Design, Synthesis, and Anticancer Activity of Novel 3,6-Diunsaturated 2,5-Diketopiperazines

Based on the marine natural products piperafizine B, XR334, and our previously reported compound 4m, fourteen novel 3,6-diunsaturated 2,5-diketopiperazine (2,5-DKP) derivatives (1, 2, 4–6, 8–16), together with two known ones (3 and 7), were designed and synthesized as anticancer agents against the A549 and Hela cell lines. The MTT assay results showed that the derivatives 6, 8–12, and 14 had moderate to good anticancer capacities, with IC50 values ranging from 0.7 to 8.9 μM. Among them, compound 11, with naphthalen-1-ylmethylene and 2-methoxybenzylidene functions at the 3 and 6 positions of 2,5-DKP ring, respectively, displayed good inhibitory activities toward both A549 (IC50 = 1.2 μM) and Hela (IC50 = 0.7 μM) cancer cells. It could also induce apoptosis and obviously block cell cycle progression in the G2/M phases in both cells at 1.0 μM. The electron-withdrawing functions might not be favorable for the derivatives with high anticancer activities. Additionally, compared to piperafizine B and XR334, these semi-N-alkylated derivatives have high liposolubilities (>1.0 mg mL−1). Compound 11 can be further developed, aiming at the discovery of a novel anticancer candidate.


Introduction
Cancer has become a main threat to human health and life around the world [1]. The migration, resistance, and occurrence of cancer cells cause treatment failure even when modern techniques, such as surgery, radiation, chemotherapy, immunotherapy, or their combination, are employed. Fighting against cancer is a hard and time-consuming war, and new approaches are needed urgently, in which developing novel anticancer agents with high potency still plays a critical role in fighting against this disease [2].
2,5-diketopiperazine (2,5-DKP) is the smallest cyclic peptide and a useful scaffold frequently found in numerous structurally diverse natural products [3][4][5]. 2,5-Diketopiperazine derivatives (2, have higher stability than their linear counterparts against enzymolysis [6] and often have conformational rigidity and are able to interact with various biological targets [7], thus giving rise to a broad range of biological activities [5,8], such as antivirus [9], anticancer [10][11][12][13][14], antifouling [15,16], antioxidation [17], and anti-PAI-1 [18,19] activities. As a result, 2,5-DKP has become an attractive and privileged scaffold for the discovery of highly active pharmaceutical agents. One class of 2,5-DKPs, which have unsaturated C-C double bonds at the 3 and/or 6 positions of the 2,5-DKP ring, such as the marine natural products XR334 [18], piperafizine B [20], (±)-phenylahistin [21], and its synthetic derivative plinabulin [22], possess two pairs of hydrogen bond donors and acceptors ( Figure 1) and show markedly different anticancer properties. The 2,5-DKPs furnished with phenyl rings at the 3 and 6 positions, such as piperafizine B and XR334, assemble in line and/or net frameworks due to the formation of the intermolecular hydrogen bond as well as the π-π stacking interactions [5,13] (Figure 2). These compounds have very poor liposolubilities and weak anticancer activities, and further biological investigations into them are thus prevented [19]. However, those 2,5-DKPs with an imidazole (or 2-pyridyl) group as a side component, such as (±)-phenylahistin and plinabulin, have an intramolecular hydrogen bond formed preferentially between the amide hydrogen of the 2,5-DKP ring and the nitrogen atom of the side imidazole moiety [23] (Figure 1), and the tendency for the formation of the intermolecular hydrogen bonds is largely weakened. Their liposolubilities and anticancer activities are greatly increased [14]. The lipophilicities of this kind of compound can also be improved by the introduction of protective groups, similar to the semi-N-methylation of piperafizine B to A or XR334 to XR330, on the amide nitrogen atom of a 2,5-DKP ring. Previous studies demonstrated that these semi-N-protected 2,5-DKPs had improved liposolubilities and could cross the cell membrane easily, along with enhanced stable capacities away from the degradation caused by enzymes [6]. Furthermore, the enhanced lipophilicity of 2,5-DKPs was beneficial to their anticancer activity [24].  In our previous study, compound 4m showed good anticancer activities against several cell lines, including the A549 and Hela cell lines [13]. An allyl group was introduced to the 1nitrogen atom of the 2,5-DKP ring to interrupt the formation of the intermolecular hydrogen bonds. Compared to the methyl group, this function with a little longer side chain could also disturb the π-π stacking interactions between the layers of the frameworks, which then contribute to the good lipophilicities of the derivatives. On the other hand, the biological results showed that the 2-methoxy group (2-OMe) on the phenyl ring at the 6 position of the 2,5-DKP ring played a critical role in improving the anticancer activity, whereas the impacts of its two chlorine atoms and some other strong electron-withdrawing groups (e.g., NO 2 , CF 3 , and CN) on the anticancer activities of derivatives have not been clearly investigated, as electron properties sometimes exhibit significant (positive or negative) impacts on a derivative's bioactivity [25,26]. Similarly, the size or skeleton of the substituent will also have different influences on the anticancer activities of derivatives [7,27]. Therefore, based on the marine natural products piperafizine B and XR334, in this context, a small library of novel 2,5-DKPs was designed as follows (Scheme 1): (1) similar to compound 4m, the allyl group on the 1-nitrogen atom was retained; (2) the electron-withdrawing groups, such as NO 2 , CN, CF 3 , Br, Cl, and F on the phenyl group (series I), and the electron donating group OMe on benzene (series I or II) or naphthalene (series II) were introduced; and (3) a few electron-rich (e.g., furan and thiophene), electron-deficient (e.g., pyridine) heteroaromatic scaffolds, or fused naphthyl or quinolyl functions (series II) were employed. Thus, fourteen novel 2,5-DKPs (1, 2, 4-6, 8-16), along with two known ones (3 and 7), were designed and synthesized, and their anticancer activities against two cell lines, A549 and Hela, were evaluated aiming at the discovery of highly active anticancer agents. Scheme 1. Designing of the 2,5-DKPs in this study.

Synthesis of Compound
By following our previously reported procedure [28], the compounds were synthesized using a one-pot operation (Scheme 2). For the synthesis of compounds 1-5 and 8-16, in the first step, after the reaction of 1,4-diAc-2,5-diketopiperazine with the corresponding aldehyde and allyl bromide under base conditions at room temperature to yield the intermediate, another corresponding aldehyde was added directly in the second step, and the reaction was subsequently heated under base conditions until the disappearance of the intermediate. Of note is that the base additive used in two steps can be added together in the first step. The syntheses of compounds 6 and 7 were similar to the above procedure, whereas the corresponding aldehydes were added just once at the first step. The products have low to moderate yields ranging from 21 to 67%. Compound 11 was selected as a representative compound to verify the correct structure of this kind of product based on its NMR and HRMS results ( Figure 3 -12). Note that one alkene terminal proton signal is overlapped (δ H 7.56-7.50, H-15), and the other alkene terminal proton signal is at 7.30 for H-7. The δ H values ranging from 5.61 to 4.31 are signals of allyl, and the signal of the methoxyl group is at 3.87. The 13 C NMR spectrum showed that two amide carbon signals (δ C 159.8 and 158.9) and one aromatic carbon (C-13, adjacent to the methoxyl group) signal (δ C 157.3) are in the far low-field. There are twenty-one carbons signals (δ C 133. 8-110.6), which belong to the groups of naphthalene (10 carbons), benzene (5 carbons), and three C-C double bonds (6 carbons Figure S29 in Supplementary Materials), which indicates the correct stereo-configuration for the 3Z,6Z-isomer and is similar to those in the previous studies [12,28].  The solubility test demonstrated that all the compounds could be soluble in common organic solvents, such as dimethyl sulfoxide (DMSO), N,N-dimethylformamide (DMF), ethyl acetate (AcOEt), dichloromethane (DCM), chloroform, and acetone, with the solubilities > 1.0 mg mL −1 at room temperature.

Inhibiting Proliferation of Cancer Cells
First, we screened the antiproliferative effects of all compounds against the cancer cell lines A549 and Hela at a preliminary concentration of 10.0 µM. Those with inhibition >50% were selected for further detailed evaluations, and their IC 50 values were subsequently calculated and obtained (Table 1). It was found that compounds 8-11 showed moderate to good inhibitory capacities against cell line A549 (IC 50 = 1.2-7.3 µM), while compounds 6, 8-12, and 14 showed moderate to good activities against the Hela cell line (IC 50 = 0.7-8.9 µM). It seemed that the active compounds had a little stronger anticancer activity against Hela than A549. In comparison to compound 4m, the antiproliferative activities of the active compounds were not obviously improved, except for compound 11, which had nearly 1.9-fold activity (IC 50 = 1.2 µM) stronger than compound 4m (IC 50 = 2.3 µM) against the A549 cell line and around 2.3-fold activity (IC 50 = 0.7 µM) stronger than compound 4m (IC 50 = 1.6 µM) against the Hela cell line. The results showed that the compounds with electron-withdrawing groups, such as 3-NO 2 (1), 4-CN (2) or 2-Cl-5-CF 3 (3), on the phenyl group at the 3 position of the 2,5-DKP ring had weak activities (IC 50 > 10.0 µM). Similar to our previous study, when replacing 2-OMe with 3-Br (4), 2-CF 3 (5), 3-CF 3 (6) or 4-CF 3 (7) on the phenyl group at the 6 position of the 2,5-DKP ring, only compound 6 had moderate inhibitory activity against Hela cells (IC 50 = 8.9 µM). Of note is that these compounds also contain electron-withdrawing groups, including 3-NO 2 (4), 2,3-Cl (5), 3-CF 3 (6) or 4-CF 3 (7) on the phenyl group at the 3 position of the 2,5-DKP ring. These results imply that electron-withdrawing groups might be not suitable functions. Delightedly, when using electron-rich heteroaromatic cycles such as furan (8) and thiophene (9) or electron-deficient pyridine (10) as the substituents on the 3 position of the 2,5-DKP ring, compounds had moderate to good activities against both the A549 (IC 50 = 3.7-7.3 µM) and Hela (IC 50 = 4.7-5.9 µM) cell lines. Changing the substituent of 1-naphthalene (11) into 2-quinoline (12) or 6-methoxy-2-naphthalene (13), compound 12 displayed only moderate activity against Hela cells (IC 50 = 6.2 µM), while compound 13 showed no obvious activities toward either of the cancer cell lines. Surprisingly, compound 14, by taking the place of the 2-OMe with 2-CF 3 on the phenyl group at the 6 position of the 2,5-DKP ring, could also suppress the growth of the Hela cell line with good activity (IC 50 = 3.9 µM). However, by moving the 2-CF 3 from the 2 position to the 3 position (15) or replacing 2-OMe with 2-F (16), the activities of the compounds were all lost. Collectively, in terms of the above results, although compound 14 with a 2-CF 3 moiety shows good activity, other examples (compounds 1-7) indicate that the electron-withdrawing functions on phenyl groups at both sites (3 and 6 positions) of the 2,5-DKP ring might be not favorable. In addition to the vital role of the 2-OMe group, the substitutive motifs with suitable sizes or skeletons at the 3 position of the 2,5-DKP ring might be another critical factor for the 2,5-DKPs with high anticancer activities. It should be mentioned that multiple factors, such as the skeleton, size, type, and/or position of the substituent, might have combined impacts on the anticancer properties of these derivatives.
0.004 ± 0.2 0.004 ± 0.8 a All results are presented as mean ± standard deviation (SD) from three parallel experiments. b Paclitaxel as a positive control.
The results also showed that compound 11 could inhibit the growth of both cell lines in dose-dependent manners after treatment for 48 h, and this compound had a slightly stronger inhibitory activity against Hela than A549 ( Figure 4). In addition, the cytotoxicity of compound 11 on human ectocervical-vaginal epithelial cells (VK2/E6E7 cells) was examined, and the results showed that it had low toxicity to normal cells (only 30% inhibition at 10.0 µM), implying its value for further optimization.

Apoptosis in Cancer Cell Lines Induced by Compound 11
Previous studies demonstrated that 3,6-diunsaturated 2,5-DKPs with anticancer properties showed a phenotype characteristic of inducing apoptosis in cancer cells. For example, the synthetic plinabulin with a side imidazole component could induce apoptosis in multiple myeloma cells [29]. Our former synthetic compound 3c with lipidic side alkyl chain could also induce apoptosis in U937 cells with dose-and time-dependent relationships [30]. Therefore, the abilities of compound 11 to induce apoptosis in both cell lines were evaluated using flow cytometry with Annexin V-FITC and PI dual staining. As shown in Figure 5, compound 11 could obviously induce apoptosis in both cells at 1.0 µM after 48 h of treatment in comparison with the untreated controls. Note that cell death occurred in both early and late stages. Dose-dependent effects of compound 11 on the induction of apoptosis in both cells were observed, especially for Hela cells, indicating that Hela cells were more sensitive to this compound, which is consistent with the MTT results that compound 11 had a stronger inhibitory ability against Hela than A549. Surprisingly, necrosis characterized by a positive correlation between the amount and the treated concentrations of compound 11 in Hela cells was also found, implying this compound might simultaneously work through other unknown pathways to induce cell death, and similar results were also found in a former study that focused on a deuterated plinabulin [31].  4) containing DMSO (1‰) was used as a negative control, and paclitaxel was used as a positive control. All results were presented as mean ± standard deviation (SD). "ns" means no statistical differences; ** p < 0.01, *** p < 0.001, Student's t-test.

Cell Cycle Arrest Induced by Compound 11
Cell cycle arrest experiments were also performed to investigate which phase of the cell cycle progression was blocked by compound 11. As shown in Figure 6, compound 11 could arrest both cancer cells in the G2/M phases, and the cell contents increased following the concentration gradients. Compound 11 also showed a stronger potency to arrest the Hela cells than A549 cells, consistent with the above MTT and apoptosis results. Compared to the untreated controls, major peak distributions for the G2/M phases could be found when cells were treated with 0.5 µM, and the cell cycle progression could be almost blocked at the G2/M phase with percentages of 85.87 (A549) and 90.37% (Hela), respectively, when treated with 1.0 µM for 24 h. The results were similar to the cell cycles blocked by the positive control paclitaxel, indicating our derivative may have the same biological targets, such as tubulin, as paclitaxel [32], while their binding modes and regulating functions might be different. It was documented in previous studies that paclitaxel bound to the microtubule at the taxoid site on the luminal face of β-tubulin and maintained the stability and destroyed the dynamics of the microtubule, thus leading to mitotic catastrophe [33]. However, according to the "similarity principle" [34], our 2,5-DKPs might have the same mechanism as plinabulin or its analogs, which acted as a tubulin binder near or at the colchicine binding site located at the interfacial region of the αand β-tubulin studied using photoaffinity labeling [35,36] or computational modeling [37][38][39][40], leading to the inhibition of tubulin polymerization and the prevention of microtubule formation, thus resulting in the apoptosis of cancer cells. The detailed mechanism will be investigated in future studies. Figure 6. Compound 11 significantly induced G2/M phase arrest in both cancer cell lines compared to control. PBS (pH 7.4) containing DMSO (1‰) was used as a negative control, and paclitaxel was used as a positive control. All results were presented as mean ± standard deviation (SD). "ns" means no statistical differences; ** p < 0.01, *** p < 0.001, Student's t-test.

Synthesis of Compounds
General Procedure for the Synthesis of Products 1-16 1,4-diacetyl-2,5-diketopiperazine (50 mg, 0.25 mmol, 1.0 equiv.) and the first corresponding aldehydes (0.25 mmol, 1.0 equiv.) dissolved in 2 mL dry DMF were added into a 4 mL vial, followed by allyl bromide (54 µL, 0.63 mmol, 2.5 equiv.) and Cs 2 CO 3 (205 mg, 0.63 mmol, 2.5 equiv.). The reaction was stirred at room temperature until the completion of the 1st aldol condensation and the alkylation of the allyl bromide. Then, the second corresponding aldehydes (0.5 mmol, 2.0 equiv.) were added, and the mixture was heated at 95 • C for about 4 h. The reaction mixture was added to water (15 mL) and extracted with AcOEt (5 mL × 3). The organic layer was dried over Na 2 SO 4 , filtered, and removed. The residues were purified by flash column chromatography on silica to afford the target products 1-5 and 8-16. The procedure for the synthesis of compounds 6 and 7 was similar to the above synthetic method; however, the aldehydes (3.0 equiv.) were added just once in the first step. Compounds 3 [41] and 7 [28] are known, and their analytic data are identical to the reported values.  13 13 13 13 13

Conclusions
Sixteen 2,5-DKPs with different functions substituted at the 3,6 positions of the 2,5-DKP ring were designed, synthesized, and investigated as anticancer agents against cell lines A549 and Hela. 2-OMe on the phenyl group at the 6 position and the suitable size or skeleton substituents at the 3 position on the 2,5-DKP ring might have combined impacts on the anticancer activities of the derivatives. The one-pot method for the synthesis of these derivatives is simple and operationally convenient, which is also a guarantee of providing enough amounts of the active derivatives for further druggability assessments. The biological evaluation disclosed that the activities of most derivatives were not obviously improved except for compound 11, which is a promising skeleton for further development to explore highly active anticancer agents; therefore, our ongoing studies are now mainly focusing on this compound.