*2.2. Synthesis of 3-Acetylthioglycosides of 2-Methoxynaphthoquinones 13a–d–16a–d*

Then, four substituted 1,4-naphthoquinones—**5**, **6**, **11**, and **12**—were condensed in acetone with per-*O*-acetylated 1-thiosugars **2a**–**d** at equimolar ratio of quinone: thiosugar under base conditions in presence of K2CO3 according the procedure described in our previous work [22] (Scheme 1). This condensation led to acetylated thioglycosides of 1,4-naphthoquinones **13a**–**d**–**16a**–**d** in good yields, 71–91% (Figure 2). The structures of new compounds were proved by NMR, IR spectroscopy, and HR mass spectrometry. The 1- ,2- -*trans*-configuration of glycosidic bond was confirmed by the value of anomeric proton doublets (*J1*- *,2*- = 7.5–10.2 Hz) in 1H-NMR spectra. The other spectral characteristics of the naphthoquinone methoxyderivatives **13a**–**d**–**16a**–**d** were in a good agreement with their proposed structures (see also Supplementary Materials file).

**Figure 2.** Acetylated thioglycosides of substituted methoxy-1,4-naphthoquinones.

#### *2.3. Preparation of Naphthoquinone–Sugar Tetracyclic Conjugates*

Under the base treatment by MeONa/MeOH thioglycosides **13a**–**d**–**16a**–**d** were readily converted in tetracycles **17a**–**d**–**20a**–**d** in good yields 82–97% (Figure 3). It is evident that tetracyclic quinone–glucoside conjugates **17a**–**d**–**20a**–**d** were formed from methoxyglucosides **13a**–**d**–**16a**–**d** through deacetylation stage and intramolecular nucleophilic substitution of the methoxy group by sugar C2-OH group.

**Figure 3.** Synthesized 1,4-naphthoquinone-thioglycoside tetracyclic conjugates.

This process proceeds with retention of the configuration of all asymmetric centers of the carbohydrate portion and formation of linear tetracyclic structure. The alternative angular structure for the obtained tetracycles **17a**–**d**–**20a**–**d** was rejected based upon on the direct comparison with the spectral data of similar angular tetracycles, which we obtained earlier [30].

#### *2.4. Biological Evaluation*

#### 2.4.1. Cytotoxic Activity

Tetracyclic conjugates **17a**–**d**–**20a**–**d** and 5-hydroxy-1,4-naphthoquinone (juglone) were examined for cytotoxicity against three cancer cell lines and two normal cell lines such as human cervical cancer (HeLa), mouse neuroblastoma (Neuro 2a), mouse ascites Ehrlich carcinoma, mouse normal epithelial cell line (JB6 Cl 41-5a), and mouse blood erythrocytes. Known cytotoxic agent cucumarioside A2-2 [31] was used as positive control. The results are presented in Table 1.

Conjugates **17a**,**c**,**d** with a 7,10-dimethoxynaphthoquinone core were inactive for all cell lines at EC50 value ≤ 25 μM. Galactoside derivative **17b** had poor solubility in DMSO, which did not allow for the measurement its activity. Introduction of two chlorine atoms in a 7,10-dimethoxynaphthoquinone core led to tetracyclic 8,9-dichloroderivatives **18a**–**d** with better solubility and promising activity in EC50 values ranging from 1.1 to 10.9 μM. Among the substances of this group, galactoside derivative **18b** showed the best activity against ascites Ehrlich carcinoma cell line with EC50 value 1.1 μM.

The following two groups of tetracycles, **19a**–**d** and **20a**–**d**, are conjugates of thiosaccharides **2a**–**d** with derivatives of 5-hydroxy-1,4-naphthoquinone **11**–**12**. Among them, six substances had promising values, EC50 < 1 μM. In the group of tetracycles **19a**–**d**, bearing a hydroxyl group at position 10, only hexapyranoside derivatives d-Glc **19a** and d-Gal **19b** showed the high cytotoxic activity with EC50 values in the range of 0.6 to 0.9 μM, while all tetracycles **20a**–**d**, bearing hydroxyl group at position 7, revealed highly toxic compounds with EC50 0.3–0.7 μM for various types of cells. Mouse epithelial Jb6 cells were more susceptible to the action of juglone tetracyclic derivatives **20a**–**d**. As evidenced from Table 1, the presence of a hydroxy group in the naphthoquinone scaffold led to the formation of naphthoquinone tetracycles **19a**–**b**–**20a**–**b** with promising cytotoxicity. However, unsubstituted 5-hydroxy-1,4-naphthoquinone (juglone) did not show any cytotoxic activity up to 100 μM. This fact proves the positive effect of heterocyclization on tetracycles cytotoxicity. Moreover, it was shown that all tested compounds did not cause lysis of murine erythrocytes up to 25 μM.

For all cytotoxic compounds **18a**–**d**–**20a**–**d** their selectivity index (SI) was calculated (Table 2). Among the tested tetracycles, the most selective substance was **19a** in relation to all studied lines of tumor cells. In comparison to non-tumor mouse epithelial Jb6 Cl 41-5a cells, the selectivity index for

Ehrlich carcinoma cells was 9.3, for HeLa—1.5, and for Neuro-2a cell cuture—1.4. Compounds **18b** and **19d** also showed rather high values of the selectivity index for Ehrlich carcinoma cells with SI = 2.6 and SI = 2.1, respectively. On HeLa cells, the most active compounds were **19a**, **19d**, and **20d**, with SI > 1.


**Table 1.** Cytotoxicity (EC50) of oxathiine fused 1,4-naphthoquinone tetracycles on cancer and non-cancer cell lines (μM).

\* Cytotoxicity evaluation with MTT reagent; \*\* cytotoxicity evaluation with with FDA; \*\*\* non-tested due to poor solubility.

**Table 2.** Tumor cell selectivity (Selectivity Index (SI)) of tested tetracycles **18a**–**d**–**20a**–**d**.

