Design and synthesis of bile acid derivatives and their activity against colon cancer

Abstract

Bile acids (BAs) containing both hydrophilic hydroxyl and carboxyl groups and hydrophobic methyl and steroid nuclei can promote the absorption of fat and other substances in the intestine, and they are synthesized by cholesterol in the liver and then returned to the liver through enteric liver circulation. Because there are many BA receptors on the cell membrane of colon tissues, BAs can improve the specific delivery and transport of medicines to colon tissues. Moreover, BAs have a certain anticancer and inflammation activity by themselves. Based on this theory, a series of BA derivatives against colon cancer including cholic acid (CA), chenodeoxycholic acid (CDCA), ursodeoxycholic acid (UDCA) and lithocholic acid (LCA) were designed and synthesized, and their antitumor activity was evaluated. For in vitro anti-tumor tests, all the compounds displayed cell proliferative inhibition to nine human malignant tumor cell lines to some degree, and in particular they showed stronger inhibition to the colon cancer cells than the other cell lines. Among them, four compounds (4, 5, 6, and 7) showed stronger activity than the other compounds as well as the positive control 5-FU against HCT116 cells, and their IC₅₀ was between 21.32 μmol L⁻¹ and 28.90 μmol L⁻¹; cell clone formation and migration tests showed that they not only effectively inhibited the formation of HCT116 cell colonies, but also inhibited the HCT116 cell migration and invasion; moreover, they induced apoptosis, arrested the mitotic process at the G2/M phase of the cell cycle, reduced the mitochondrial membrane potential, increased the intracellular ROS levels, and reduced the expression of Bcl-2 and p-STAT3 in HCT 116 cells. In addition, they also displayed intermediate anti-inflammatory activity by inhibiting inflammatory mediators NO and downregulating TNF-α expression, which also is one of the causes of colon cancer. This suggests that they deserve to be further investigated as candidates for colon cancer treatment drugs.

The BA derivatives designed in this subject showed good antitumor activity, especially with significant selectivity for the HCT116 cell line.

1. Introduction

Bile acids (BAs) are endogenous steroid compounds synthesized via cholesterol in the liver, and are mainly found in mammalian bile. BAs participate in various physiological processes, and they not only act as signaling molecules to regulate the cellular metabolism of glucose, lipids and energy, but also can participate in immune response.¹ Numerous studies have shown that BA derivatives have many biological functions, such as anti-tumor (Fig. 1),^2–8 anti-inflammatory,^9–11 antibacterial,^12–15 antiviral,^16,17 regulation of metabolism, and so on. Some BA derivatives (Fig. 2), such as obeticholic acid (6-ethylchenodeoxycholic acid, 6-ECDCA), have been used in clinical practice for the treatment of cholestatic liver diseases. The structures of some BA derivatives reported are listed in Fig. 1 and 2.

Fig. 1. The typical bile acid derivatives with antitumor activity reported.^2–8.

Fig. 2. The bile acid derivatives that have been used clinically.

BAs are a large class of sterol derivatives that contain hydrophilic hydroxyl and carboxyl groups, as well as hydrophobic methyl and steroid nuclei, which enable them to promote the absorption of fat and other substances in the gut. BAs are synthesized by cholesterol in the liver and returned to the liver through enteric liver circulation. This circulating property can be used as a carrier to design drugs for gastrointestinal diseases. As endogenous compounds, BAs have no toxicity to the human body, with improved drug bioavailability, tissue selectivity and site-specific delivery.¹⁸

There are many BA receptors in the colon tissue, so the compounds using BAs as carriers have dual function. They are not only easily distributed into colon tissues, but also beneficial for colon cancer treatment. Based on this theory, a series of BA derivatives, including cholic acid (CA), chenodeoxycholic acid (CDCA), ursodeoxycholic acid (UDCA), and lithocholic acid (LCA), were synthesized through combination of BAs and phenolic or benzyl methanol derivatives, and then evaluated for their biological activities on colon cancer cells. By which, we hope to provide a theoretical basis for the discovery of new anti-colon tumor drug candidates.

2. Results and discussion

2.1. Design, synthesis, and characterization of the compounds

There are many active small-molecule substances for diseases in nature. Caseol (4-hydroxyphenyl ethanol) from the natural rhotinine was reported to have pressure protection, anti-oxidant, anti-inflammation, anti-cancer, geriatric protection, cardio-protection, anti-atherosclerosis, neuroprotection and other pharmacological effects.^19–23 And 4-hydroxybenzyl methanol (HBA) is one of the main components of the traditional Chinese medicine Gastrodia elata, and many testing data have shown that HBA could inhibit inflammation by reducing cyclooxygenase activity and inhibiting the release of inflammatory mediators such as NO; moreover, HBA also inhibited tumor blood vessel formation and growth, making it a potential anticancer medicine.^24,25 2-Hydroxybenzamide (2-HOBA, also known as salicylamine) is a natural phenolic compound found in buckwheat, and it could remove a highly reactive lipid γ-keto-aldehyde (γ-KA) 980 times faster than a KA-forming protein adduct,²⁶ thereby protecting cells from the adverse effects of KA adducts and weakening the damaging effects of oxidative stress in cells or tissues.²⁷ 3,5-Dihydroxybenzyl alcohol isolated from Polygonum cuspidatum was reported to inhibit DNA topoisomerase I more strongly than camptothecin, and its inhibitory activity against DNA topoisomerase II was stronger than that of NSC-141540.²⁸ These small molecule compounds not only directly inhibit the proliferation of cancer cells but also indirectly their growth through the anti-inflammatory pathway. When these small molecules are combined with BAs by the ester bond, it is possible to obtain some drug molecules which have selectivity to colon cancer. Based on this theory, compound 1 was obtained by the reaction of CA and 2-hydroxybenzamide; compound 3 was afforded by esterification of CA and acetaminophen. Six BA derivatives containing one or two phenol hydroxyls (2, 9, 10, 17, 18 and 19) were obtained from the four BAs reacting with 3-hydroxyl benzyl alcohol and 3,5-dihydroxybenzyl alcohol. To further clarify the effect of the substituents on the benzene ring on the anticancer activity, CH₃, NO₂ and COOCH₃ were used to replace the hydroxyl on benzene, and some new BA derivatives were obtained; in addition, the introduction of methionine into CA and CDCA obtained compounds 8 and 16. The synthetic routes and structures of the compounds are shown in Scheme 1 and Fig. 3.

Scheme 1.

Fig. 3. The structures of the compounds.

All the compounds were prepared by a one-pot method. Compound 3 was obtained by esterification between CA and acetaminophen in the presence of EDCI and DAMP, and compounds 10 and 18 were obtained respectively by esterification of CDCA, LCA and 3,5-dihydroxybenzyl alcohol under the same conditions (reaction conditions b). Compounds 1, 8, and 16 were obtained by the condensation reaction catalyzed by catalyst HOBt and EDCI (reaction conditions c), and the remaining compounds were prepared with different bromobenzyl compounds by nucleophilic substitution reactions (conditions a).

Of the 20 compounds, compounds 1–8 are cholic acid (CA) derivatives, compounds 9–16 are chenodeoxycholic acid (CDCA) derivatives, compounds 17 and 18 are lithocholic acid (LCA) derivatives, and compounds 19 and 20 are ursodeoxycholic acid (UDCA) derivatives. Spectral data of the compounds are as expected. In the ¹H NMR, the chemical shift of two protons in PhCH₂O appeared at 4.88–5.24 ppm as singlets, and the signals of protons in benzene rings at 7.04–7.36 ppm appeared as multiplets. All the compounds were white solid, stable, soluble in organic solvents such as chloroform, toluene, DMF, and DMSO, and insoluble in water. Moreover, the preparation process of the compounds was safe and effective, and the reaction conditions were mild.

2.2. The proliferative inhibition of compounds against tumor cells

Using the MTT method, we first evaluated the proliferative inhibition of all the compounds against nine human malignant tumor cell lines (HCT 116, HT-29, RKO, LOVO, SW480, MGC-803, A549, HuH7, and SK) and 5-fluorouracil (5-FU) as the positive control. All the IC₅₀ values are shown in Table 1.

The IC₅₀ values of compounds for tumor cell lines^a.

Compd	IC50 (μmol L−1)
	HCT 116	HT-29	RKO	LOVO	SW480	HuH7	SK	A549	MGC-803
1	36.30 ± 1.95	32.72 ± 1.36	48.47 ± 1.36	51.08 ± 2.01	35.87 ± 1.82	63.66 ± 1.44	25.39 ± 0.89	40.47 ± 2.31	42.32 ± 1.78
2	33.01 ± 0.87	24.26 ± 0.98	47.58 ± 1.32	28.67 ± 1.04	39.26 ± 0.49	101.9 ± 3.54	61.83 ± 1.76	47.77 ± 1.01	54.87 ± 1.26
3	54.31 ± 2.24	53.26 ± 1.04	36.33 ± 1.04	58.02 ± 1.76	38.69 ± 0.70	70.37 ± 2.54	41.12 ± 1.28	59.40 ± 1.48	36.71 ± 1.67
4	21.32 ± 0.84	22.67 ± 1.21	27.62 ± 1.87	18.67 ± 0.56	23.61 ± 2.26	51.60 ± 2.32	60.21 ± 2.46	40.53 ± 1.07	21.91 ± 0.44
5	21.99 ± 0.58	34.78 ± 1.53	29.51 ± 1.58	24.39 ± 0.78	31.75 ± 2.18	62.99 ± 1.71	71.91 ± 2.74	45.70 ± 2.13	21.82 ± 0.35
6	23.66 ± 2.32	27.89 ± 0.77	31.69 ± 0.43	23.07 ± 1.61	28.03 ± 0.92	125.7 ± 2.86	50.66 ± 1.82	34.55 ± 1.36	26.42 ± 0.28
7	24.97 ± 0.73	27.15 ± 1.41	41.24 ± 2.21	28.45 ± 1.74	31.32 ± 5.77	52.10 ± 1.34	41.64 ± 1.27	51.15 ± 1.21	28.53 ± 1.23
8	85.16 ± 2.32	126.7 ± 1.74	97.29 ± 2.83	51.61 ± 1.38	64.65 ± 1.65	189.7 ± 4.31	173.3 ± 4.83	>200	96.88 ± 1.45
9	42.28 ± 1.23	33.61 ± 1.61	127.7 ± 4.19	40.90 ± 1.21	42.77 ± 2.36	124.1 ± 2.75	99.00 ± 2.21	69.39 ± 0.64	49.04 ± 1.78
10	64.48 ± 1.44	40.87 ± 1.54	76.88 ± 2.41	56.67 ± 2.03	69.37 ± 1.61	115.7 ± 3.46	156.3 ± 2.42	127.3 ± 2.31	81.78 ± 2.43
11	149.0 ± 1.88	56.31 ± 2.23	142.1 ± 3.20	76.92 ± 1.32	100.0 ± 3.53	60.72 ± 1.27	85.02 ± 2.21	>200	109.3 ± 4.65
12	53.72 ± 0.88	25.41 ± 1.12	42.05 ± 1.72	27.39 ± 0.87	43.06 ± 3.39	27.25 ± 1.02	54.87 ± 1.65	86.62 ± 1.69	37.53 ± 1.15
13	40.38 ± 1.19	29.57 ± 1.83	43.19 ± 1.56	30.37 ± 1.49	44.13 ± 0.83	44.62 ± 0.76	111.2 ± 2.34	87.22 ± 1.82	50.63 ± 1.19
14	44.97 ± 0.15	25.97 ± 2.21	45.40 ± 0.72	21.34 ± 1.63	37.61 ± 1.62	63.11 ± 1.74	65.08 ± 2.13	99.64 ± 2.21	36.58 ± 1.46
15	28.90 ± 1.36	29.17 ± 1.02	44.43 ± 1.32	28.27 ± 0.99	36.14 ± 0.76	74.76 ± 1.56	49.9 ± 2.23	120.9 ± 2.74	31.64 ± 0.72
16	27.43 ± 1.21	27.33 ± 1.89	27.32 ± 1.52	22.79 ± 1.56	23.57 ± 1.72	67.95 ± 0.93	38.56 ± 1.47	79.84 ± 2.32	31.53 ± 1.29
17	142.3 ± 1.99	173.0 ± 3.01	150.1 ± 3.44	157.5 ± 3.20	183.3 ± 2.28	185.0 ± 2.31	125.2 ± 2.19	>200	>200
18	50.28 ± 1.27	45.43 ± 2.26	42.87 ± 1.78	59.75 ± 1.38	134.2 ± 3.37	>200	>200	123.7 ± 3.28	82.29 ± 2.01
19	54.18 ± 2.21	58.81 ± 1.98	68.81 ± 1.64	44.89 ± 1.92	99.04 ± 2.15	152.5 ± 2.16	155.3 ± 3.62	>200	59.36 ± 2.12
20	102.0 ± 1.21	123.0 ± 2.19	187.2 ± 2.19	167.5 ± 4.14	>150	185.0 ± 2.31	177.0 ± 3.39	187.2 ± 2.75	111.0 ± 2.39
CA	>200	>200	>200	>200	>200	>200	>200	>200	>200
CDCA	>200	>200	>200	>200	>200	>200	>200	>200	>200
UDCA	>200	>200	>200	>200	>200	>200	>200	>200	>200
LCA	>200	>200	>200	>200	>200	>200	>200	>200	>200
5-FU	32.26 ± 0.88	17.48 ± 2.24	12.71 ± 0.69	12.60 ± 0.87	21.93 ± 1.61	1.06 ± 0.65	12.50 ± 0.88	108.7 ± 2.31	6.06 ± 1.08

^aFor compounds 4, 5, 6, 7, 15 and 16, their IC₅₀ values are over 500 μmol L⁻¹ to the WI38 cell line. The small molecules related to them displayed lower anticancer activity, and their IC₅₀ values were over 200 μmol L⁻¹.

The results showed that CA, CDCA, LCA and UDCA, the four parent compounds, showed very weak proliferative inhibitory activity for the nine human malignant tumor cell lines, and their IC₅₀ values were above 200 μmol L⁻¹; but all the synthesized compounds displayed strong proliferative inhibition, with most IC₅₀ values between 20 μmol L⁻¹ and 50 μmol L⁻¹. Generally, the compounds had good effects on five human colon cancer cell lines namely HCT 116, HT-29, RKO, LOVO, SW480, and one human gastric cancer cell line MGC-803, and lower activity for the human lung cancer cell line A549 as well as HuH7 and SK, two human liver cancer cell lines. This may be related to the selective uptake of the compounds by colon cancer cells.

In contrast, of the five colon cancer cell lines, all the compounds are highly selective to the colon cancer cell line HCT 116; among them, compounds 4, 5, 6, 7, 15 and 16 were more prominent and showed higher cell proliferative inhibition than the positive controls, with IC₅₀ values up to 21.32–28.9 μmol L⁻¹.

Most compounds displayed lower activity than the control 5-FU, and compounds of different skeletons exhibited distinct structure–activity relationships. For compounds 2, 9, 17 and 19, they are the derivatives obtained by the reaction of 3-hydroxybenzanol and CA, CDCA, LCA, and UDCA, respectively. For all the nine human malignant tumor cell lines, the order of compound activity from strong to weak according to the BA backbone was: CA > CDCA > UDCA > LCA. This indicates the more hydroxyl units in the BA backbone, the stronger the activity. For the CA derivatives 1–8, the substituents in the p-position of the benzene ring increased the anticancer activity (IC₅₀ 21.0 μmol L⁻¹) and those in the o- and m-positions decreased the activity; however, 3,5-dimethoxy substitution increased the anti-colon cancer activity (compound 6, IC₅₀ 23.66 μmol L⁻¹). For the CDCA derivatives 9–16, the carboxymethyl in the p-position of the benzene ring, or the compound with methyl methionine displayed good activity (compound 15, IC₅₀ 28.9 μmol L⁻¹ and 16, 27.43 μmol L⁻¹, respectively); compound 9 containing one phenolic hydroxyl was more active against colon cancer cells than compound 10 containing two phenolic hydroxyl; compound 14 containing methoxy is slightly better than compound 10 with phenolic hydroxyl. However, for LCA derivatives, compound 18 containing 3,5-dihydroxyl showed greater activity than compound 17 with 2-hydroxyl. In summary, we can see that the different BA skeletons have different structure–activity relationships, and even the results are the opposite due to the difference in the number of hydroxyls and substitution position on the benzene ring. And we can see that compounds 4, 5, 6, 7, 15 and 16 have no phenolic hydroxyls, but they showed strong anti-proliferation activity of colon cancer HCT116 cells. Possibly, the free hydroxyl group did not favor binding of molecules to the target of action of cancer cells.

2.3. Effect of compounds on the proliferative inhibition of HCT 116 cells

To illustrate the relationship between the concentration of the test compounds and the proliferative inhibition of HCT 116 cells, we treated HCT116 cells with different concentrations of compounds for 24 h, and then measured the rates of cell growth inhibition. The results are shown in Fig. 4. As can be seen from Fig. 4, for all six tested compounds, with increasing concentration, the growth inhibition rates of the cells gradually increased in a concentration-dependent manner. For compounds 4, 5 and 6, when they were less than 20 μmol L⁻¹, the rates of cell proliferative inhibition caused by them are less than 40%; when they were within 20–25 μmol L⁻¹, the inhibitory rate increased sharply with increasing concentration, indicating that the growth inhibition caused by them is accelerated; when it was higher than 40 μmol L⁻¹, the growth inhibition rate of cells had reached about 90%. As for compound 16, the cell inhibitory rate was less than 30% at 25 μmol L⁻¹, but over 70% at 30 μmol L⁻¹, indicating a fast increase in cell proliferative inhibition in the range of 25–30 μmol L⁻¹. For all the tested compounds, when it was less than 20 μmol L⁻¹ or higher than 40 μmol L⁻¹, the cytoinhibition rate did not differ much after the HCT116 cells were treated with compounds for 24 h, 48 h or 72 h. But when it was within 20–40 μmol L⁻¹, the longer the incubation time, the higher the rate of cellular inhibition. However, the difference between incubating for 48 h and 72 h becomes small, so in subsequent tests the incubating time was set to 48 h.

Fig. 4. Growth-inhibition rate curves of HCT116 cells after being treated with the compounds. Data from each experiment were obtained from at least three independent triplicate experiments, as indicated according to mean ± SD.

The colony formation assay is an important means to detect the proliferation of tumor cells in vitro. To further evaluate the anti-proliferation of the tested compounds, the colony formation assay was performed. The results are shown in Fig. 5. We can see that all the tested compounds significantly inhibited the number of clonal colonies and the volume size of HCT 116 cells compared with the control cells. By comparison, compounds 4, 5 and 7 greatly reduced the number of cell clonal colonies at the same concentrations, indicating that their proliferative inhibition is stronger than the others. This suggests that the compounds have ability to inhibit the proliferation of HCT 116 tumor cells. This is consistent with the conclusions before.

Fig. 5. HCT116 cell colony formation assay at 25 μmol L⁻¹ for 48 h.

2.4. Effect of compounds on HCT116 cell migration and invasion

Tumor cell metastasis is a major feature of malignant tumors; this metastasis involves cancer cells from the primary tumor spreading to the surrounding tissues and distant organs. This is also the main cause of cancer incidence and mortality. Therefore, to inhibit and prevent the cancer cells from migrating and invading is an effective means to treat cancers, which also is an effective way to prolong the life of patients. We wanted to know whether the compounds can inhibit the migration and invasion of HCT116 cells. In order to confirm our thoughts, we used the Transwell compartment model and conducted counting after crystal violet staining, and preliminarily evaluated the effects of the compounds on cell migration and invasion. The cells with migration and invasion are shown in Fig. 6. As can be seen from Fig. 6A and C, the cells of the control group were dense and stained deeply, and the number of migrating cells was lower. But for the cells treated with the compounds, the cells were sparse and stained shallowly, and the number of migrating cells was higher. Moreover, the number of migrating cells was different for all the tested compounds. This indicates that when they are 20 μmol L⁻¹, all the compounds were able to significantly inhibit HCT 116 cell migration. Of these, compounds 4, 5, 6 and 7 prevented about 80% of HCT116 migration and the migrating cells account for only about 20% of the total cells; but for compound 15, it inhibited cell migration by nearly 30%. The reason why the compounds inhibited the migration of cancer cells is possibly not only related to the drug mechanism of action, but also related to the speed of the drug onset; and drugs that readily enter cells are also more strongly able to inhibit their migration.

Fig. 6. Effect of compounds on HCT 116 cell migration and invasion. A) The effect of compounds on HCT116 cell migration; B) the effect of compounds on HCT116 cell invasion; C) the statistical analysis of HCT116 cell migration; D) the statistical analysis diagram of HCT116 cell invasion number; at 20 μmol L⁻¹ for 24 h; ruler 100 μm, magnification 20 fold; each value is represented by mean ± SD, **** p < 0.0001.

As shown in Fig. 6B and D, after treatment of HCT116 cells with the compounds at 20 μmol L⁻¹, the invasion and metastasis of HCT116 cells were significantly inhibited compared with the untreated group. Among all the compounds, compounds 4, 5, 6 and 7 had a strong effect, inhibiting nearly 80% of cell invasion and metastasis, whereas compounds 15 and 16 showed weak inhibition; in particular compound 16 only inhibited about 50% HCT116 cell invasion and metastasis. In summary, compounds 4, 5, 6, and 7 showed better inhibition in both cell migration and invasion and metastasis assays, indicating that they more readily cross the matrix membrane of tumor cells and thus enter the cells to function. Although the two trials differ from the specific microenvironment of the tumor, the results indicate that these four compounds do indeed have the ability to inhibit the invasion and metastasis of HCT116 cells and potentially to treat the disease.

2.5. Effect of compounds on HCT116 cell cycle and apoptosis

For normal cells, cell cycle progression is controlled by multiple factors. The different signaling cascades regulate cyclin, gene expression, post-translational modifications, etc. Meanwhile in tumors, the cells over-proliferate due to dysregulated cell cycle regulatory mechanisms. Based on the activity of compounds against HCT116 cells' proliferative inhibition, we speculated that this effect may be related to the blocking of the cell cycle. In order to confirm our ideas, HCT116 cells were treated with the compounds for 48 h and stained with propidium iodide (PI), and then we determined the effect of tested compounds on the cell cycle.

The results showed the HCT116 cell cycle had a great change after the HCT116 cells were treated with 25 μmol L⁻¹ compounds for 48 hours; most of the cells stagnated in the G2M phase, in which DNA increased substantially and it did not change almost in other periods. Fig. 7 shows that the cells at the G2M phase in the control group accounted for only 11.00%, but after the HCT116 cells were treated with compounds 4 and 5, the cells at the G2M phase reached 90.66% and 93.18%, respectively; for compounds 15 and 16, they were 63.96% and 72.91%, respectively. This indicates that the compounds arrested the mitotic process in the G2M phase. Therefore, we can draw a conclusion that blocking the cell cycle at the G2M phase is also one of the important reasons for the compounds to inhibit cell proliferation.

Fig. 7. Effect of the compounds at 25 μmol L⁻¹ on the HCT116 cell cycle.

In addition to the cell cycle, the induction of apoptosis is also one of the common mechanisms of antitumor drugs. FITC-labeled annexin-V (annexin-V) can be used to distinguish between normal viable cells and early apoptotic cells, and propidium iodide (PI) can distinguish between surviving early cells, late apoptotic cells, and necrotic cells. To evaluate the effect of the compounds on HCT 116 apoptosis, HCT116 cells were incubated with different concentrations of compounds for 48 h, and 5-FU was used as a positive control; then the cells were double stained with Annexin V-FITC/PI, and then detected by flow cytometry. The results are shown in Fig. 8. We can see that: for 5-FU at 30 μmol L⁻¹, the apoptotic cells were about 71.51%, and the late apoptotic cells account for 69.38% and 2.13% in early apoptosis. But for compounds 4, 5, 6 and 7 under the same concentration, they induced a total apoptosis of 73.91%, 85.2%, 78.89% and 75.73%, respectively. Apparently, these four compounds displayed stronger activity to induce HCT116 apoptosis than 5-FU; as for compounds 15 and 16, the apoptotic cells account for 64.25% and 56.21%, respectively, indicating that their activity is lower than that of 5-FU. The results show that all the tested compounds induced HCT116 cell programmed apoptosis, and they are mainly dominated by late apoptosis, and almost no cells are in early apoptosis. Among the six compounds, compound 5 displayed the highest activity, and it increased cell apoptosis by 75.38% compared with the control. Therefore, it can be inferred that the compounds that block the cell cycle and induce apoptosis are important reasons for HCT 116 cell proliferation inhibition.

Fig. 8. Effect of the compounds on apoptosis in HCT 116 cells. 5-FU was 30 μmol L⁻¹, and all the compound were 30 μmol L⁻¹. The first row is the control group, compounds 4, 5 and 6, and the second row is 5-FU, compounds 7, 15 and 16.

To more clearly present the apoptotic process induced by the compounds, we dyed the HCT116 cells by Giemsa staining and DAPI staining, and observed the HCT116 cells' morphology and state by microscopy (Fig. 9). From the Giemsa staining results, the morphology of HCT116 cells treated with all the compounds showed contraction of cell volume, unclear cell rupture edges, apoptotic body formation, and retraction of the nucleus and cytoplasm. DAPI is a common nuclear and chromosomal counteragent that emits blue fluorescence upon binding to the DNA region, and it is semi-permeable and selective. By DAPI staining, the nuclei of HCT116 cells were fixed and appeared bright blue, indicating that the apoptotic cells induced by the tested compound appeared. The weaker the bright blue light, the greater the number of apoptotic cells, which further indicates a stronger compound activity. It is clear that the cells treated with compounds 4, 5, 6 and 7 have fewer bright blue cell clusters than the 5-FU group, while compounds 15 and 16 cause more bright blue nuclear clusters. It suggests that the former has stronger activity to induce HCT116 cell apoptosis than 5-FU, while the latter has lower activity. This is in accord with that reported before.

Fig. 9. A shows HCT116 cells stained with Jiemsa dye; B shows DAPI fluorescence stained HCT116 cells. 5-FU and compounds were 30 μmol L⁻¹; scale = 100μm, magnification: 10 fold.

2.6. Effect of compounds on expression of biomarkers in HCT116 cells

There are many biomarkers involved in growth, proliferation and apoptosis of cells. The anti-apoptotic protein Bcl-2 and its family members play an important role in regulating cell apoptosis and making the tumor cells resistant to numerous apoptotic stimuli.^29,30 The anti-proliferation of compounds is related to anti-apoptotic protein Bcl-2. The western blot results show that the Bcl-2 level in HCT116 cells treated with the compounds for 48 h was significantly downregulated (Fig. 10). Among them, compound 5 downregulated Bcl-2 more significantly than the other compounds. Caspases are a family of endogenous proteases, and they are the key mediators of programmed cell apoptosis. Among them, caspase-3 is a death protease which is commonly activated, and it can catalyze the specific cleavage of many key cellular proteins. Caspase-3 is originally produced in the form of inactive monomer, and requires dimerization and cleavage for activation; however, the cleavage of caspase 3 can be detected by protein blotting.³¹ In order to further determine the effect of compounds on caspase 3, the cleaved caspase 3 levels of the HCT116 cells treated with the compounds were determined by immunoblot tests (Fig. 10). The results showed that all the compounds upregulated cleaved caspase 3. In comparison, compounds 7, 15 and 16 upregulated cleaved caspase 3 much more strongly than the control, while compounds 4 and 5 were not much different from the control. This suggests that the apoptosis induced by the compounds may be achieved by downregulating Bcl-2 and upregulating the cleaved caspase 3 expression level, but the activity order of the tested compounds is not consistent with their activity order of blocking the cell cycle and promoting apoptosis.

Fig. 10. Changes in apoptosis-related proteins after HCT116 cells were treated with 30 μmol L⁻¹ compounds for 48 h. For n = 3, each value is represented by mean ± SD, * p < 0.05, ** p < 0.01, and **** p < 0.0001.

Signal transduction and transcription activator 3 (STAT3) is a member of the family of transcription factors that regulate cell proliferation, cell transformation, tumor formation, and immune response.³² In normal cells, the expression level of STAT3 is properly controlled. But in many cancers, STAT3 is hyperpersistently activated, and STAT3 is often overexpressed in tumor cells and tissues. This aberrant activation provides favorable conditions for tumor metastasis as well as tumor cell proliferation, angiogenesis, migration, and invasion.³³ In the majority of malignant tumor cells, STAT3 keeps on activating resulting in the phosphorylation of STAT3. In addition, high activation of STAT3 can drive tumorigenesis by transcription promoting angiogenesis inducers, which encode cell cycle regulators (e.g., cyclin D1 and c-Myc), apoptosis inhibitors (e.g., Bcl-xL, Mcl-1 and survivin) and vascular endothelial growth factor (VEGF).³⁴ In order to further determine the effect of compounds on STAT3, the STAT3 and p-STAT3 levels of the HCT116 cells treated with the compounds were determined by immunoblot tests. The results are shown in Fig. 11. We can see that all the compounds downregulated the expression of p-STAT3 to some degree, and their activities were stronger than that of the control drug. This indicates that the inhibition of compounds on STAT3 signaling may also be one of the reasons for their anti-HCT116 cell proliferation.

Fig. 11. Expression levels of p-STAT3 and STAT3 after HCT116 cells were treated with 30 μmol L⁻¹ compounds, respectively. For n = 3, each value is represented by mean ± SD, **** p < 0.0001.

2.7. Effect of compounds on mitochondrial membrane potential and ROS in HCT116 cells

Mitochondria are important organelles in eukaryotic cells that not only integrate death signals, but also provide energy to maintain the cellular metabolic demands.³⁵ Mitochondrial membrane potential (Δψ_m) is highly associated with cancer, and the cancer-derived cells (e.g., breast, prostate, melanoma, etc.) have higher Δψ_m, at least 60 mv, than normal epithelial cells.^36,37 To confirm whether the antitumor activity of the compounds was related to the mitochondrial membrane potential, after the HCT116 cells were treated with the compounds, we stained them with tetramethylrhodamine ethyl ester (TMRE) dye, in which the transmembrane distribution of the dye is directly correlated to the membrane potential. Then the cell mitochondrial membrane potentials were measured by flow cytometry. The results are shown in Fig. 12. We can see that all the tested compounds reduced the mitochondrial membrane potential in HCT116 cells (Fig. 12A and C). Among the tested compounds, compounds 4 and 5 displayed stronger activity than the control, and reduced the mitochondrial membrane potential to 50–55% of the control, while the other compounds decreased the potential to approximately 70–80% of the control group.

Fig. 12. Effect of compounds on mitochondrial membrane potential changes (Δψ_m) and reactive oxygen species (ROS) after treatment of HCT 116 cells. A: Change of compounds on mitochondrial membrane potential for 48 h; B: ROS after compound treatment for 24 h; C: relative change of cell mitochondrial membrane potential; D: relative change of cell ROS; n = 3, each value is represented by mean ± SD, **** p < 0.0001.

In normal metabolism, the cells produce reactive oxygen species (ROS), and most ROS are produced in cells through the mitochondrial respiratory chain. The ROS is an important signaling molecule in every stage of cancer development, including initiation, promotion, and progression. High levels of ROS can cause mitochondrial dysfunction, further giving rise to cell apoptosis. In order to confirm whether the anti-proliferation activity is related to the ROS level, we used DCFH-DA dye to detect the intracellular changes in ROS. The results show all the tested compounds significantly increased the ROS levels in HCT116 cells, and it increased by 20–25% compared with the control (Fig. 12B and D). Therefore, we draw a conclusion: the induction of apoptosis by reducing the mitochondrial membrane potential and elevating the intracellular ROS levels is also one of the effects of the compounds against colon cancer.

2.8. Anti-inflammatory activity of the compounds

The occurrence of colon cancer has great relevance with colon inflammation. BAs have anti-inflammatory activities, especially for gut-associated inflammation.^38,39 In the innate immune system, BAs exert an anti-inflammatory effect by suppressing NF-κB signaling pathways and NLRP3-dependent inflammasome activities.^40–42 To evaluate the anti-inflammatory activity of the compounds, we chose RAW 264.7 cells stimulated with LPS (1 μg mL⁻¹) as a model, then treated them with the compounds for 24 h. After that, the inflammatory factor NO levels of the cells were measured.

The results showed all 20 compounds displayed activity to inhibit NO production to some degree. Compounds 10 and 18 showed the highest activity in inhibiting NO release among them, with IC₅₀ of 16.3 μM and 12.1 μM; compounds 4, 5, 6 and 7 displayed intermediate activity, with IC₅₀ of 20.6–25.4 μM. As can be seen in Fig. 13, when compound 10 was 10 μmol L⁻¹ and 25 μmol L⁻¹, the NO inhibitory rates of RAW 264.7 cells were 35.46% and 73.19% of the LPS control, respectively; as for compound 18 under the same conditions, they reached 48.99% and 82.09%, respectively. With the concentration of the compound increasing, the stronger the anti-inflammatory activity. Distinctly, this effect is concentration-dependent.

Fig. 13. The NO-inhibition rate of the RAW 264.7 cells stimulated with LPS (1 μg ml⁻¹) after being treated with compounds 10 and 18 for 24 h. Take the LPS group as 100%.

To further investigate the anti-inflammatory activity of the compounds, the TNF-α level of the cells was also determined. The results show that all the tested compounds downregulated the TNF-α level, and this effect is parallel with their inhibition of NO release. In the LPS-induced cells, the TNF-α level rose sharply as compared to that of the blank control, and it was about 1800 pg ml⁻¹. After the RAW 264.7 cells were treated with the compounds, the TNF-α level was reduced in a concentration-dependent manner. Compound 18 displayed the highest activity among all the tested compounds. When compound 18 was 25 μmol L⁻¹, the TNF-α level in RAW 264.7 cells was reduced to 50% of the control (Fig. 14). In contrast, compounds 4, 5, 6 and 7 showed slightly lower activity to downregulate the TNF-α level.

Fig. 14. Effect of compounds acting on the RAW264.7 cells stimulated with LPS for 24 h when on TNF-α. ### represents the significant difference between the LPS and blank groups, p < 0.001; * represents a significant difference between the compound groups and the LPS groups, with p < 0.05, ** indicates p < 0.01, and *** indicates p < 0.001.

Compared with compound 18, compound 10 showed lower anti-inflammatory activity. This seemly suggests that the presence of more hydroxyl units in the BA framework is disadvantageous to anti-inflammation. Generally, the hydroxyl group on the benzene ring is beneficial against inflammation, but not against cancer. Compounds 4, 5, 6, 7, 15 and 16, though they displayed lower anti-inflammatory activity, showed strong anti-proliferation activity of colon cancer HCT116 cells. It is possible that the ability of the hydroxyl group to form hydrogen bonds prevents the binding of molecules to the target of action of cancer cells.

3. Conclusion

BAs have the amphipathic molecular properties to promote the absorption of fat and other substances in the intestine. Active small molecules were combined with them to obtain a series of BA derivatives. Because there are many bile acid receptors in the colon tissues and cells, the derivatives maybe are absorbed selectively by colon tissues and cells, which facilitates the treatment of colon cancer.

Our testing results showed that all the compounds showed good selectivity against 5 human colon cancer cell lines among the nine cancer cell lines, especially showing high selectivity for HCT116 colon cancer cells. Among all the compounds, compounds 4, 5, 6 and 7 exhibited high anticancer activity, and their activity was stronger than that of the positive control 5-FU. Their effects were concentration-dependent; the longer the action time, the better the action effect. Cell cloning tests show that the four compounds effectively inhibited HCT116 cell colony formation and reduced the colony size; moreover, they also effectively inhibited HCT116 cell migration and invasion. Pharmacological studies found that they not only affected the cycle of HCT 116 cell division and arrested the cells at the G2/M phase, but also induced the apoptosis of HCT116 cells. Meanwhile, the four compounds also reduced the mitochondrial membrane potential of HCT116 cells, with increasing intracellular ROS levels, downregulating the anti-apoptotic protein Bcl-2, upregulating the expression level of cleaved caspase 3 and inhibiting the phosphorylation of STAT3. In addition, in the LPS (1 μg mL⁻¹)-induced macrophage RAW 264.7 as inflammatory model tests, the four compounds showed intermediate inhibitory activity for NO release, as well as downregulating TNF-α expression. This indicates that they have an anti-inflammation activity to some degree.

On the whole, the BA derivatives designed in this subject showed good antitumor activity, especially with significant selectivity for the HCT116 cell line. For 4, 5, 6 and 7, they not only inhibited colony formation, migration and invasion, but also induced apoptosis, downregulating Bcl-2, upregulating cleaved caspase 3 and inhibiting phosphorylation of STAT3. And they have intermediate anti-inflammation activity (Fig. 15), which is one of causes of colon cancer; furthermore, they have a targeted selectivity. Therefore, they deserve to be further investigated as drug candidates to cure colon cancer.

Fig. 15. The mechanism of action of the compounds.

4. Experimental

4.1. Synthesis of the compounds

Synthesis of some intermediates

3-Hydroxyl benzyl bromine

In a nitrogen atmosphere, the chloroform solution of 6 mmol phosphorus tribromide was dropped into a chloroform suspension containing 3 mmol m-hydroxyl benzyl alcohol within 30 minutes in an ice bath. Then, the mixture was stirred at room temperature, and monitored by TLC until the reaction was complete (about 2 h). The resulting gold solution was poured into a pre-cooled saturated sodium chloride solution. The organic phase was separated, the aqueous phase was extracted three times with chloroform, and the combined chloroform phase was dried with anhydrous sodium sulfate and evaporated to dry. The resulting brown oil was purified by silica gel column chromatography using dichloromethane as the eluate (R_f 0.55) to produce a white solid with a yield of 83.40%.

3,5-Dihydroxyl-benzylbromine

3,5-dihydroxybenzyl methanol was used as the starting material, THF as the solvent, PE/EA as the eluate, and the rest was similar to the preparation process of m-hydroxyl benzyl bromine. It is a white solid, with a yield of 78.93%.

4-Ethoxyl benzyl methanol

In N₂ atmosphere, 10 mmol 4-acetyloxybenz-aldehyde was dissolved in THF in an ice bath, a small amount of NaBH₄ was added slowly to the reaction bottle, and the mixture was removed from the ice water bath when there are no more bubbles. The reaction continued at room temperature, with TLC monitoring until the reaction was complete (about 1.5 h); with saturated NH₄Cl aqueous solution quenching (carefully add slowly), a lot of EA was added to the reaction mixture, which was washed several times with saturated salt water. The organic phase was combined and dried over anhydrous sodium sulfate, silica gel column chromatography separated using PE/EA as elute. The product is a colorless oil liquid, with a yield of 70.45%.

General method of compound synthesis

2 mmol (1 eq.) CA/CDCA/UDCA/LCA was added in a round bottom flask in N₂ atmosphere, and 5 ml of DMF, different small molecule compounds (1.1 eq.) and cesium carbonate (1.2 eq.) or other reagents were added. TLC was used to monitor the reaction until the reaction was complete. After filtration, ethyl acetate was added, DMF was removed with a small amount of saturated salt water, anhydrous sodium sulfate was used for drying, the solvent was drained with a rotary evaporator, and silica column chromatography of the petroleum ether and ethyl acetate (PE/EA) system was used to separate the target compound.

Compound 1

White solid and yield 47.2%. ESI-MS: calcd for C₃₁H₄₇NO₅ [M–H]⁻ 512.3376, found 512.3805. ¹H NMR (400 MHz, methanol-d₄) δ 7.07 (dd, J = 21.9, 7.7 Hz, 2H, Ar–H), 6.75 (d, J = 8.6 Hz, 2H, Ar–H), 5.43 (s, 1H, NH), 4.28 (s, 2H, Ph–CH₂), 3.89 (s, 1H, 3-CH), 3.75 (s, 1H, 7-CH), 3.28 (s, 1H, 12-CH), 0.97 (d, J = 5.7 Hz, 3H, 21-CH₃), 0.87 (s, 3H, 18-CH₃), 0.61 (s, 3H, 19-CH₃). ¹³C NMR (101 MHz, methanol-d₄) δ 177.2, 156.5, 130.5, 129.7, 125.9, 120.5, 116.6, 73.9, 72.8, 69.0, 47.9, 47.4, 43.1, 42.9, 40.9, 40.4, 39.9, 36.7, 36.5, 35.8, 33.9, 33.3, 31.1, 29.5, 28.6, 27.8, 24.2, 23.2, 17.7, 12.9.

Compound 2

White solid and yield 54.4%. ESI-MS: calcd for C₃₁H₄₆O₆ [M + Na]⁺ 537.3187, found 537.3201. ¹H NMR (400 MHz, DMSO-d₆) δ 9.43 (s, 1H, Ph–OH), 7.13 (t, J = 7.9 Hz, 1H), 6.77–6.67 (m, 3H), 4.98 (s, 2H, Ph–CH₂), 3.97 (s, 3H, OH), 3.78 (br, 1H, 12-CH), 3.61 (br, 1H, 7-CH), 3.19 (m, J = 10.9, 5.9 Hz, 1H, 3-CH), 0.92 (d, J = 5.7 Hz, 3H, 21-CH₃), 0.80 (s, 3H, 19-CH₃), 0.56 (s, 3H, 18-CH₃). ¹³C NMR (101 MHz, DMSO-d₆) δ 173.6, 157.9, 138.1, 129.8, 118.7, 115.3, 115.1, 71.5, 70.9, 66.8, 65.7, 46.5, 46.2, 42.0, 41.8, 35.8, 35.5, 35.3, 34.8, 31.2, 30.9, 28.9, 27.7, 26.7, 23.3, 23.1, 17.3, 12.7.

Compound 3

White solid and yield 42.4%. ESI-MS: calcd for C₃₂H₄₇NO₆ [M + H]⁺ 542.3476, found 542.3510. ¹H NMR (400 MHz, methanol-d₄) δ 7.55 (d, J = 8.9 Hz, 2H, Ar–H), 7.01 (d, J = 8.9 Hz, 2H, Ar–H), 3.96 (d, J = 2.9 Hz, 1H, 12-CH), 3.79 (d, J = 3.0 Hz, 1H, 7-CH), 3.37 (m,1H, 3-CH), 2.11 (s, 3H), 1.06 (d, J = 5.8 Hz, 3H, 21-CH₃), 0.91 (s, 3H, 18-CH₃), 0.72 (s, 3H, 19-CH₃). ¹³C NMR (101 MHz, methanol-d₄) δ 174.3, 171.3, 148.1, 137.3, 122.7, 121.7, 73.8, 72.6, 68.8, 47.8, 47.3, 42.9, 42.8, 40.8, 40.2, 36.5, 36.3, 35.7, 35.6, 31.9, 31.9, 30.9, 29.4, 28.5, 27.7, 24.0, 23.6, 23.0, 17.5, 12.8.

Compound 4

White solid and yield 70.3%. ESI-MS: calcd for C₃₂H₄₈O₅ [M + Na]⁺ 535.3394, found 535.3363; ¹H NMR (400 MHz, CDCl₃) δ 7.23 (d, J = 7.8 Hz, 2H, Ar–H), 7.14 (d, J = 7.8 Hz, 2H, Ar–H), 5.05 (s, 2H, Ph–CH₂), 3.92 (br, 1H, 12-CH), 3.82 (br, 1H, 7-CH), 3.40 (m, J = 11.2, 6.2 Hz, 1H, 3-CH), 2.33 (s, 3H, Ph–CH₃), 0.96 (d, J = 5.3 Hz, 3H, 21-CH₃), 0.86 (s, 3H, 18-CH₃), 0.63 (s, 3H, 19-CH₃); ¹³C NMR (101 MHz, CDCl₃) δ 173.9, 137.7, 132.9, 128.9, 128.1, 72.8, 71.6, 68.2, 65.8, 46.7, 46.1, 41.3, 39.2, 39.1, 35.2, 35.0, 34.5, 34.4, 31.2, 30.7, 29.9, 27.9, 27.3, 26.0, 23.0, 22.2, 20.9, 17.1, 12.2.

Compound 5

White solid and yield 70.3%. ESI-MS: calcd for C₃₁H₄₅NO₇ [M + Na]⁺ 566.3077, found 566.3097. ¹H NMR (400 MHz, methanol-d₄) δ 8.24 (d, J = 8.8 Hz, 2H, Ar–H), 7.61 (d, J = 8.8 Hz, 2H, Ar–H), 5.24 (s, 2H, Ph–CH₂), 3.93 (br, 1H, 12-CH), 3.79 (br, 1H, 7-CH), 3.37 (m, J = 3.5 Hz, 1H, 3-CH), 1.01 (d, J = 5.8 Hz, 3H, 21-CH₃), 0.91 (s, 3H, 18-CH₃), 0.66 (s, 3H, 19-CH₃). ¹³C NMR (101 MHz, CDCl₃) δ 173.7, 147.4, 143.3, 128.2, 123.6, 72.9, 71.6, 68.3, 64.4, 46.7, 46.2, 41.4, 41.3, 39.2, 35.2, 35.1, 34.6, 34.5, 30.9, 30.7, 30.0, 27.9, 27.4, 26.1, 23.0, 22.2, 17.1, 12.2.

Compound 6

White solid and yield 83.3%. ESI-MS: calcd for C₃₃H₅₀O₇ [M + Na]⁺ 581.3449, found 581.3460. ¹H NMR (400 MHz, CDCl₃) δ 6.46 (d, J = 2.3 Hz, 2H, Ar–H), 6.37 (s, 1H, Ar–H), 5.00 (s, 2H, Ph–CH₂), 3.90 (br, 1H, 12-CH), 3.81 (br, 1H, 7-CH), 3.75 (s, 6H, OCH₃), 3.41–3.34 (m, 1H, 3-CH), 0.95 (d, J = 5.3 Hz, 3H, 21-CH₃), 0.83 (s, 3H, 18-CH₃), 0.61 (s, 3H, 19-CH₃). ¹³C NMR (101 MHz, CDCl₃) δ 173.8, 160.6, 138.1, 105.6, 99.7, 72.8, 71.5, 68.1, 65.7, 55.0, 46.7, 46.0, 41.2, 39.2, 39.1, 35.1, 35.0, 34.5, 34.4, 31.0, 30.6, 29.9, 27.8, 27.3, 25.9, 22.9, 22.2, 17.0, 12.1.

Compound 7

White solid and yield 76.4%. ESI-MS: calcd for C₃₃H₄₈O₇ [M + Na]⁺ 579.3292, found 579.3307; ¹H NMR (400 MHz, CDCl₃) δ 7.96 (d, J = 8.0 Hz, 2H, Ar–H), 7.34 (d, J = 8.0 Hz, 2H, Ar–H), 5.09 (s, 2H, Ph–CH₂), 3.87 (br, 1H, 12-CH), 3.84 (s, 3H, CH₃), 3.75 (br, 1H, 7-CH), 3.35 (m, J = 11.4, 6.0 Hz, 1H, 3-CH), 0.92 (d, J = 5.3 Hz, 3H, 21-CH₃), 0.80 (s, 3H, 18-CH₃), 0.57 (s, 3H, 19-CH₃). ¹³C NMR (101 MHz, CDCl₃) δ 173.8, 166.6, 141.0, 129.6, 127.5, 72.9, 71.6, 68.2, 65.1, 51.9, 46.7, 46.2, 41.3, 39.3, 35.2, 35.1, 34.6, 34.5, 31.0, 30.7, 30.0, 27.9, 27.3, 26.1, 23.0, 22.2, 17.1, 12.2.

Compound 8

White solid and yield 64.7%. ESI-MS: calcd for C₃₀H₅₁NO₆S [M + Na]⁺ 576.3329, found 576.3305. ¹H NMR (400 MHz, CDCl₃) δ 6.62 (d, J = 7.9 Hz, 1H, NH), 4.72–4.60 (m, 1H, –NHCH̲–), 3.92 (s, 1H, 12-CH), 3.79 (d, J = 3.6 Hz, 1H, 7-CH), 3.71 (s, 3H, CH₃COO), 3.40–3.35 (m, 1H, 3-CH), 2.48 (t, J = 7.5 Hz, 2H, –CH̲₂SCH₃), 2.06 (s, 3H, SCH₃), 0.96 (d, J = 5.4 Hz, 3H, 21-CH₃), 0.84 (s, 3H, 18-CH₃), 0.63 (s, 3H, 19-CH₃). ¹³C NMR (101 MHz, CDCl₃) δ 173.7, 172.6, 72.9, 71.7, 68.3, 52.3, 51.2, 46.4, 46.2, 41.4, 41.3, 39.3, 39.3, 35.1, 34.6, 34.6, 32.8, 31.5, 31.3, 30.3, 29.9, 27.9, 27.4, 26.1, 23.1, 22.3, 17.3, 15.3, 12.3.

Compound 9

White solid and yield 54.2%. ESI-MS: calcd for C₃₁H₄₆O₅ [M + Na]⁺ 521.3237, found 521.3166. ¹H NMR (400 MHz, DMSO-d₆) δ 9.43 (s, 1H, Ph–OH), 7.13 (t, J = 7.9 Hz, 1H, Ar–H), 6.77–6.67 (m, 3H, Ar–H), 4.98 (s, 2H, Ph–CH₂), 4.32 (d, J = 4.7 Hz, 1H, 20-CH), 4.09 (d, J = 3.3 Hz, 1H, 7-CH), 3.62 (s, 1H, 7-OH), 3.36 (s, 1H, 3-OH), 3.19 (m, J = 10.7, 5.4 Hz, 1H, 3-CH), 0.87 (d, J = 6.5 Hz, 3H, 21-CH₃), 0.83 (s, 3H, 18-CH₃), 0.58 (s, 3H, 19-CH₃). ¹³C NMR (101 MHz, DMSO-d₆) δ 173.5, 157.8, 138.0, 129.8, 118.8, 115.3, 115.1, 70.8, 66.6, 65.7, 55.9, 50.4, 42.4, 41.9, 35.8, 35.3, 35.2, 31.1, 31.1, 31.0, 28.2, 23.6, 23.2, 20.7, 18.5, 12.0.

Compound 10

White solid and yield 53.3%. ESI-MS: calcd for C₃₁H₄₆O₅ [M + Na]⁺ 537.3187, found 537.3288. ¹H NMR (400 MHz, DMSO-d₆) δ 9.25 (s, 2H, Ph–OH), 6.18 (s, 2H, Ar–H), 6.15 (s, 1H, Ar–H), 4.88 (s, 2H, Ph–CH₂), 4.32 (d, J = 4.7 Hz, 1H, 7-CH), 4.10 (d, J = 3.3 Hz, 1H, 3-CH), 3.63 (s, 1H, 3-OH), 3.18 (m, J = 5.4 Hz, 1H, 7-OH), 0.88 (d, J = 6.6 Hz, 3H, 21-CH₃), 0.84 (s, 3H, 18-CH₃), 0.59 (s, 3H, 19-CH₃). ¹³C NMR (101 MHz, DMSO-d₆) δ 173.5, 158.9, 138.6, 106.2, 102.4, 70.8, 66.7, 65.7, 55.9, 50.5, 42.4, 41.9, 35.8, 35.3, 35.2, 32.8, 31.1, 28.3, 23.6, 23.2, 20.7, 18.6, 12.1.

Compound 11

White solid and yield 53.3%. ESI-MS: calcd for C₃₃H₄₈O₆ 541.3529, found 541.3533; [M + Na]⁺ 563.3349, found 563.3342. ¹H NMR (400 MHz, methanol-d₄) δ 7.34 (d, J = 8.2 Hz, 2H, Ar–H), 7.04 (d, J = 8.5 Hz, 2H, Ar–H), 5.05 (s, 2H, Ph–CH₂), 3.74 (d, J = 2.6 Hz, 1H, 7-CH), 3.37–3.29 (m, 1H, 3-CH), 2.22 (s, 3H, CH̲₃COO), 0.88 (d, J = 7.4 Hz, 6H, 21-CH₃, 18-CH₃), 0.60 (s, 3H, 19-CH₃). ¹³C NMR (101 MHz, methanol-d₄) δ 174.1, 169.6, 150.7, 133.9, 129.1, 121.5, 71.5, 67.6, 65.1, 55.9, 50.1, 42.3, 41.8, 39.6, 39.4, 39.1, 35.3, 35.2, 34.8, 34.5, 32.6, 30.9, 30.7, 29.9, 27.8, 23.2, 22.1, 20.4, 19.6, 17.4, 10.6.

Compound 12

White solid and yield 64.6%. ESI-MS: calcd for C₃₂H₄₈O₄ [M + Na]⁺ 519.3445, found 519.3373. ¹H NMR (400 MHz, methanol-d₄) δ 7.23 (d, J = 7.9 Hz, 2H, Ar–H), 7.15 (d, J = 7.8 Hz, 2H, Ar–H), 5.04 (s, 2H, Ph–CH₂), 3.78 (br, J = 2.9 Hz, 1H, 7-CH), 3.36 (m, J = 4.4 Hz, 1H, 3-CH), 2.32 (s, 3H, Ph–CH₃), 0.92 (d, J = 4.7 Hz, 6H, 21-CH₃, 13-CH₃), 0.62 (s, 3H, 19-CH₃). ¹³C NMR (101 MHz, CDCl₃) δ 173.9, 137.7, 132.9, 128.9, 128.2, 71.7, 68.2, 65.8, 55.7, 50.2, 42.4, 41.3, 39.5, 39.2, 35.2, 35.1, 34.8, 34.5, 32.6, 31.0, 30.8, 30.4, 27.9, 23.4, 22.6, 21.0, 20.4, 18.1, 11.5.

Compound 13

White solid and yield 54.5%. ESI-MS: calcd for C₃₁H₄₅NO₆ [M + Na]⁺ 550.3139, found 550.3147. ¹H NMR (400 MHz, CDCl₃) δ 8.19 (d, J = 8.7 Hz, 2H, Ar–H), 7.49 (d, J = 8.6 Hz, 2H, Ar–H), 5.18 (s, 2H, Ph–CH₂), 3.82 (m, J = 2.9 Hz, 1H, 7-CH), 3.43 (m, J = 11.1, 5.5 Hz, 1H, 3-CH), 0.90 (d, J = 6.2 Hz, 3H, 21-CH₃), 0.87 (s, 3H, 18-CH₃), 0.61 (s, 3H, 19-CH₃). ¹³C NMR (101 MHz, CDCl₃) δ 173.5, 128.1, 123.5, 71.5, 68.0, 64.3, 55.5, 50.1, 39.4, 39.4, 39.1, 35.1, 34.9, 34.8, 34.4, 32.6, 30.7, 30.6, 30.3, 27.9, 23.3, 20.3, 18.0, 11.5.

Compound 14

White solid and yield 52.8%. ESI-MS: calcd for C₃₃H₅₀O₆ [M + H]⁺ 543.3680, found 543.3653, [M + Na]⁺ 565.3500, found 565.3512. ¹H NMR (400 MHz, CDCl₃) δ 6.47 (d, J = 2.3 Hz, 2H, Ar–H), 6.39 (s, 1H, Ar–H), 5.02 (s, 2H, Ph–CH₂), 3.82–3.79 (m, 1H, 7-CH), 3.76 (s, 6H, OCH₃), 3.43 (d, J = 6.6 Hz, 1H, 3-CH), 0.90 (d, J = 6.2 Hz, 3H, 21-CH₃), 0.87 (s, 3H, 18-CH₃), 0.61 (s, 3H, 19-CH₃). ¹³C NMR (101 MHz, CDCl₃) δ 174.2, δ 161.1, 138.6, 106.1, 100.3, 72.1, 68.3, 66.2, 56.1, 55.5, 39.9, 39.6, 35.6, 35.5, 35.3, 34.9, 33.1, 31.4, 31.2, 30.9, 28.4, 23.9, 23.0, 20.8, 18.5, 11.9.

Compound 15

White solid and yield 66.4%. ESI-MS: calcd for C₃₃H₄₈O₆ [M + H]⁺ 541.3524, found 541.3463, [M + Na]⁺ 563.3343, found 563.3265. ¹H NMR (400 MHz, CDCl₃) δ 8.00 (d, J = 7.9 Hz, 2H, Ar–H), 7.38 (d, J = 8.0 Hz, 2H, Ar–H), 5.13 (s, 2H, Ph–CH₂), 3.89 (s, 3H, COOCH₃), 3.81 (s, 1H, 7-CH), 3.47–3.36 (m, 1H, 3-CH), 0.90 (d, J = 6.4 Hz, 3H, 21-CH₃), 0.87 (s, 3H, 18-CH₃), 0.61 (s, 3H, 19-CH₃). ¹³C NMR (101 MHz, CDCl₃) δ 173.7, 166.5, 140.9, 129.6, 129.6, 127.5, 71.7, 68.2, 65.1, 55.6, 51.9, 50.2, 42.4, 41.3, 39.5, 39.4, 39.2, 35.2, 35.1, 34.8, 34.5, 32.6, 30.9, 30.7, 30.4, 27.9, 23.4, 22.6, 20.4, 18.1, 11.5.

Compound 16

White solid and yield 48.2%. ESI-MS: calcd for C₃₀H₅₁NO₅S [M + Na]⁺ 560.3386, found 560.3358. ¹H NMR (400 MHz, CDCl₃) δ 6.50 (s, 1H, NH), 4.63 (dd, J = 7.8, 4.9 Hz, 1H, –NHCH̲–), 3.75 (s, 1H, 3-CH), 3.67 (s, 3H, CH₃COO), 3.35 (q, J = 7.2, 5.9 Hz, 1H, 7-CH), 2.44 (t, J = 7.5 Hz, 2H, CH₂S), 2.01 (s, 3H, SCH₃), 0.86 (d, J = 6.3 Hz, 3H, 21-CH₃), 0.82 (s, 3H, 18-CH₃), 0.58 (s, 3H, 19-CH₃). ¹³C NMR (101 MHz, CDCl₃) δ 173.3, 172.4, 71.5, 67.9, 55.7, 52.2, 51.1, 50.1, 42.3, 41.3, 39.4, 39.1, 35.2, 35.1, 34.7, 34.4, 33.0, 32.5, 31.4, 30.3, 29.7, 27.9, 23.4, 22.5, 20.3, 18.1, 15.2, 11.5.

Compound 17

White solid and yield 39.8%. ESI-MS: calcd for C₃₁H₄₆O₄ [M–H]⁻481.3318, found 481.3601. ¹H NMR (400 MHz, DMSO-d₆) δ 9.43 (s, 1H, Ph–OH), 7.14 (t, J = 7.8 Hz, 1H, Ar–H), 6.78–6.66 (m, 3H, Ar–H), 4.99 (s, 2H, Ph–CH₂), 4.44 (s, 1H, 3-OH), 3.37 (s, 1H, 3-CH), 0.86 (d, J = 4.0 Hz, 6H, 21-CH₃, 18-CH₃), 0.58 (s, 3H, 19-CH₃). ¹³C NMR (101 MHz, DMSO-d₆) δ 173.5, 157.8, 138.0, 129.8, 118.8, 115.3, 115.1, 70.3, 65.7, 56.5, 55.9, 42.7, 41.9, 36.7, 35.8, 35.6, 35.2, 34.6, 31.1, 30.8, 28.1, 27.3, 26.6, 24.3, 23.7, 20.8, 18.5, 12.3.

Compound 18

White solid and yield 36.5%. ESI-MS: calcd for C₃₁H₄₆O₅ [M + H]⁺ 499.3418, found 499.3429. ¹H NMR (400 MHz, DMSO-d₆) δ 9.27 (s, 2H, Ph–OH), 6.17 (s, 2H, Ar–H), 6.14 (s, 1H, Ar–H), 4.88 (s, 2H, Ph–CH₂), 4.43 (s, 1H, 3-OH), 3.37 (m, J = 10.8, 4.5 Hz, 1H, 3-CH), 0.87 (d, J = 6.8 Hz, 6H, 21-CH₃, 18-CH₃), 0.59 (s, 3H, 19-CH₃). ¹³C NMR (101 MHz, DMSO-d₆) δ 173.5, 158.9, 138.5, 106.2, 102.4, 70.3, 65.7, 56.5, 55.9, 42.7, 41.9, 36.7, 35.8, 35.6, 35.2, 34.7, 31.1, 31.1, 30.8, 28.2, 27.4, 26.6, 24.3, 23.7, 20.9, 18.5, 12.3.

Compound 19

White solid and yield 70.4%. ESI-MS: calcd for C₃₁H₄₆O₅ [M–H]⁻ 497.3267, found 497.2931. ¹H NMR (400 MHz, DMSO-d₆) δ 9.43 (s, 1H, OH), 7.14 (t, J = 7.8 Hz, 1H, Ar–H,), 6.77–6.66 (m, 3H, Ar–H), 4.98 (s, 2H, Ph–CH₂), 4.43 (d, J = 4.5 Hz, 1H, 7-CH), 3.86 (d, J = 6.7 Hz, 1H, 3-CH), 3.29 (d, J = 14.0 Hz, 2H, 7-OH, 3-OH), 0.87 (d, J = 6.2 Hz, 6H, 21-CH₃, 18-CH₃), 0.59 (s, 3H, 19-CH₃). ¹³C NMR (101 MHz, DMSO-d₆) δ 173.5, 157.8, 138.1, 129.8, 118.8, 115.3, 115.1, 70.2, 69.9, 65.7, 56.3, 55.1, 43.5, 43.4, 42.6, 38.2, 37.7, 35.3, 35.2, 34.2, 31.2, 31.1, 30.7, 23.6, 21.3, 18.7, 12.4.

Compound 20

White solid and yield 48.9%. ESI-MS: calcd for C₃₃H₄₈O₆ [M + Na]⁺ 563.3343, found 563.3563. ¹H NMR (400 MHz, methanol-d₄) δ 7.38 (d, J = 8.6 Hz, 2H, Ar–H), 7.08 (d, J = 8.5 Hz, 2H, Ar–H), 5.09 (s, 2H, Ph–CH₂), 3.51–3.41 (m, 2H, 3-CH, 7-CH), 2.26 (s, 3H, CH̲₃COO), 0.95 (s, 3H,19-CH₃), 0.93 (d, J = 6.4 Hz, 3H, 21-CH₃), 0.66 (s, 3H, 15-CH₃). ¹³C NMR (101 MHz, methanol-d₄) δ 175.7, 171.2, 152.3, 135.6, 130.7, 123.1, 72.3, 72.1, 66.7, 57.7, 56.7, 44.9, 44.7, 44.2, 41.7, 40.9, 38.8, 38.2, 36.8, 36.3, 35.39, 32.5, 32.4, 31.3, 29.8, 28.1, 24.2, 22.6, 21.2, 19.1, 12.9.

4.2. Cell culture

Malignant tumor cell lines (HCT 116, HT-29, SW 489, MGC, 4803, A549, HuH7, SK) and RAW264.7 cells were treated and sub-cultured when the cells grew to about 80% density in the culture flask. The medium in the plate was discarded in a ultra-clean workbench, washed with sterile PBS buffer, and complete medium with 10% FBS was added into the cell suspension. RAW264.7 cells were grown in DMEM complete medium containing 10% FBS. An appropriate amount of cell suspension was added to a new cell culture flask, fresh complete medium was added and the cells were cultured in a constant incubator in 5% CO₂ at 37 °C.

4.2.1. Cell proliferation inhibition assay

The proliferation inhibitory activity of the 20 compounds and four BAs against nine human malignant cell lines (HCT 116, HT-29, SW 489, MGC, 4803, A549, HuH7, and SK) was evaluated by the MTT method. The positive control was 5-fluorouracil (5-FU). Specifically, cells grown to the log-growth phase were seeded on 96-well plates (4000 cells per well) containing 200 μL of medium per well, and cells were incubated with medium containing the compounds, where the compound concentrations range from 200, 100, 50, 25, 12.5, 6.25, 3.125, to 1.5625 μmol L⁻¹. After that, 20 μL of MTT solution at 5 mg mL⁻¹ was added, the medium was dropped after 4–6 h, and 100 μL DMSO was added to each well to dissolve purple crystalline methyl. The OD value was determined at 570 nm on a microplate reader. All the experiments were performed independently at least three times, and half of the inhibitory concentration (IC₅₀) was calculated for each compound. The inhibition rate of tumor cells by the tested compounds under different administration concentration conditions was calculated by Excel software. IBM SPSS 25.0 statistical software was used to calculate the IC₅₀ value, the IC₅₀ value was expressed as the mean ± standard deviation (mean ± SD), inhibition rate = 1 − (experimental group OD value-blank group OD value)/(control group OD value-blank group OD value). The concentration–time-inhibition curve of drug action is obtained by the cell proliferation inhibition assay.

RAW264.7 cells grown to the log-growth phase were seeded on 96-well plates (7000 cells per well) and the cells were incubated with the medium containing the compounds, where the compound concentrations range from 200, 100, 50, 25, 12.5, 6.25 to 3.125 μmol L⁻¹. Thereafter, 10 μL of MTT solution at 5 mg ml⁻¹ concentration was added, the medium was dropped after 4–6 h, and DMSO (100 μL) was added to each well. The OD value was determined at 570 nm on a microplate reader. All the experiments were performed independently at least three times, and half of the inhibitory concentration (IC₅₀) was calculated for each compound. The inhibition rate of tumor cells by the test compounds under different drug administration concentration conditions was calculated by Excel software. IBM SPSS 25.0 statistical software was used to calculate the IC₅₀ value, the IC₅₀ value was expressed as the mean ± standard deviation (mean ± SD), inhibition rate = 1 − (experimental group OD value-blank group OD value)/(control group OD value-blank group OD value).

4.2.2. Cell clone formation assay

With 2000 cells per well seeded in 6-well plates and after cells were adherent, the medium was replaced with compounds at 25 μmol L⁻¹. The culture medium was changed every 3 days. After about 10 days, the cell colonies consisted of at least 50 surviving cells, and the culture was terminated. The medium was discarded, washed with PBS and fixed with 4% paraformaldehyde cell fixative for 30 min. After fixation, it was stained with crystal violet solution (1%) for 20 min. After washing, cell clone formation in each well was recorded with a camera.

4.2.3. Cell-cycle assay

HCT 116 cells were seeded in 6-well plates at 1 × 10⁵ cells per well and after cells were adherent, the medium was replaced with compounds at 25 μmol L⁻¹. Cells were harvested after 48 h. Cells were harvested after trypsin digestion, centrifuged at 1200 r min⁻¹ 4 °C for 5 min and washed three times with PBS. According to the kit instructions, the cells were stained with propidium iodide (PI) DNA staining solution for 30 min in the dark. Finally, the DNA content was determined by flow cytometry.

4.2.4. Cell migration and invasion tests

Cell migration: 600 μL of DMEM medium containing 20% FBS was added to the Transwell subchamber, then HCT 116 cells were digested with serum-free medium and a HCT 116 cell suspension with a cell density of 1 × 10⁵ ml⁻¹ was prepared. 200 μL was added to the upper chamber of the Transwell chamber, followed by addition of test compounds to a concentration of 20 μmol l⁻¹. After 48 h in the incubator, the medium in the chamber was discarded and cells were washed with PBS buffer solution once. 1 mL of 4% paraformaldehyde cell fixative was added to each well over 30 min. After fixation, when stained with crystal violet solution (1%) for 20 min, the cells in the upper chamber layer were wiped off with a cotton ball, and the residual crystal violet was cleaned with distilled water. The number of migrated cells in at least 5 visual fields was observed and recorded under the microscope.

Cell invasion: Matrigel base glue was dissolved on ice in a 4 °C refrigerator overnight, diluted with DMEM medium, with 50 μL of Matrigel in the upper chamber of the Transwell chamber, and glued at 37 °C for 30 minutes. Next, 600 μL of DMEM medium containing 20% FBS was added to the Transwell subchamber, digested with good colon cancer HCT 116 cells with serum-free medium and a HCT 116 cell suspension with a cell density of 2 × 10⁵ ml⁻¹ was prepared. 200 μL was added to the upper chamber, and test compounds were added to a concentration of 20 μmol l⁻¹. After 48 h in the incubator, the medium in the chamber was discarded and cells were washed with PBS buffer solution once. One ml of 4% paraformaldehyde cell fixative was added to each well over 30 min. After fixation, when stained with crystal violet solution (1%) for 20 min, the cells in the upper chamber layer were wiped off with a cotton ball, and the residual crystal violet was cleaned with distilled water. The number of invading cells in at least five visual fields was observed and recorded under the microscope.

4.2.5. Apoptosis assay

HCT 116 cells were seeded in 6-well plates of 1 × 10⁵ cells per well, and the medium was replaced with tested compounds at 25–30 μmol L⁻¹. Cells were harvested after 48 h. Cells were harvested after trypsin digestion, centrifuged at 1200 r min⁻¹ 4 °C for 5 min and washed three times with PBS. Cells were resuspended with 100 μL of binding buffer according to the kit instructions, followed by addition of 5 μL of Annexin-V-FITC and 10 μL of PI staining solution, and incubated at room temperature in the dark for 10–15 min. 400 μL of binding buffer was added, vortex mixed and placed on ice, and the samples were detected by flow cytometry within 1 h.

4.2.6. Staining

Well-established colon cancer HCT 116 cells were seeded in 6-well plates of 8000 cells per well, and after the cells were adherent, the medium was replaced with tested compounds at 30 μmol L⁻¹. After 48 h of drug administration, the culture medium was discarded, washed with PBS buffer, and fixed with 1 ml of 4% paraformaldehyde cell fixative in each well for 30 min. After fixation, when stained with Giemsa for 20 min, the excess dye solution was washed away with distilled water, and photographed under the microscope to observe the morphological changes of the cells after drug treatment.

Well-established colon cancer HCT116 cells were seeded in 6-well plates of 8000 cells per well, and after the cells were adherent, the medium was replaced with tested compounds at 30 μmol L⁻¹. After 48 h of drug administration, the culture medium was discarded, washed with PBS buffer, and fixed with 1 ml of 4% paraformaldehyde cell fixative in each well for 30 min. After fixation, the cells were stained with DAPI dye solution for 15 min, washed five times in distilled water, and photographed under a fluorescence microscope to observe the fluorescence strength of the cells after drug action.

4.2.7. Mitochondrial membrane potential determination

HCT 116 cells were seeded in 6-well plates with 1 × 10⁵ cells per well and after cells were adherent, the medium was replaced with fresh medium containing tested compounds. The culture medium in each well was discarded after 48 h of drug treatment, and 700 μL of serum-free medium with diluted tetramethylethyl rhodamine ester (TMRE) dye was added to each well and incubated at 37 °C for 30 min. After trypsin digestion, the cells were harvested, centrifuged at 1200 r min⁻¹ 4 °C for 5 min and washed three times with PBS. Changes in cell mitochondrial membrane potential were measured by flow cytometry.

4.2.8. Measurement of ROS

HCT 116 cells were seeded in 6-well plates with 1 × 10⁵ cells per well and after cells were adherent, the medium was replaced with fresh medium containing tested compounds. The culture medium in each well was discarded after 24 h of drug treatment, then 700 μL of serum-free medium with diluted 2′,7′-dichlorofluorescein diacetate (DCFH-DA) dye was added to each well, and incubated at 37 °C for 30 min. After trypsin digestion, the cells were harvested, centrifuged at 1200 r min⁻¹ 4 °C for 5 min and washed three times with PBS. Changes in cellular ROS were determined by flow cytometry.

4.2.9. Western blot

The comb was pulled out vertically up, and then the calculated amount of sample was added (4 μL). The electrophoresis liquid was added to the electrophoresis tank, and using 80 V constant voltage electrophoresis for 30 min to the concentration gel and separation line, the voltage was adjusted to 120 V constant voltage electrophoresis. After about 1 h, the electrophoresis was stopped when the bromophenol blue strip was close to the bottom of the glass plate. A single PVDF membrane was cut, the methanol was activated for 5 min, and a sandwich model was mounted with a 250 mA constant flow membrane for 2 h. The electric transfer process releases heat and requires an ice bath. After membrane transfer, it was blocked on a shaker with TBST buffer containing 5% skim milk powder for 2 h at room temperature. After blocking, the cells were washed five times with 1 TBST buffer for 5 min each. The band was placed in a preconfigured primary antibody at 4 °C overnight. The bands were removed, washed five times with TBST buffer for 5 min each and incubated for secondary antibody. Incubation with secondary antibody was completed and the washing process was repeated. The protein band was placed onto a tray, the ECL luminescent solution was added, imaged using a fully automated chemiluminescence imaging system, and gray scale analysis was conducted using ImageJ software.

4.3. NO inhibition tests

RAW264.7 cells grown to the log growth phase were seeded on 96-well plates (7000 cells per well), and the cells were incubated with compound-containing medium for 24 h, where the compound concentration gradient range was set to the safe concentration range of different compounds, while 1 μg ml⁻¹ of LPS was added. Setting of the blank and LPS was required to stimulate the control. After being cultured for 24 h, 50 μL of cell culture supernatant was removed, and 50 μL of Griess A and then 50 μL of Griess B were added according to the Griess kit instructions. In the LPS group, the OD values were read at 540 nm of the microplate reader. The nitric oxide inhibition rate was calculated for all the compounds.

4.4. ELISA tests

RAW264.7 cells were seeded in 6-well plates of approximately 1 × 10⁵ cells per well. After the cells were attached overnight, the drugs were added. The blank group, LPS + compound 10 (10 μM and 25 μM) and LPS + compound 18 (10 μM and 25 μM), three groups, were set. After 24 h, the cell culture supernatant was removed first to calculate the protein concentration, and the volume of the required cell culture supernatant was determined according to the quantitative results. TNF-α was determined according to the instructions used for the Elabscience mouse TNF-α enzyme-linked immunosorbent assay kit.

4.5. Statistical analysis

Each value from all statistical analysis is expressed as mean ± SD, all derived from independent replicates in at least triplicate in the experiment. Experimental data were analyzed using IBM SPSS 26.0 software and GraphPad Prism 8.0 software, and one-way analysis of variance (ANOVA) was used to analyze the statistical differences between the control and experimental groups, and a p-value less than 0.05 was statistically significant.

Conflicts of interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.