Keywords

4.1 Introduction

(+)-Siladenoserinol A (1) is a marine natural product extracted from a tunicate of the family Didemnidae sampled from North Sulawesi in Indonesia. It contains a 6,8-dioxabicyclo[3.2.1]octane skeleton and two long side chains, one of which terminates in sulfamated serinol, and the other terminates in a glycerophosphocholine moiety [1]. Siladenoserinol A inhibits the protein–protein interaction (PPI) between p53 and Hdm2 that downregulates p53. As the upregulation of Hdm2 in cancer cells induces the inhibition of cancer suppressive p53, this PPI inhibitor is expected to be a tumor suppressive agent [2,3,4]. The study of the total synthesis of siladenoserinol A can be applied to its derivatives. It may also contribute to the elucidation of its mechanism of action and structure–activity relationship (SAR) and hence may lead to the discovery of a new anticancer drug. This chapter describes a new method for synthesizing the 6,8-dioxabicyclo[3.2.1]octane skeleton using the Au(III)-catalyzed cycloisomerization of dihydroxyalkynoate and its application in the total synthesis of siladenoserinol A [5].

4.2 Synthetic Plan of (+)-Siladenoserinol A (1)

The SAR of twelve siladenoserinols A–L has been previously reported. Based on this previous study, three substructures, namely the 6,8-dioxabicyclo[3.2.1]octane skeleton, sulfamated serinol, and glycerophosphocholine, are expected to play a major role in the biological activity of the siladenoserinols (Tables 4.1 and 4.2) [1]. Therefore, a convergent synthetic approach combining these three substructures was planned.

Table 4.1 Structures of siladenoserinols A–G and their IC50 values
Full size table
Table 4.2 Structures of siladenoserinols H–L and their IC50 values
Full size table

The initial synthetic plan is as follows. The glycerophosphocholine-containing side chain of 1 can be constructed by the esterification of α,β-unsaturated carboxylic acid 2 with glycerophosphocholine 3. The selective formation of the sulfamate group in the serinol moiety can be carried out in the final step. Acid 2 can be synthesized via the Julia–Kocienski olefination of aldehyde 4 with 1-phenyl-1H-tetrazole sulfone (PT-sulfone) 5, followed by the Horner–Wadsworth–Emmons (HWE) reaction with phosphonoacetate 6 after the conversion of the benzyloxyalkyl group into an aldehyde. This strategy involves stepwise side chain elongation in the late stage of the synthesis and could be suitable for the synthesis of the analogs of 1. As the formation of the bicyclic acetal moiety in 4 is challenging, we planned the Lewis acid-catalyzed cyclization of 7,8-dihydroxytetradec-2-ynoate 7 (Fig. 4.1).

Fig. 4.1
A complex chemical synthesis process leading to the formation of a compound labeled 7 has various reactions leading to different intermediate compounds. Some compounds involved are labeled siladenoserinol A, 2, 3, 4, 5, and 6. These compounds have open carbon chains with various functional groups.

Retrosynthesis of (+)-siladenoserinol (1)

Full size image

4.3 Model Study for the Synthesis of a 6,8-Dioxabicyclo[3.2.1]octane Skeleton

To synthesize the above bicyclic acetal, the cycloisomerization of 7,8-dihydroxy-2-alkynoate 8 was investigated (Table 4.3). Initially, 5 mol% of PdCl2(MeCN)2, which is often used for the hydroalkoxylation of alkynes [6], was used in MeCN to obtain the desired bicyclic acetal 9 in 52% yield (entry 1); however, this reaction required 12 h to complete. To improve the yield, a gold catalyst was used as a soft Lewis acid. Gold catalysts are not only capable of coordinating alkynes but also of catalyzing the intramolecular hydroalkoxylation of electron-deficient alkynes [7]. When AuClPPh3 was used as the catalyst, the reaction did not proceed, and the starting material was recovered (entry 2). However, when a cationic Au(I) complex prepared from AuClPPh3/AgOTf was used, the reaction was completed in 9 h, and 9 was obtained in 69% yield (entry 3). Surprisingly, when Au(III) chloride was used as the catalyst, the reaction was completed in 5 min, and the desired compound 9 was obtained in the highest yield of 79% (entry 4). When the amount of the catalyst was reduced to 1 mol%, 9 was obtained in almost the same yield; however, the reaction required 45 min to complete (entry 5). Both Au(III) and Au(I) can activate alkynes as soft Lewis acids, but Au(III) is known to exhibit a higher oxygen affinity than Au(I) [8]. Therefore, it is conceivable that Au(III) can activate both the alkyne and the carbonyl group as a Lewis acid and can increase the electrophilicity of the β-position of the alkynoate, thereby forming the desired bicyclic acetal 9 (Fig. 4.2). In contrast, the reaction did not proceed when AgOTf, AlCl3, p-TsOH, and HCl/dioxane were used (entries 6–9, respectively).

Table 4.3 Investigation of bicyclic acetal formation
Full size table
Fig. 4.2
A reaction presents the conversion of 7, 8 dihydroxy, 2, alkynoate to a bicyclic acetal using P d C l 2, M e C N 2. An alternative route presents formation of the bicyclic acetal via dual-activation in the presence of 6 exo and 5 exo.

Mechanism for the formation of bicyclic acetal 9

Full size image

4.4 Synthesis of Aldehyde 4 and PT-Sulfone 5

The preparation of 7 was carried out as follows. After the lithiation of terminal alkyne 11 prepared from D-malic acid (10) through a four-step transformation, its alkylation with benzyl 6-bromohexyl ether afforded 12 in 88% yield. Subsequently, 12 was converted into a trans-alkenediol by removing the benzylidene acetal moiety under acidic conditions and reduction with Red-Al® [9]. The free diol was then converted into benzylidene acetal again, and its regioselective cleavage using DIBAL afforded 13 [10]. The Sharpless asymmetric dihydroxylation of trans-alkene 13 [11] was followed by acetonide formation, which yielded 14 as a single diastereomer. Next, the primary hydroxy group in 14 was tosylated, and its alkylation was performed using lithium acetylide-ethylenediamine; this reaction afforded 15 in 95% yield. Subsequently, 15 was lithiated and added to methyl chloroformate. The removal of the acetonide under acidic conditions afforded the desired cyclization precursor 7 (Scheme 4.1).

Scheme 4.1
A complex chemical synthesis process leading to the formation of a compound labeled 7 has various reactions leading to different intermediate compounds. Some compounds involved are labeled D-malic acid, 11, 12, 13, 14, and 15. These compounds mostly have open carbon chains with various functional groups.

Synthesis of cyclization precursor 7

Full size image

Aldehyde 4 was then synthesized from 7. Cycloisomerization of 7 proceeded rapidly in the presence of 5 mol% of AuCl3 in MeCN at 30 °C leading to the desired product 16 in 83% yield. The stereochemical configuration of 16 was determined by the NOE correlation between the two hydrogen atoms, Ha and Hb, in the 6,8-dioxabicyclo[3.2.1]octane skeleton. Partial reduction of ester 16 using DIBAL was performed at − 78 °C to provide aldehyde 4 in quantitative yield (Scheme 4.2).

Scheme 4.2
A complex chemical synthesis process leading to the formation of a compound labeled 4 has various reactions leading to different intermediate compounds. Some compounds involved are labeled 7 and 16. The catalysts are A u C l 3, DIBAL, M e C N, and C H 2 C L 2.

Synthesis of aldehyde 4

Full size image

After the successful synthesis of aldehyde 4, PT-sulfone 5 was prepared as follows: L-Serine was converted into 17 in four steps using a previously reported method [12]. Etherification of primary alcohol 17 with 12-bromododecyl benzyl ether afforded 18 in 60% yield. The benzyl group in 18 was removed by hydrogenolysis, and its Mitsunobu reaction with 1-phenyl-1H-tetrazole-5-thiol (PT-SH, 19) was performed [13]. The ammonium molybdate-catalyzed oxidation of the resultant sulfide furnished PT-sulfone 5 in a good yield [14] (Scheme 4.3).

Scheme 4.3
A complex chemical synthesis process leading to the formation of a compound labeled 5 has various reactions leading to different intermediate compounds. Some compounds involved are labeled L serine, 17, and 18.

Synthesis of PT-sulfone 5

Full size image

4.5 Total Synthesis of (+)-Siladenoserinol (1)

4.5.1 Approach to 1 via Esterification of 2 with 3

A serinol moiety was introduced by the Julia–Kocienski olefination of aldehyde 4 with PT-sulfone 5 [15, 16]. Removal of the two benzyl groups and concomitant hydrogenation of the alkene moiety produced diol 20. Selective oxidation of the primary hydroxy group in 20 with TEMPO/PhI(OAc)2 yielded aldehyde 21. The HWE reaction of 21 using phosphonoacetate 6 [17], acetylation of the free secondary hydroxy group, and removal of the allyl group using Pd(PPh3)4/phenylsilane [18] afforded carboxylic acid 2 (Scheme 4.4).

Scheme 4.4
A complex chemical synthesis process leading to the formation of a compound labeled 2 has various reactions leading to different intermediate compounds. Some compounds involved are labeled 4, 20, 21, and 6. These compounds mostly have open carbon chains with various functional groups.

Conversion of aldehyde 4 to carboxylic acid 2

Full size image

Next, the esterification of the obtained α,β-unsaturated carboxylic acid 2 with the pre-prepared glycerophosphocholine 3 [19] was attempted. However, the desired esterification did not proceed under any of the following conditions: (a) DIC, DMAP, DMF; (b) PyBroP, DIEA, DMAP, DMF [20]; (c) triphosgene, DIEA, DMAP, 2,4,6-collidine, DMF [21]. A few studies have reported on the esterification of α,β-unsaturated carboxylic acids with glycerophosphocholine; the yields of these reactions were poor, and the isomerization of the alkene moiety was observed [22, 23]. In addition to the low reactivity of α,β-unsaturated carboxylic acid 2 in the condensation, the low solubility of glycerophosphocholine 3 made it difficult to synthesize 22a by esterification (Scheme 4.5).

Scheme 4.5
A complex chemical synthesis process leading to the formation of a compound labeled 22 a has with reactants labeled 2 and 3. These compounds mostly have open carbon chains with various functional groups.

Attempted coupling of carboxylic acid 2 with glycerophosphocholine 3

Full size image

4.5.2 Next Approach to 1

As the condensation of 2 and 3 did not proceed, we next focused on the introduction of a glycerophosphocholine precursor by the HWE reaction. Based on a previous study [24], phosphonoacetate 23 could be used as a precursor of glycerophosphocholine. After the HWE reaction of aldehyde 24a with 23, the bromine atom can be replaced with trimethylamine to afford the desired glycerophosphocholine 22a (Fig. 4.3).

Fig. 4.3
A chemical reaction has reactant compounds labeled 23 and 24 leading to a compound labeled 22 a in the presence of catalysts H W E and N M e 3. These compounds have open carbon chains with various functional groups.

Plan for the stepwise synthesis of 22a from aldehyde 24a

Full size image

Phosphonoacetate 23 was prepared as follows. Selective acetylation of the primary hydroxy group in 25 using dibutyltin oxide/AcCl afforded 26 in 58% yield [25]. Acylation of the secondary hydroxy group in 26 with diethylphosphonoacetic acid (27) using DCC/DMAP furnished phosphonoacetate 28 in quantitative yield. The benzyl group in 28 was then removed using hydrogenolysis. Next, the coupling of the resulting primary alcohol 29 with phosphoramidite 30 and subsequent oxidation by tert-butyl hydroperoxide afforded 23 (Scheme 4.6).

Scheme 4.6
A complex chemical synthesis process leading to the formation of a compound labeled 23 has various reactions leading to different intermediate compounds. Some compounds involved are labeled 25, 26, 28, and 29. These compounds mostly have open carbon chains with various functional groups.

Synthesis of phosphonoacetate 23

Full size image

Using 23, we then examined the HWE reaction of aldehyde 24a, which was prepared from 21 by acetylation (Table 4.4). As 23 contains a bromoethyl moiety, it may decompose via β-elimination in the presence of a strong base such as sodium hydride. Therefore, we performed the HWE reaction using LiBr and 10 equivalents of Et3N in THF, as reported by Rathke et al. [26], and observed that the desired α,β-unsaturated ester 31 was obtained in 54% yield (entry 1). Use of DIEA instead of NEt3 increased the yield up to 76% (entry 2) [27]. This may have occurred because DIEA was hindered and hence unable to attack the bromide. However, a decrease in yield was observed when only 5 equivalents of DIEA was used (entry 3). When the reaction was carried out in MeCN, the starting material was consumed within 30 min; however, the yield of 31 decreased to 59% (entry 4). The use of a polar solvent, such as MeCN, may accelerate the HWE reaction; however, it may also lead to the undesired nucleophilic substitution with DIEA. Based on the above results, we decided to use the conditions shown in entry 2 as the optimal conditions for the HWE reaction.

Table 4.4 Investigation of the HWE reaction of aldehyde 24a with phosphonoacetate 23
Full size table

The conversion of the glycerophosphocholine precursor 31 into glycerophosphocholine using trimethylamine was examined (Table 4.5). Nucleophilic substitution of compound 31 with trimethylamine was performed in THF and the concomitant removal of the tert-butyl group afforded glycerophosphocholine 22a in low yield (30%, entry 1). As some of the starting material remained unreacted, MeCN was added to accelerate the reaction. As expected, the reaction time decreased and the yield increased (55%, entry 2). At 80 °C, the reaction was complete in 12 h to afford 22a in 59% yield (entry 3). Thus, we succeeded in synthesizing 22a in a moderate yield through the HWE reaction of aldehyde 24a with phosphonoacetate 23, followed by its nucleophilic substitution with trimethylamine. This method could be applied to the synthesis of glycerophosphocholine derivatives containing an α,β-unsaturated ester.

Table 4.5 Investigation of the formation of glycerophosphocholine 22a
Full size table

Compound 1 was synthesized from 22a as follows. The Boc group and acetonide were removed by treating 22a with 3 M HCl in dioxane. Next, the nitrogen atom-selective sulfamate formation of amino alcohol 32 using SO3 pyridine in the presence of triethylamine in THF—water furnished 1. The 1H and 13C NMR spectra and the specific rotation of 1 were in good agreement with those of (+)-siladenoserinol A. Hence, the total synthesis of (+)-siladenoserinol A (1) was accomplished (Scheme 4.7).

Scheme 4.7
A chemical reaction has a compound labeled 22 a leading to a compound labeled 32 which finally yields siladenoserinol A in the presence of S O 3 dot P y, E t 3 N, and T H F H 2 o.

Total synthesis of (+)-siladenoserinol A (1)

Full size image

4.5.3 Convergent Approach to 1 via the HWE Reaction Using Glycerophosphocholine-Containing Phosphonoacetate 33

We then considered a unique and further convergent approach to 22a through the simultaneous construction of the α,β-unsaturated ester and the introduction of the glycerophosphocholine moiety through the HWE reaction of aldehyde 24a with phosphonoacetate 33, which contained a glycerophosphocholine moiety (Fig. 4.4).

Fig. 4.4
A chemical reaction. It begins with a compound with a 7 carbon ring with one C replaced with O. O B n group, carbon chain with O B n and an aldehyde group are attached. This converts to a diol that further converts to an aldehyde by selection oxidation. This aldehyde further converts to a carboxylic acid.

Plan for the one-step synthesis of 22a by the HWE reaction using 33

Full size image

Although the coupling of α,β-unsaturated acid 2 and alcohol 3 did not proceed (Scheme 4.5), the simpler phosphonoacetic acid 27 reacted with 3 using DCC/DMAP in refluxing CH2Cl2 to afford 33 in moderate yield (Scheme 4.8).

Scheme 4.8
A chemical reaction has a reactant compound labeled 3 yielding a compound labeled 33 in the presence of a compound labeled 27 and catalysts D C C, D M A P, and C H 2 C l 2.

Synthesis of phosphonoacetate 33

Full size image

To investigate the optimal reaction conditions for the HWE reaction using glycerophosphocholine-containing phosphonoacetate 33, the reaction with benzaldehyde was investigated (Table 4.6). When NaH was used as a base, only a trace amount of the desired α,β-unsaturated ester 34 was obtained (entry 1). This probably occurred because of the decomposition of phosphonoacetate 33 under strongly basic conditions. In contrast, the Masamune–Roush method [27] afforded the desired α,β-unsaturated ester 34 in moderate yield (entries 2–5). Especially, the use of LiBr/DBU in THF produced the best result (entry 4).

Table 4.6 Investigation of the HWE reaction of benzaldehyde and phosphonoacetate 33
Full size table

The optimal reaction conditions were applied to the coupling of 33 with aldehyde 24a and its benzoyl analog 24b. To our delight, the HWE reactions afforded 22a and 22b in 64% and 51% yields, respectively. Subsequently, acetonide and the Boc group were removed, and the selective sulfamate formation described in Scheme 4.7 furnished (+)-siladenoserinol A (1) and its benzoyl analog 35 in moderate yields (Scheme 4.9). Thus, the convergent synthesis of siladenoserinol A and its analogs was accomplished [5]. In addition to our synthesis, the Tong group succeeded in the total synthesis of siladenoserinols A and H [28]. Moreover, the Liu and Du group achieved the total synthesis of siladenoserinols A and D [29].

Scheme 4.9
A complex chemical synthesis process yielding a compound labeled 1 and 35 has various reactions leading to different intermediate compounds. Some compounds involved are labeled 33, 24 a and b, 22 a and b, and 1 and 35. These compounds have open carbon chains with various functional groups.

Convergent synthesis of 1 and its benzoyl analog 35

Full size image

4.6 Biological Evaluation

The inhibitory activities of siladenoserinol A, synthetic compounds 1, 32, and 35 against p53–Hdm2 interaction were evaluated by the Tsukamoto group, who had isolated siladenoserinols [1]. The synthetic compound 1 exhibited potent inhibitory activity comparable to that of the natural product (Table 4.7, entries 1 and 2). The desulfamate derivative 32 did not exhibit potent activity (entry 3). Therefore, either a sulfamate or a sulfate group in the serinol moiety is crucial for p53–Hdm2 inhibition (see Tables 4.1 and 4.2). Notably, the benzoyl analog 35 was found to be more potent than 1 (entry 4) [5].

Table 4.7 Inhibitory activity of siladenoserinol and its derivatives against p53-Hdm2 interaction
Full size table

4.7 Conclusion

The total synthesis of (+)-siladenoserinol A (1) was accomplished. The bicyclic acetal skeleton of siladenoserinol A was synthesized by the Au(III)-catalyzed cycloisomerization of 7,8-dihydroxy-2-alkynoate. The serinol moiety was introduced via the Julia–Kocienski reaction. Although the condensation reaction of the glycerophosphocholine moiety with the α,β-unsaturated acid did not proceed, an α,β-unsaturated ester containing the glycerophosphocholine moiety was successfully synthesized by the HWE reaction using an originally developed glycerophosphocholine-containing phosphonoacetate. Deprotection and selective sulfamate formation furnished 1. Using this method, a benzoyl analog 35 was also synthesized. Their inhibitory activities against the p53–Hdm2 interaction were evaluated. Compound 1 exhibited comparable activity to siladenoserinol A, but the desulfamate derivative 32 showed lower activity. Therefore, the presence of sulfamate or sulfate in the serinol moiety is essential for p53–Hdm2 inhibition. Notably, the p53–Hdm2 inhibitory activity of the benzoyl analog 35 was better than that of 1, which indicates that its analogs have high potential for this PPI inhibition.