Design, Synthesis, and Cell Lines Studies of Oleanolic Acid—Hydrogen Sulﬁde Donor Hybrids

: In order to develop new oleanolic acid (OA) derivatives endowed with improved antitumor activities, for the ﬁrst time, a number of new hybrid compounds were reported by combining OA or 3-oxooleanolic acid with appropriate H 2 S-donor moiety, coupled via a suitable linker. The anti-tumor evaluation indicated that they exhibited excellent anti-cancer activities against the tested cancer cell lines. Moreover, 18d with 5-(4-hydroxyphenyl)-3 H -1,2-dithiole-3-thione moiety as H 2 S donor and β -alanine as the linker, showed more potent cytotoxicity against the tested cancer cell lines than OA and 3-oxooleanolic acid, especially for A549 cells. Furthermore, the preferred compound, 18d , prefer-entially accumulates in cancer cells (13.6 µ M) over the matched normal cells LO2 (>100 µ M) in vitro. The improved antitumor activity of this hybrid was probably due to its H 2 S-releasing capability.


Introduction
One of the major causes of death in patients is cancer, and a major factor to explore in cancer treatment is novel chemotherapy. Approximately 90% of patient deaths are due to cancer or metastatic cancer. Current anticancer agents cannot cure primary tumor and metastatic cells. Thus, the design of novel antitumor therapeutic agents, including oleanolic acid (OA) activities, represents an area in need of urgent attention.
The optimization of compounds derived from natural origins represents one of the promising strategies which is widely used in seeking and developing anticancer drug molecules [1]. Oleanolic acid (OA), a well-known natural pentacyclic triterpenoid, has anti-inflammatory and antitumor activities [2,3]. A number of studies have demonstrated that OA can act at various stages of tumor development to inhibit tumor initiation and promotion, as well as to induce tumor cell differentiation and apoptosis [4]. However, the antitumor activity of OA is relatively weak, and the aqueous solubility of OA is also very poor, which greatly hinders its clinical application [5]. Thus, on the basis of the structure of OA, there have been multiple studies concerning the structural modification and improvement of their antitumor activities and aqueous solubility [6,7]. More importantly, many hybrid OA compounds have been synthesized, including amino acid/dipeptide prodrugs of OA [4,8,9] and nitric oxide (NO) donor hybrid OA compounds [4,[10][11][12][13], which showed significantly improved antitumor activity and pharmacokinetic properties. However, to the best of our knowledge, there were no reports focused on hydrogen sulfide (H 2 S) donor hybrid OA compounds.
Similar to the other two gasotransmitters, NO and carbon monoxide (CO), hydrogen sulfide (H 2 S) is a signaling molecule and plays an important part in various physiological processes [14][15][16][17]. Recently, H 2 S donor hybrid compounds have focused on the treatment of tumors; H 2 S is a signaling molecule involved in the apoptosis of tumor cells, and H 2 S donors have reported antiangiogenic effects [18][19][20][21]. In previous reports, several H 2 S donor hybrid compounds have been synthesized and biologically evaluated, which were identified to display increased effectiveness and reduced toxicity when compared with their parent drugs due to synergic effects [22][23][24][25][26][27][28][29][30]. However, the linkage between drugs and hydrogen sulfide donors was relatively simple, most of which was connected by ester bonds, not a heterozygous modification [26][27][28][29][30]. Additionally, compared with the other reported slow-releasing hydrogen sulfide (H 2 S) donors, only a slow-releasing hydrogen sulfide (H 2 S) donor in aqueous media over a period of hours to days reportedly compared with sulfide salts such as sodium hydrosulfide (NaHS) and sodium sulfide (Na 2 S) in its evaluation of biological effects in cells, tissues and animals, but there are few examples of connecting it to other parent anticancer drugs [26][27][28][29][30]. In recent years, it has been found that the efficiency can be improved by coupling different types of specific or sensitive linkages. Among them, the coupling of amino acids has captured the greatest attention in pro-drug design. The coupling of amino acids can be selectively identified and taken up by peptide transporter 1 (PepT1) to improve the oral absorption of parent drugs with undesirable biopharmaceutical characteristics [4,9]. PepT1 appears to be an attractive target with its high capacity, broad substrate specificity, and high level of expression in the intestinal epithelium. Many PepT1 targeted prodrugs have been designed and synthesized with the aim of improving oral bioavailability of parent drugs, in which valacyclovir and valganciclovir are two excellent examples (they are L-Val ester prodrugs of acyclovir and ganciclovir, respectively) [8]. Furthermore, the redox sensitiveness disulfide structure is a promising linkage because the redox process is widely present in physiological environments. Especially, there is a significant redox gradient for tumors between the intracellular components and the extracellular environments. Many studies have reported that polymer micelles with disulfide structures have a good ability to transfer antitumor drug to cancer cells and effectively perform intracellular drug release [31,32].
Regarding the abovementioned studies, it may be interesting to study the hybrid OA compounds containing H 2 S-donating species through different linkers. Herein, oleanolic acid and 3-oxooleanolic acid (3-oxo-OA) were coupled with a number of classic H 2 S-releasing moieties via different linkers, including ester bonds, amino acids, reduction-sensitive disulfide bonds, and the chemically and metabolically stable ether bonds. Their anticancer, H 2 Sreleasing ability and the effects of the type of linker were biologically evaluated.

General Procedure for the Synthesis of Targeted Compounds
We purchased all reagents from Shanghai Chemical Reagent Company. The silica gel of the Column chromatography was 200-300 mesh and the Column chromatography was examined by thin layer chromatography performed on GF/UV 254 plates and was visualized with UV light at 365 and 254 nm. Results of the 1 H NMR spectra: BrukerAVANCE III equipment at 300 or 400 MHz, in CDCl 3 unless otherwise specified; δ in ppm relative to Me4Si, J in Hz. The 13 C NMR spectra: BrukerAVANCE III equipment at 75 or 100 MHz, in CDCl 3 unless otherwise specified; δ in ppm relative to Me4Si. High-resolution mass spectra (HRMS; m/z) were taken using a Thermo QE spectrometer.
The starting materials OA and 4-OH-TBZ were purchased directly. The compound 3-oxo-OA was synthesized using the reported method [33]. The H 2 S-releasing groups ADT-OH and N-benzyl-4-hydroxybenzothioamide were prepared according to the reported methods [34,35] (20 mL) solution. The mixture was stirred for 20 min under room temperature [36,37]. Then, ADT-OH or 4-OH-TBZ or N-benzyl-4-hydroxybenzothioamide (2.0 mmol, dissolved in 10 mL dry CH 2 Cl 2 ) was added dropwise to the mixture and the reaction solution was whisked until the end of the reaction. Then, we washed the whole solution using water (20 mL × 3), saturated NaHCO 3 aqueous solution (20 mL × 3) and saturated brine (20 mL × 3), dried and filtered it with anhydrous Na 2 SO 4 , inspissated it under reduced pressure to obtain a remainder, and purified it by flash column chromatography to acquire pure product.

General Procedure for the Preparation of 4a-d
The corresponding dibromoalkane (3.0 mmol), KOH (1.0 mmol) and KI (0.1 mmol) were added to OA (1.0 mmol) which was dissolved in anhydrous tetrahydrofuran (20 mL) solution. The reaction solution was heated to reflux and whisked until the end of the reaction. The solvent was removed under vacuum. The remainder was diluted with water and extracted with EtOAc. The organic layer was washed with brine and then dried with sodium sulfate, filtered, and evaporated in vacuo to give corresponding middle product, 3a or 3b, respectively, which was used in the next step without further purification. ADT-OH or 4-OH-TBZ (2.0 mmol), KOH (2.0 mmol) and KI (0.1 mmol) were added to compounds 3a or 3b (1.0 mmol), which were dissolved in anhydrous tetrahydrofuran (20 mL) solution. The reaction mixture was heated to reflux and whisked until the end of the reaction. Then, the solvent was removed under vacuum, and the remainder was diluted with water and extracted with EtOAc. The organic layer was washed with brine, dried over anhydrous sodium sulfate, filtered, and evaporated under reduced pressure. The resulting crude product was purified by flash chromatography on silica gel column to provide the target compound 4a-d, respectively.
Oleanolic acid 28-[4-(4-carbamothioylphenoxy)] butyrate (4a). Yield: 24%. Oil: 1 13 13    OA (1.0 mmol) was dissolved in 20 mL anhydrous dichloromethane solvent, to which 1,5,6-oxadithionane-2,9-dione (4.0 mmol) or succinic anhydride (4.0 mmol) and a certain amount of catalyst DMAP were added. This reaction mixture was heated to reflux while continuously stirring to ensure full reaction. The resulting reaction mixture was first diluted with anhydrous dichloromethane solvent (10 mL). The diluted solution was washed with water (20 mL × 3), saturated aqueous sodium bicarbonate solution (20 mL × 3), and saturated brine (20 mL × 3) three times. The organic layer was collected, dried over anhydrous sodium sulfate, and concentrated in vacuo to obtain intermediate product 5 or 7. The intermediate product 5 or 7 (1.0 mmol) was dissolved in 20 mL of anhydrous dichloromethane solvent, dehydrating agent EDCI (2.4 mmol) was added, and a certain amount of dimethylaminopyridine catalyst was present in the flask. This was stirred for 20 min at room temperature. Then, N-benzyl-4-hydroxybenzothioamide or 4-OH-TBZ or ADT-OH 2.0 mmol was dissolved in 10 mL of anhydrous dichloromethane solvent, dropwise added to the flask, and continued to react until the reaction was completed. Then, the whole solution was washed 3 times with water (20 mL × 3), saturated aqueous sodium bicarbonate solution (20 mL × 3), and saturated brine (20 mL × 3), in that order. The organic layer was collected, dried over anhydrous sodium sulfate, filtered, evaporated in vacuo, and passed through the column; the residues obtained after separation and purification were the corresponding products 6a, 6b and 8a-c.        A mixture of OA (1.0 mmol) and Et 3 N (1.0 mmol) in anhydrous CH 2 Cl 2 (25 mL) was stirred at room temperature for 10 min. The reaction mixture was cooled to 0 • C, oxalyl chloride (5.0 mmol) in anhydrous CH 2 Cl 2 (2 mL) was added dropwise, and the mixture was allowed to stir at 0 • C for an additional 30 min. Then, the solvent was removed in vacuo to give the middle product 9, which was used in the next step without further purification. To a solution of different ethyl-amino acid (5.0 mmol) and TEA (10.0 mmol) in dry CH 2 Cl 2 (25 mL), the middle product 9 (1.0 mmol) in anhydrous CH 2 Cl 2 (2 mL) was added dropwise at 0 • C, and the mixture was stirred at room temperature for about 5 h. Then, the solvent was removed in vacuo, and the residue was diluted with water and extracted with EtOAc. The organic layer was washed with saturated NH 4 Cl aqueous solution, brine, dried over anhydrous sodium sulfate, filtered, and evaporated under reduced pressure to obtain the middle products 10a or 10b, respectively. The middle product, 10a or 10b (1.0 mmol), was dissolved in dry THF (25 mL), and 4 N NaOH (20.0 mmol) was added into the flask. The mixture was stirred at room temperature until the starting material was totally consumed, as indicated by TLC. The reaction mixture was diluted with H 2 O (20 mL), neutralized with 1 N HCl, and extracted three times with CHCl 3 . The combined organic layers were washed with brine, dried over Na 2 SO 4 , and concentrated under vacuum to afford the middle products 11a or 11b. The reaction mixture with 20 mL of water was diluted and neutralized with 1 N HCl for acid-base, the neutralized solution was extracted three times with chloroform, and the organic layer was collected. The brine was used to wash the composite organic layer, dried over anhydrous sodium sulfate, and concentrated in vacuo to obtain intermediate product 11a or 11b. A solution of 20 mL anhydrous dichloromethane was added to 11a or 11b (1.0 mmol) to dissolve it, and certain amounts of dimethyl-aminopyridine and dehydrating agent EDCI (490 mg, 2.4 mmol) were added. This mixture was stirred for 20 min at room temperature. ADT-OH 2.0 mmol was dissolved in 10 mL anhydrous dichloromethane solution and added to the reaction solution, and the reaction continued at room temperature until the end of the reaction. Then, the reaction solution was washed with water (20 mL × 3), saturated aqueous sodium bicarbonate solution (20 mL × 3), and saturated brine (20 mL × 3) successively, dried over anhydrous sodium sulfate, filtered, and the solvent was distilled off under reduced pressure; after separation and purification by column chromatography, the pure products 12a or 12b could be obtained.

General Procedure for the Preparation of 14a-c
Compound 1 was treated in acetone at 0 • C with Jones reagent to form the C-3 oxidized derivative 13 in an almost quantitative yield. Then, using compound 13 as a starting material, the target products 14a-c were obtained according to the method of preparing 2a-c described above.

General Procedure for the Preparation of 18a-f
Using compound 13 instead of compound 1 as raw material, the target products 18a-f were obtained according to the method of preparing 12a and 12b described above.      13

Cell Lines
Human hepatocellular carcinoma cell line (HepG2), breast cancer cell line (MCF-7) and lung adenocarcinoma cell line (A549) were cultured in a humidified 5% CO 2 atmosphere at 37 • C and incubated in RPMI-1640 medium supplemented with 10% fetal bovine serum (FBS).

MTT Assay
HepG2, A549, or MCF-7 cell lines were grown on 96-well plates at a cell density of 5000 cells/well in RPMI-1640 medium with 10% FBS. The plates were incubated at 37 • C in a humidified atmosphere of 5% CO 2 /95% air overnight. The cells were then exposed to different concentrations of selected compounds and cisplatin for another 48 h in an atmosphere of 95% air with 5% CO 2 at 37 • C. The cells were stained with MTT in the incubator for 4 h at 37 • C, and the absorbance was read at 490 nm on a microplate reader (Thermo Fisher Scientific). Experiments were carried out in triplicates and independently repeated three times.

Evaluation of Hydrogen Sulfide Release Activity
DNS-Az by itself is nonfluorescent. Nevertheless, the DNS-Az solution with an addition of hydrogen sulfide formed dansyl amide, which revealed strongly enhanced fluorescence. Without specimen pre-processing, the rapid response by DNS-Az to sulfide can be used to detect the instantaneous changes of hydrogen sulfide levels. The initial H 2 S-OA concentration of 200 µM was added to 60 µL of dansylazide solution (10 mM in ethanol) and 40 µL of H 2 S-OA stock solution (10 mM in DMSO) in preheated 1900 µL of culture medium (DMEM) at 37 • C. All trials were carried out at least three times. The Waters e2695-2998 chromatograph system which performed the HPLC analyses was equipped with a 2475 fluorescence detector integrated in the Waters e2695 system. Data analysis was implemented using an Empower system (Waters Technologies). The sample was eluted on Symmetry C 18 (250 × 4.6 mm, 5 µm; Waters); the amount of injection was 20 µL. The mobile phase was composed of 0.1% aqueous HCOOH/ CH 3 CN (18:82, v/v); elution-adopted isasteric mode and flow rate was 1.0 mL/min. Using excitation and emission wavelengths of 340 and 535 nm, respectively, the fluorescence signals were obtained (gain factor = 10). Data operation was implemented by Empower 3. The values acquired from integration of the peak of dansyl amide were interpolated in NaHS as a standard calibration line; therefore, the concentration of dansyl amide in each specimen was an indicatrix of H 2 S amount [24]. The detection limit was about 1 µM with a signal:noise ratio (S/N) of 3:1; the amounts of H 2 S could be tested and quantified at 0.5 mM combined concentrations. A linear relationship between the fluorescence intensity and in DMEM against hydrogen sulfide was shown, and the assessment of the linear range of the H 2 S release was [3 µM, 250 µM]. The measured concentrations were within the range of the standard curve.

Chemistry
The synthesis of oleanolic acid-hydrogen sulfide donor hybrids and 3-oxooleanolic acid-hydrogen sulfide donor hybrids are outlined in Schemes 1 and 2, respectively. In this study, we selected 5-(4-hydroxyphenyl)-3H-1,2-dithiole-3-thione (ADT-OH), 4-hydroxythiobenzamide (4-OH-TBZ), and N-benzyl-4-hydroxybenzothioamide (N-Bn-4-OH-TBZ) as H 2 S-releasing moieties. OA or 3-oxo-OA was directly modified with different H 2 S donors in the presence of 1-(3-dimethylaminopropyl)-3-ethylcarbodiimide hydrochloride (EDCI) and 4-dimethylaminopyridine (DMAP) to produce the corresponding C-28 ester derivatives (2a-c and 14a-c). In order to enrich the types of linking bonds, we tried to connect the dibromoalkane to 3-OH of OA to obtain C-3 ether derivatives. Reaction of OA with dibromoalkane in the presence of potassium hydroxide and 0.1 mmol of potassium iodide in dry tetrahydrofuran afforded intermediates 3a,b. Unexpectedly, the NMR spectra of the intermediates 3a,b showed that they were C-28 ester derivatives. These results indicated that the reactivity of 3-OH was too low to participate in the reaction under these conditions. Thereafter, treating the intermediates 3a,b with H 2 S donors under basic conditions afforded the hybrids 4a-d. Furthermore, the preparation of C-3 ester derivatives (6a,b and 8a-c) was carried out in two steps. OA was reacted with succinic anhydride or 1,5,6-oxadithionane-2,9-dione in the presence of DMAP to form 3-O-acyl derivatives (5 and 7), which were further esterified with corresponding H 2 S donors in the presence of EDCI/DMAP to generate target compounds. Then, we tried to couple the amino acid moiety to the 28carboxyl group. In order to improve the reactivity and avoid the generation the C-28 ester by-products, the reaction was carried out in multiple steps. Generally, OA or 3-oxo-OA first reacted with oxalyl chloride to give intermediates (9 or 15), and then a different ethylamino acid in the presence of triethylamine (TEA) was added to yield intermediates (10a,b  or 16a-c). Thereafter, the ethyl group was successfully removed by treating the above intermediates with sodium hydroxide, resulting in 11a,b or 17a-c, respectively. In the presence of EDCI/DMAP, the OA-or 3-oxo-OA-amino acid-H 2 S donor trihybrids (12a,b or 18a-f) were successfully obtained by further coupling with different H 2 S donors. Their structures were characterized by 1 H NMR, 13 C NMR, and HRMS.

The Antiproliferative Activities of Compounds in Different Cancer Cell Lines
The target complexes 2a-c, 4a-d, 6a,b, 8a-c, 12a,b, 14a-c, and 18a-f, in parallel with OA and 3-oxo-OA, were initially evaluated by the MTT assays for their cytotoxicities on human hepatocellular carcinoma cell line (HepG2), human lung adenocarcinoma cell line (A549), and human breast cancer cell line (MCF-7), using general chemotherapeutic drug oxaliplatin as positive control. As the results show in Table 1, compared with their lead compounds OA or 3-oxo-OA, the synthesized prodrugs showed significantly improved

The Antiproliferative Activities of Compounds in Different Cancer Cell Lines
The target complexes 2a-c, 4a-d, 6a,b, 8a-c, 12a,b, 14a-c, and 18a-f, in parallel with OA and 3-oxo-OA, were initially evaluated by the MTT assays for their cytotoxicities on human hepatocellular carcinoma cell line (HepG2), human lung adenocarcinoma cell line (A549), and human breast cancer cell line (MCF-7), using general chemotherapeutic drug oxaliplatin as positive control. As the results show in Table 1, compared with their lead compounds OA or 3-oxo-OA, the synthesized prodrugs showed significantly improved cytotoxicities against all of the tested tumor cell lines. In contrast, the precursor compounds (3a, 3b, 5, 7, 11a, and 11b) without the H 2 S-donor moiety, showed only a weak activity, indicating that the H 2 S-donor moiety exerts an important influence on the cytotoxic activities of the prodrugs (data not shown). Furthermore, 5-(4-hydroxyphenyl)-3H-1,2-dithiole-3-thione (ADT-OH), 4-hydroxythiobenzamide (4-OH-TBZ), and N-benzyl-4-hydroxybenzothioamide (N-Bn-4-OH-TBZ), three kinds of hydrogen sulfide donors (Figure 1), were initially evaluated by the MTT assays for their cytotoxicities on human hepatocellular carcinoma cell line (HepG2), human lung adenocarcinoma cell line (A549), and human breast cancer cell line (MCF-7) as well. (Table 2). These results indicated that the enhanced cytotoxicity of the H2S donor derivatives may be due to their H2S-releasing ability, which can achieve the synergic action of cytotoxic H2S and the parent compounds OA or 3-oxo-OA. In addition, because the cytotoxic activity of 3-oxo-OA was higher than that of OA, the cytotoxic activities of H2S-donating 3-oxo-OA derivatives (14a-c, 18b, 18d) were also superior to those of the corresponding H2S-donating OA derivatives (2a-c, 12a,b).
The above results revealed that the type of the H2S-donor moiety was crucial for the cytotoxicity of H2S-donating OA or 3-oxo-OA derivatives. Under the same conditions of the connecting arm and coupling site, the order of potencies was ADT-OH > 4-OH-TBZ > N-Bn-4-OH-TBZ (Table 1), which was almost the same (2b > 2a > 2c, 8b > 8a > 8c, and 14b > 14a > 14c) against the tested tumor cell lines.
In addition, the linkers which linked the H2S donor to OA or 3-oxo-OA had a great influence on the cytotoxic potency of the derivatives. Firstly, the compounds that directly coupled H2S donor to the 28-COOH of OA (2a-c) or 3-oxo-OA (14a-c) exhibited lower cytotoxicities. Secondly, the 3-OH ester derivatives of OA (6a,b, and 8a-c) showed moderate cytotoxicities. Unexpectedly, the 3-OH ester derivatives with redox-sensitized disulfide bonds did not significantly promote cytotoxicities. IC50 values of the compounds with redox-sensitized disulfide bond (8a, 8b) were slightly lower than those without redox-sensitized disulfide bond derivatives (6a, 6b). In contrast, compared with their lead compounds OA or 3-oxo-OA, the amino acid prodrugs 12a,b and 18a-f showed significantly improved cytotoxicity against the three tested cell lines. Among them, compounds 18c and 18d with β-alanine as a linker showed potent cytotoxicity against HepG2 and A549 cell lines. Especially, the cytotoxicities of 18c and 18d against A549 were higher than those of other cell lines, which may be due to the expression of PepT1 in A549, but not in other cells.
On the basis of the above results, the most promising compounds, 18c and 18d, were evaluated for possible cytotoxicity towards normal human liver cells (LO2) to investigate whether the prodrugs have a tumor-cell-selecting activity. Interestingly, the growth of LO2 cell lines was not significantly affected by 18c and 18d, indicating that 18c and 18d selectively inhibit the growth of tumor cells (Table 3). Table 3. The antiproliferative activities of compounds in human normal liver cell lines by the MTT assay.  These results indicated that the enhanced cytotoxicity of the H 2 S donor derivatives may be due to their H 2 S-releasing ability, which can achieve the synergic action of cytotoxic H 2 S and the parent compounds OA or 3-oxo-OA. In addition, because the cytotoxic activity of 3-oxo-OA was higher than that of OA, the cytotoxic activities of H 2 S-donating 3-oxo-OA derivatives (14a-c, 18b, 18d) were also superior to those of the corresponding H 2 S-donating OA derivatives (2a-c, 12a,b).
The above results revealed that the type of the H 2 S-donor moiety was crucial for the cytotoxicity of H 2 S-donating OA or 3-oxo-OA derivatives. Under the same conditions of the connecting arm and coupling site, the order of potencies was ADT-OH > 4-OH-TBZ > N-Bn-4-OH-TBZ (Table 1), which was almost the same (2b > 2a > 2c, 8b > 8a > 8c, and 14b > 14a > 14c) against the tested tumor cell lines.
In addition, the linkers which linked the H 2 S donor to OA or 3-oxo-OA had a great influence on the cytotoxic potency of the derivatives. Firstly, the compounds that directly coupled H 2 S donor to the 28-COOH of OA (2a-c) or 3-oxo-OA (14a-c) exhibited lower cytotoxicities. Secondly, the 3-OH ester derivatives of OA (6a,b, and 8a-c) showed moderate cytotoxicities. Unexpectedly, the 3-OH ester derivatives with redox-sensitized disulfide bonds did not significantly promote cytotoxicities. IC 50 values of the compounds with redox-sensitized disulfide bond (8a, 8b) were slightly lower than those without redox-sensitized disulfide bond derivatives (6a, 6b). In contrast, compared with their lead compounds OA or 3-oxo-OA, the amino acid prodrugs 12a,b and 18a-f showed significantly improved cytotoxicity against the three tested cell lines. Among them, compounds 18c and 18d with β-alanine as a linker showed potent cytotoxicity against HepG2 and A549 cell lines. Especially, the cytotoxicities of 18c and 18d against A549 were higher than those of other cell lines, which may be due to the expression of PepT1 in A549, but not in other cells.
On the basis of the above results, the most promising compounds, 18c and 18d, were evaluated for possible cytotoxicity towards normal human liver cells (LO2) to investigate whether the prodrugs have a tumor-cell-selecting activity. Interestingly, the growth of LO2 cell lines was not significantly affected by 18c and 18d, indicating that 18c and 18d selectively inhibit the growth of tumor cells (Table 3). Table 3. The antiproliferative activities of compounds in human normal liver cell lines by the MTT assay.

H 2 S-Releasing Ability
To investigate whether the H 2 S-donating prodrugs possessed the ability to release H 2 S, 18c and 18d were selected to perform a fluorometric assay. In brief, the confirmation of H 2 S release was based on the reaction of the H 2 S donor compound with the H 2 S fluorescence probe to produce the fluorescent related amide detected by HPLC [24,25]. H 2 S released in DMEM is shown in Figure 2. As expected, compounds 18c and 18d did have the ability to release H 2 S. The percentages of H 2 S released from compound 18c in DMEM were 1.2%, 4.2%, and 7.3% at 1 h, 12 h, and 24 h, respectively, which was indicative of a slow and steady H 2 S release. The trend of compound 18d was very similar to that of 18c. Interestingly, at all points, the % H 2 S released from 18d were higher than that of 18c, which were consistent with the order of the cytotoxicity results. This finding further confirmed that the H 2 S-releasing abilities were quite well associated with their cytotoxic activities.
Appl. Sci. 2021, 11, x FOR PEER REVIEW 14 of 16 slow and steady H2S release. The trend of compound 18d was very similar to that of 18c.
Interestingly, at all points, the % H2S released from 18d were higher than that of 18c, which were consistent with the order of the cytotoxicity results. This finding further confirmed that the H2S-releasing abilities were quite well associated with their cytotoxic activities.

Conclusions
In conclusion, with the hydrogen sulfide donor, 23 new hybrid compounds derived from the combination of OA as well as 3-oxo-OA via various chemical linkers were prepared to improve their pharmaceutical profiles. In this study, the synthesized compounds were screened for anticancer activities against three human cancer cell lines using the MTT assay. The H2S donors or materials were controllable or slow-releasing. We can deduce that the specific H2S-releasing moiety and linker strongly influenced the anticancer activities. Additionally, all of the synthesized pro-drugs showed moderate to potent cytotoxicity against the three tested cell lines. In particular, the cytotoxicity of 18d (IC50 = 11.8 μM) against PepT1 expressed in A549 cell lines was basically equal to that of oxaliplatin. The higher activity of 18d may be attributable to the introduction of the H2S donor and amino acid moieties. In addition, our tests revealed that 18d could release H2S in vitro and has little toxicity against normal human liver cells. Overall, for the first time, H2S-releasing derivatives as anticancer drugs have been reported and indicate a promising strategy for the further development of anticancer drugs.

Conclusions
In conclusion, with the hydrogen sulfide donor, 23 new hybrid compounds derived from the combination of OA as well as 3-oxo-OA via various chemical linkers were prepared to improve their pharmaceutical profiles. In this study, the synthesized compounds were screened for anticancer activities against three human cancer cell lines using the MTT assay. The H 2 S donors or materials were controllable or slow-releasing. We can deduce that the specific H 2 S-releasing moiety and linker strongly influenced the anticancer activities. Additionally, all of the synthesized pro-drugs showed moderate to potent cytotoxicity against the three tested cell lines. In particular, the cytotoxicity of 18d (IC50 = 11.8 µM) against PepT1 expressed in A549 cell lines was basically equal to that of oxaliplatin. The higher activity of 18d may be attributable to the introduction of the H 2 S donor and amino acid moieties. In addition, our tests revealed that 18d could release H 2 S in vitro and has little toxicity against normal human liver cells. Overall, for the first time, H 2 S-releasing derivatives as anticancer drugs have been reported and indicate a promising strategy for the further development of anticancer drugs.
Funding: This research received no external funding.