

[0119] D.C.Wallace,Mitochondria and cancer.Nature Reviews Cancer 12,685-698(2012).

[0120] N.Stepanyants et al.,Cardiolipin's propensity for phase transitionand itsreorganization by dynamin-related protein 1 form a basis formitochondrialmembrane fission.Mol Biol Cell 26,3104-3116(2015).

[0121] J.Dudek,Role of Cardiolipin in Mitochondrial Signaling Pathways.FrontCellDev Biol 5,90(2017).

[0122] S.T.Ahmadpour et al.,the Mitochondrial Signature Lipid:Implication inCancer.Int J Mol Sci 21(2020).

[0123] X.Miliara et al.,Structural insight into the TRIAP1 / PRELI-like domainfamilyof mitochondrial phospholipid transfer complexes.EMBO Rep 16,824-835(2015).

[0124] B.Y.Kim et al.,Effects of PRELI in Oxidative-Stressed HepG2 Cells.AnnClinLab Sci 45,419-425(2015).

[0125] A.E.Gillen et al.,Alternative Polyadenylation of PRELID1RegulatesMitochondrial ROS Signaling and Cancer Outcomes.Mol Cancer Res 15,1741-1751(2017).

[0126] T.MacVicar et al.,Lipid signalling drives proteolytic rewiring ofmitochondriaby YME1L.Nature 575,361-365(2019).

[0127] J.P.Medema,Cancer stem cells:the challenges ahead.Nat Cell Biol 15,338-344(2013).

[0128] X.Miliara et al.,Structural determinants of lipid specificity withinUps / PRELIlipid transfer proteins.Nat Commun 10,1130(2019).

[0129] P.Mishra et al.,Metabolic regulation of mitochondrial dynamics.J CellBiol 212,379-387(2016).

[0130] T.Ban et al.,Molecular basis of selective mitochondrial fusion byheterotypicaction between OPA1 and cardiolipin.Nat Cell Biol 19,856-863(2017).M.J.Aaltonen et al.,MICOS and phospholipid transfer by Ups2-Mdm35organize membrane lipid synthesis in mitochondria.J Cell Biol 213,525-534(2016).

[0131] K.Quan et al.,Rapid preparation of rare ginsenosides by acidtransformation andtheir structure-activity relationships against cancercells.Sci Rep 5,8598(2015).K.G.CSH Tan et al.,Thermal proximity coaggregationfor system-wideprofiling of protein complex dynamics in cells.Science 359,1170-1177(2018).K.V.Huber et al.,Proteome-wide drug and metaboliteinteraction mapping bythermal-stability profiling.Nat Methods 12,1055-1057(2015).

[0132] S.Sridharan et al.,Proteome-wide solubility and thermal stabilityprofilingreveals distinct regulatory roles for ATP.Nature Communications 10,1155(2019).J.X.Huang et al.,High throughput discovery of functional proteinmodificationsby Hotspot Thermal Profiling.Nat Methods 16,894-901(2019).

[0133] I.Becher et al.,Pervasive Protein Thermal Stability Variation duringthe CellCycle.Cell 173,1495-1507.e1418(2018).

[0134] C.Potting et al.,TRIAP1 / PRELI complexes prevent apoptosis bymediatingintramitochondrial transport of phosphatidic acid.Cell Metab 18,287-295(2013).W.C.Potting et al.,Regulation of mitochondrial phospholipids byUps1 / PRELI-like proteins depends on proteolysis and Mdm35.EMBO J.29,2888-2898(2010).J.W.Ma et al.,The role of mitochondrial dynamics in human cancers.AmJCancer Res 2020;10(5):1278-1293 10,1278-1293(2020).

[0135] G.R.Anderson et al.,Dysregulation of mitochondrial dynamics proteinsare atargetable feature of human tumors.Nat Commun 9,1677(2018).

[0136] M.R.McKeller et al.,Vital function of PRELI and essential requirementof itsLEA motif.Cell Death Dis 1,e21(2010).

[0137] D.C.Chan,Fusion and fission:interlinked processes critical formitochondrialhealth.Annu Rev Genet 46,265-287(2012).

[0138] C.R.Chang,et al.,AMP-dependent protein kinase phosphorylation ofDrp1regulates its GTPase activity and mitochondrial morpho logy.J Biol Chem282,21583-21587(2007)。

Claims

1. An anticancer compound, which is a 20-O-β-(D-glucopyranosyl)-20(S)-protopanaxadiol (CK) derivative (CKD), said derivative having at least one glucose hydroxyl group substituted with at least one acetal or ketal group, wherein said acetal group provides a hydrophobic moiety, said hydrophobic moiety enhancing binding to the mitochondrial membrane protein PRELID3b, said mitochondrial membrane protein being capable of assembling a lipid transfer complex with TRIAP1 in mitochondria, and wherein said CKD has greater cytotoxicity than CK, wherein said CKD has the following structural formula: Where X is H or F, R1 is independently H, F, or methyl, and R2 is independently F or methyl; or Where R1 is H and R2 is -C6H x R 5-x , where x is 3 to 5, and R is independently selected from methoxy, nitro, trifluoromethyl and trifluoromethoxy.

2. The anticancer compound according to claim 1, wherein R1 is H and R2 is -C6H4OCH3, and wherein OCH3 is in the para, meta, or ortho position.

3. The anticancer compound according to claim 2, wherein R1 is H and R2 is p-methoxyphenyl.

4. The anticancer compound according to claim 1, wherein R1 is H, and R2 is phenyl, o-methoxyphenyl, m-methoxyphenyl, p-nitrophenyl, p-trifluoromethylphenyl, or p-trifluoromethoxyphenyl.

5. The anticancer compound according to claim 1, wherein the glucose hydroxyl group forms a ketal of 2,2-propane or a ketal of 2,2-difluoromethane.

6. An anticancer drug comprising at least one anticancer compound according to claim 1.

7. The anticancer drug according to claim 6, wherein R1 is H and R2 is -C6H. x X 5-x , where x is 3 to 5, and X is independently selected from methoxy, nitro, trifluoromethyl and trifluoromethoxy.

8. The anticancer drug according to claim 6, wherein R1 is H and R2 is -C6H4OCH3, and wherein OCH3 is in the para, meta, or ortho position.

9. The anticancer drug according to claim 7, wherein R1 is H and R2 is p-methoxyphenyl.

10. The anticancer drug according to claim 6, wherein R1 is H, and R2 is phenyl, o-methoxyphenyl, m-methoxyphenyl, p-nitrophenyl, p-trifluoromethylphenyl, or p-trifluoromethoxyphenyl.

11. A method for preparing the anticancer compound according to claim 1, comprising: Provides 20-O-β-(D-glucopyranosyl)-20(S)-protopanaxadiol (CK); Reagents containing aldehydes, acetals, ketones, or ketals are provided. The CK is combined with the reagent to form the CKD; and Separate the CKD.

12. The method of claim 11, wherein the reagent comprising a ketal or acetal is 2,2-dimethoxypropane or dimethoxydifluoromethane.

13. The method of claim 11, wherein the reagent comprising the acetal is 1-(dimethoxymethyl)-4-methoxybenzene.

14. The method according to claim 11, wherein the reagent comprising the acetal is dimethoxymethylbenzene, 1-(dimethoxymethyl)-3-methoxybenzene, 1-(dimethoxymethyl)-2-methoxybenzene, 1-(dimethoxymethyl)-4-nitrobenzene, 1-(dimethoxymethyl)-4-trifluoromethylbenzene, or 1-(dimethoxymethyl)-4-trifluoromethoxybenzene.

15. The method of claim 11, further comprising adding a solvent.

16. The method of claim 11, further comprising adding a catalyst, wherein the catalyst is added together with the CK and the reagent.

17. The method of claim 11, wherein separating the CKD comprises performing chromatography.

Images