Synthesis of 6-Halo-Substituted Pericosine A and an Evaluation of Their Antitumor and Antiglycosidase Activities

The enantiomers of 6-fluoro-, 6-bromo-, and 6-iodopericosine A were synthesized. An efficient synthesis of both enantiomers of pericoxide via 6-bromopericosine A was also developed. These 6-halo-substituted pericosine A derivatives were evaluated in terms of their antitumor activity against three types of tumor cells (p388, L1210, and HL-60) and glycosidase inhibitory activity. The bromo- and iodo-congeners exhibited moderate antitumor activity similar to pericosine A against the three types of tumor cell lines studied. The fluorinated compound was less active than the others, including pericosine A. In the antitumor assay, no significant difference in potency between the enantiomers was observed for any of the halogenated compounds. Meanwhile, the (−)-6-fluoro- and (−)-6-bromo-congeners inhibited α-glucosidase to a greater extent than those of their corresponding (+)-enantiomers, whereas (+)-iodopericosine A showed increased activity when compared to its (−)-enantiomer.


Introduction
Chemical modification of bioactive natural products is one of the most common methods used in drug development [1][2][3][4][5][6][7][8]. In addition to increasing the pharmacological activity, which is the major aim of chemical modification, decreasing side effects, improving the solubility or stability, and reducing costs remain problems to be resolved during drug development. Chemical modification based on marine natural products has been extensively studied due to the discovery of new lead compounds suitable for drug discovery [9][10][11][12].
The isolation of pericosine A (1 Cl ) and B (2) as metabolites of the marine-derived fungus Periconis byssoides N133 was first reported in 1997, and pericosines C-E (3)(4)(5)(6) were discovered in 2008 [13][14][15]. The unique carbasugar structures constituting the highly functionalized cyclohexene ring are shown in Figure 1. Recently, the chemistry of carbasugars has received an increasing amount of research attention due to their wide range of biological activity [16][17][18]. In addition, many total syntheses of pericosines A-C (1-3) have been reported due to the antitumor activity of pericosine A (1 Cl ) [19][20][21][22][23][24][25][26][27][28][29][30]. However, the chemical modification of the pericosines has not been reported to date with the exception of our synthetic study on pericosine E analogs bearing chloro-or methoxy-substituents at C-6, which were used as α-glucosidase inhibitors [31,32]. Figure 1 also includes (+)pericoxide (7) and (−)-maximiscin, which were discovered by Cichewitcz and coworkers from Tolypocladium sp. [33,34]. The fact that pericosine 1 Cl was obtained together with pericoxide 7 in their work suggests that 7 may be a biosynthetic precursor of 1 Cl . Therefore, 7 should be classified as a member of the pericosine family. In our recent report on pericosine A, natural 1 Cl was isolated from Periconia sp. as an enantiomeric mixture. In addition, we confirmed that both synthesized enantiomers of 1 Cl exhibit antitumor activity with similar potency against three tumor cell lines (P388, LH-6, and L1210). Furthermore, (−)-1 Cl showed moderate α-glucosidase inhibitory activity (IC 50 = 2.25 mM), whereas (+)-1 Cl was inactive [35]. As we are interested in the biological activities of pericosine congeners bearing other halogen atoms, the fermentation of Periconia sp in artificial seawater containing fluoride, bromide, or iodide sources has been examined as an alternative to chloride. It should be mentioned here that 6-halogenated pericosines bearing F, Br, or I atoms were not obtained in our preceding paper. Consequently, we have attempted the synthesis of non-natural 6-halo-substituted pericosine A. Our recent synthetic work on pericosine E analogs has suggested that the presence of a chlorine atom at C6 is an important factor for α-glucosidase inhibitory activity. The preparation of other 6-halo-congeners is also required for our continuous synthetic studies on new pericosine E analogs.
Herein, we describe the synthesis and evaluation of the antitumor and antiglycosidase inhibitory activities of both enantiomers of newly designed 6-halo-congeners of pericosine A. In addition, an efficient synthesis of closely related pericoxide 7 via 6-bromo pericosine A (1 Br ) has also been reported.

Synthesis of Both Enantiomers of 6-Halo-Substituted Pericosine A
A similar reaction to the hydrochlorination of epoxide intermediate 8 used in the synthesis of pericosine A 1 Cl [21,29] using commercially available HBr and HI aqueous solutions was envisioned in this study, in which the deprotection of the hydroxyl groups and hydrolysis of the ester moiety will occur (Scheme 1). Subsequently, we began this work by searching for suitable reagents used to introduce the required fluorine, bromine, or iodine atoms. Scheme 1. Synthesis of 6-halogenated pericosine A. (a) Synthesis of (-)-1 from (-)-shikimic acid; (b) Synthesis of (+)-1 from (-)-quinic acid.
For bromination, BBr 3 was initially examined as the bromide source to react with epoxide (−)-8, which can be prepared from bromohydrin (+)-9 via an intramolecular S N 2 reaction [29,31]. However, careful addition of BBr 3 [0.33 equivalents (eq)] to (−)-8 in dry diethyl ether (Et 2 O) at −78 • C afforded the desired product (10 Br ) in only 14% yield. Using 1.0 eq of BBr 3 slightly improved the yield (38%), but increasing the reaction temperature led to the formation of a complicated mixture including the undesired regioisomer. After investigating a variety of brominating reagents, we found mono-bromoborane dimethyl sulfide complex (BH 2 Br·SMe 2 ) to be the most suitable for the desired reaction [36,37]. The reaction of 8 with 1.0 eq of BH 2 Br·SMe 2 at −78 • C in Et 2 O dramatically improved the yield of (−)-10 Br to 94% yield. The reactions performed at higher temperatures led to a decreased yield of (−)-10 Br (78% at −20 • C and 58% at 0 • C). Subsequent deprotection of the cyclohexylidene moiety was also a delicate process. The optimum reaction conditions were found after investigating a variety of conditions. Treatment of (−)-10 Br with Dowex ® 50WX8 hydrogen form (Dowex ® 50WX8-H: Acidic ion-exchange resin) in MeOH at room temperature (rt) for 56 h gave (−)-1 Br in an excellent 87% yield when compared to the conventional reaction using trifluoroacetic acid (TFA) in MeOH (66%) (Scheme 1A). This deprotection process with Dowex-XW50-H was examined in detail because could also be applied in the subsequent synthesis of pericoxide 7 (see Supplementary Material, Table S1).
The introduction of a fluorine atom into (−)-8 was accomplished using the (HF) n /py complex. The reaction of (−)-8 with this complex at 0 • C for 15 min in a polypropylene tube afforded the desired fluorohydrine product [(−)-10 F ] in 46% yield. A longer reaction time (1 h) led to a more complex product mixture, giving a low yield of (−)-10 F (32%). Subsequent treatment of (−)-10 F with TFA in MeOH afforded (−)-1 F in 43% yield. Similarly, (+)-1 F was successfully obtained using (+)-8. It should be noted that 10 I and 1 I are relatively unstable when compared to the other halogenated congeners, so they were stored in a freezer prior to further use.

Synthesis of Pericoxide
The synthesis of 7, which was not reported prior to 2019, was also conducted, as shown in Scheme 2. The direct deprotection of epoxide 8 was examined in our preliminary efforts to prepare 7. Treatment of (−)-8 with trifluoroacetic acid in t-BuOH at room temperature (rt) did not afford pericoxide 7 despite the complete consumption of 8. Changing the acid catalyst to Dowex ® 50WX8-H resulted in no reaction and microwave (MW) heating afforded a small amount of methyl 3,4-dihydroxybenzoate along with the recovery of the starting material [(−)-8]. In our next experiment, bromotriol 11 derived from bromohydrin 9 was treated with 3.0 eq of lithium hexamethyldisilazide (LHMDS) in THF at −78 • C because if the C6 hydroxide anion attacked the C5 center faster than the C4 hydroxide anion, the formation of 7 via an intramolecular S N 2 process was expected to occur. However, the spectral data of epoxide 12 did not agree with those of 7. The structure of 12 was determined using NMR spectroscopy; epoxy carbon atoms C6 (δ 55.5 ppm) and C1 (δ 56.0 ppm) were detected in the high field region in the 13 C-NMR spectrum and the HMBC cross peaks corresponding to H1/C3 and H6/C4 indicated the structure of 12 (see Supplementary Materials).
Finally, (−)-bromopericosine A (1 Br ) was treated with 3.0 eq of LHMDS at −78 • C to give (−)-7, which is an enantiomer of the natural product. The isolated yield of (−)-7 upon purification via conventional acidic silica gel column chromatography was only 23%, but this was improved to 77% using neutral silica gel. The spectral data agree with those of natural pericoxide with the exception of the sign of its specific rotation. Unfortunately, (−)-7 was so unstable that it decomposed upon storage at rt in methanol, exhibiting a smaller optical rotation within a couple of days. Furthermore, the decomposed residue exhibits a positive specific rotation, although freshly synthesized 7 was negative. Regrettably, biological assays of 7 could not be performed because of this inherent instability.

Antitumor Assay
The antitumor activity of halo-compounds 1 was evaluated against three types of tumor cell lines: Basic P388 (mouse lymphocytic leukemia), L1210 (mouse lymphocytic leukemia), and LH60 (human promyelocytic leukemia) cell lines, along with a previously reported procedure using 5-fluorouracil (5-FU) as a positive control [35]. The results are presented in Table 1, including those obtained for perocosine A (1 Cl ), which has been previously reported in the literature, for comparison. All compounds showed antitumor activity against the three types of tumor cell lines studied. Bromo-and iodo-pericosine (1 Br and 1 I ) show similar activities to 1 Cl , but fluorinated compound 1 F was less active than all of the other compounds, including 1 Cl . Because pericosine C (3), which exists as an enantiomeric mixture in nature and has the same relative configuration to 1s, was reported to be inactive against P-388 cell line [14], present results implied the importance of the presence of halogen atom at C-6 in pericosine core structure for antitumor activity. In addition, it is noteworthy that no significant difference in the potency was observed between the enantiomers, similar to pericosine A.

Biological Assay
Antitumor and glucosidase inhibitory assays were performed using the same procedures as those described in our previous paper [35].

Conclusions
The synthesis of both enantiomers of pericosine A analogs bearing F, Br, and I atoms was achieved for the first time and their antitumor activity against P388, L1210, and HL-60 cell lines was evaluated. Although all of the synthesized compounds were moderately active against the three types of tumor cell lines studied, significant differences between their enantiomers and differences between the halogens, except for fluorine, were not observed. The fluorinated derivatives showed weaker activities than the other analogs and pericosine A.
In addition, both enantiomers of pericoxide were synthesized using 6-bromopericosine A as a suitable synthetic precursor.

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