Design, Synthesis, and Mechanism of Dihydroartemisinin–Coumarin Hybrids as Potential Anti-Neuroinflammatory Agents

Cancer patients frequently suffer from cancer-related fatigue (CRF), which is a complex syndrome associated with weakness and depressed mood. Neuroinflammation is one of the major inducers of CRF. The aim of this study is to find a potential agent not only on the treatment of cancer, but also for reducing CRF level of cancer patients. In this study, total-thirty new Dihydroartemisinin–Coumarin hybrids (DCH) were designed and synthesized. The in vitro cytotoxicity against cancer cell lines (HT-29, MDA-MB-231, HCT-116, and A549) was evaluated. Simultaneously, we also tested the anti-neuroinflammatory activity of DCH. DCH could inhibit the activated microglia N9 release of NO, TNF-α, and IL-6. The docking analysis was shown that MD-2, the coreceptor of TLR4, might be one of the targets of DCH.


Introduction
Chemotherapy is an important method in the therapy of cancer, but the adverse side effects of anticancer agents should also be focused on. Gastrointestinal injury induced by chemotherapeutic agents could result in bacterial/endotoxin translocation from the gut to the systemic blood circulation [1,2]. The released endotoxin would cause a systemic inflammatory response including neuroinflammation [3]. Cancer patients frequently suffer from cancer-related fatigue (CRF), which is a complex syndrome associated with weakness and depressed mood. Diana M., Norden et al. demonstrated that the neuroinflammation is one of the major reasons which induce CRF [4]. Currently, there were no effective method to reduce CRF [5]. The aim of this study is to find a potential anti-neuroinflammatory agents which might reduce CRF of cancer patients.
Central nervous system (CNS) consists of two major types of cells, nerve cells and glial cells. Microglia is a major glial cell and plays an important role as the resident macrophage population in the CNS [6,7]. There has been wide recognition that the fully activated microglial cells are neurotoxic. They release reactive oxygen species, nitric oxide (NO), tumor necrosis factor-alpha (TNF-α), and other potentially cytotoxic molecules [8,9]. It means that activation of microglia could induce neuroinflammation.
Lipopolysaccharide (LPS), also known as endotoxin, is a natural Toll-like receptor 4 (TLR4) ligand and it is also a potent microglial activator that can trigger the production of inflammatory mediators. TLR-4, a transmembrane receptor, plays a vital role in the initiation and acceleration of the inflammatory response induced by LPS [10,11]. TLR4 alone does not directly bind LPS and needs the coreceptor MD-2 [12]. MD-2 is linked with the extracellular domain of TLR4 and is indispensable for LPS recognition [13]. MD-2 is a member of the MD-2-related lipid-recognition protein family [14]. MD-2 directly binds to LPS in its hydrophobic cavity with high affinity [15].
Artemisinin and its derivatives have been the first-line antimalarial drug since the late 1990s. Dihydroartemisinin (DHA) is the primary bioactive metabolite of artemisinin and its derivatives, but not artemisinin per se [16]. Additionally, DHA has higher bioavailability and less side effects than artemisinin [17]. Recent research showed that besides antimalarial activity, DHA also exhibits anticancer and anti-inflammatory activity [18][19][20][21]. As the anti-inflammatory agent, DHA could decrease LPS-induced protein expression of TLR4 [22].
In our previous study, Dihydroartemisinin-Coumarin hybrids (DCH) were designed and synthesized referring to the principle of hybridization [34,35]. Each of them contains coumarin moiety, DHA moiety and a linker. DCH could inhibit HT-29 cells proliferation and arrest the G0/G1 phase of HT-29 cells [35]. They could also suppress cancer cell migration, and induce apoptosis and ferroptosis of HT-29 cancer cells [35]. It also has been confirmed that the DCH having 1,2,3-triazole linker generally exhibit better cytotoxicity exhibition than the others [35].
With this background, total-thirty new DCH were designed and synthesized, and the synthetic approaches were shown in Figure 1. All of them contained 1,2,3-triazole linker. We focused on the modifications of positions 3 and 4 of the coumarin moiety. Due to the new synthesized DCH were for the cancer patients, we evaluated the in vitro cytotoxicity against cancer cell lines (HT-29, MDA-MB-231, HCT-116, and A549). In addition, DCH could inhibit the activated N9 cell release of NO, TNF-α and IL-6. Furthermore, to investigate the mechanism of the anti-neuroinflammatory activity, docking analysis was applied. The results might support that the anti-neuroinflammatory activity was related to inhibiting TLR4/MD-2 complex.

Chemistry
The synthesis of the intermediates a-p and I-IV is presented in Scheme 1. The intermediates I and II were obtained by 2-bromoethanol or 3-bromo-1-propanol, with dihydroartemisinin (DHA), respectively, followed by treating the product with sodium azide. The intermediates III and IV were synthesized by propargyl alcohol or 3-butyn-1-ol with DHA, respectively. The required 3methylcoumarin derivatives were obtained by substitution of salicylaldehyde with propionic anhydride in the presence of sodium propionate and triethylamine. The substituted-3-
The synthesis of the desired new compounds was presented in Scheme 2. The target compounds 1a-1e and 2a-2e were obtained by the reaction of the intermediates III or IV with a-e via click chemistry, respectively. The target compounds 3f-3i and 4f-4i were synthesized by intermediates III or IV with f-i via click chemistry, respectively. The target compounds 5j, 5l-5o, and 6j-6p were prepared by the reaction of the intermediates I or II with j-p via click chemistry, respectively.

In Vitro Cytotoxic Activity of the Compounds
As the anti-neuroinflammatory agents for cancer patients, all newly synthesized compounds were evaluated in vitro for antitumor activity against human breast cancer cell line MDA-MB-231, colorectal cancer cells HT-29 and HCT-116, and lung cancer cell line A549 by using the standard MTT method, respectively. Doxorubicin (DOX) and DHA were used as the reference standard. Additionally, owing to the rapid proliferation, the tumor cells actually was always in hypoxia condition. Therefore, the in vitro cytotoxicity was evaluated in two different conditions (hypoxia and Scheme 2. Synthetic routes of target compounds. Reagents and conditions: (1) CuI, triethylamine, dichloromethane, r.t.

In Vitro Cytotoxic Activity of the Compounds
As the anti-neuroinflammatory agents for cancer patients, all newly synthesized compounds were evaluated in vitro for antitumor activity against human breast cancer cell line MDA-MB-231, colorectal cancer cells HT-29 and HCT-116, and lung cancer cell line A549 by using the standard MTT method, respectively. Doxorubicin (DOX) and DHA were used as the reference standard. Additionally, owing to the rapid proliferation, the tumor cells actually was always in hypoxia condition. Therefore, the in vitro cytotoxicity was evaluated in two different conditions (hypoxia and normoxia). All data are summarized in Table 1. Most of the target compounds exhibited moderate activities against the tested cell lines and they showed a general promotion under hypoxia conditions.

Morphological Changes and Effect of Compound 5n on Cell Viability
As shown in A-F of Figure 2, there showed a concentration-dependent cell death after treatment with compound 5n for 48 h. As the concentration up to 100 μM, the compound 5n could obviously inhibit the proliferation of HT-29. As shown in the (G) of Figure 2, after HT-29 cells were treated with different concentrations of 5n for 48 h, the cell viability decreased in a dose-dependent manner.

Anti-Neuroinflammatory Activity
The anti-neuroinflammatory activity of DCH was evaluated and then the representative compounds were selected for the further study. As shown in Table 2, murine microglial N9 cells were treated with 1μg/mL LPS for 48 h and the NO level in the medium was measured as the preliminary screening. Most of the tested compounds exhibited good anti-inflammatory activity, and especially the IC50 of some compounds were even less than 1 μM. Among them, compound 5n had the best activity and the IC50 reached 0.22 μM. In addition, in order to exclude the death influence, MTT-IC20

Anti-Neuroinflammatory Activity
The anti-neuroinflammatory activity of DCH was evaluated and then the representative compounds were selected for the further study. As shown in Table 2, murine microglial N9 cells were treated with 1µg/mL LPS for 48 h and the NO level in the medium was measured as the preliminary screening. Most of the tested compounds exhibited good anti-inflammatory activity, and especially the IC 50 of some compounds were even less than 1 µM. Among them, compound 5n had the best activity and the IC 50 reached 0.22 µM. In addition, in order to exclude the death influence, MTT-IC20 of all the compounds were measured. The results showed that DCH did not exhibit cytotoxic activity in N9 cells. The exact reason was still unclear, but we supposed that it might be related to the homeostasis of cells. The tumor cells actually were always in hypoxia condition, in which DCH could promote the cytotoxic activity against tumor cells. In brief, DCH could reduce the release of NO caused by LPS-induced microglial cells.
3.36 7.13 5m 1.08 NO Capture Capability Sodium nitroprusside (SNP) is a NO donor. In this experiment, the reason of the DCH reducing NO level was preliminary explored. As shown in Figure 3, the results showed that the compounds only partially reduced NO levels by direct NO capture. That means the main reason of the NO level reduction is due to species-specific effect. only partially reduced NO levels by direct NO capture. That means the main reason of the NO level reduction is due to species-specific effect. After reacted for 1 h under normal sun light at room temperature, Griess reagent was added and the optical density was observed at 540 nm. Data were shown as mean ± SD. *** p < 0.001 compared with control group.
Effect of Representative Compounds on TNF-α and IL-6 Release As shown in Table 3, four kinds of DCH were involved to measure the effect on the release of TNF-α and IL-6 from the microglia cells which induced by LPS for 48 h. The results showed that compounds 2b, 2c, 5n, and 6p could inhibit the release of TNF-α with the inhibition rate of 22% to Effect of Representative Compounds on TNF-α and IL-6 Release As shown in Table 3, four kinds of DCH were involved to measure the effect on the release of TNF-α and IL-6 from the microglia cells which induced by LPS for 48 h. The results showed that compounds 2b, 2c, 5n, and 6p could inhibit the release of TNF-α with the inhibition rate of 22% to 32% and the release of IL-6 with the inhibition rate of 26% to 43%. The results indicated that the inhibition effect of DCH on NO release was not only by nonspecific effects, but also by slightly influencing the synthesis and release of upstream inflammatory pathway.

Docking Analysis
Toll-like receptor 4 (TLR4) and its coreceptor-MD-2-are the target of LPS, which could influence innate immune response to release potentially cytotoxic molecules. MD-2 is linked with the extracellular domain of TLR4 and is indispensable for LPS recognition [13]. Therefore, this docking analysis was applied to explore whether or not the target compounds inhibit MD-2 to reduce the release of inflammatory mediators. Umeharu Ohto et al. [39] had revealed the crystal structure of mouse TLR4/MD-2/LPS complex, which was shown in Figure 4. In their study, MD-2 directly bound to LPS in its hydrophobic cavity with high affinity. With this background, compound 5n was used as the ligand to explore the relationship between DCH and TLR4/MD-2/LPS complex. As shown in Figure 5, compound 5n could integrate closely with the hydrophobic cavity of MD-2, which is the important TLR4/MD-2/LPS binding position. That means the DCH might disturb the interactions between LPS and TLR4/MD-2 complex. The minimum binding energy of (∆G) is −10.08 kcal/mol. Given to this result, we supposed TLR4/MD-2 complex is likely to be one of the drug targets of DCH.

Conclusions
Total-thirty new DCH were designed and synthesized. Most of the target compounds exhibited moderate activities against the tested cell lines and they showed a general promotion under hypoxia conditions. Meanwhile, the new potential activity of DCH was found that DCH could exhibit anti-neuroinflammatory activity. DCH could inhibit the release of NO, TNF-α, and IL-6, which were induced by LPS. The docking analysis was shown that MD-2, the coreceptor of TLR4, might be one of the targets of DCH. Additionally, DCH could also reduce NO levels by direct NO capture. In brief, DCH were the potential anti-neuroinflammatory agents which might reduce CRF of cancer patients.

Chemistry
Unless otherwise specified, all solvents and reagents were commercially available and used without further purification. Analytical TLC was performed using silica gel precoated GF254 plates. The 1 H-NMR and 13 C-NMR spectra were recorded at 600 MHz and 150 MHZ on a Bruker AV-400 spectrometer (Bruker Bioscience, Billerica, MA, USA), using CDCl 3 as solvent with tetramethylsilane as the internal standard. NMR spectra were analyzed using MestReNova. High-resolution mass spectra (HRMS) were measured using an Agilent Accurate-Mass Q-TOF 6530 (Agilent, Santa Clara, CA, USA) instrument in ESI mode. The 1 H-NMR and 13 C-NMR spectra of synthesized compounds, please see the supplementary materials.

General Procedure for the Synthesis of Compounds 1a-1e and 2a-2e
Substituted coumarin (a-e) (10 mmol), substituted DHA (III or IV) (10 mmol), and triethylamine (10 mmol) were dissolved in 20 mL CH 2 Cl 2 . In addition, CuI (100 mg) was added and the reaction mixture was stirred at room temperature for 8 h under the protection of nitrogen. The mixture was filtered, washed with water, dried over Na 2 SO 4 and evaporated to dryness. The crude product was purified through column chromatography.  13 13 13 13 13 13 13 13 13    Substituted coumarin (f-i) (10 mmol), substituted DHA (III or IV) (10 mmol) and triethylamine (10 mmol) were dissolved in 20 mL CH 2 Cl 2 . In addition, CuI (100 mg) was added and the reaction mixture was stirred at room temperature for 8 h under the protection of nitrogen. The mixture was filtered, washed with water, dried over Na 2 SO 4 and evaporated to dryness. The crude product was purified through column chromatography.  13 13 13

Conflicts of Interest:
The authors declare no conflict of interest.