Stereochemistry of Marine Sterols : ( 22 E )-24-Ethyl-24-methylcholesta-5 , 22-dien-3 bol and 24-Ethyl-24-methylcholest-5-en-3 bol

As configurações dos carbonos C-24 nos compostos (22E)-24-etil-24-metilcolesta-5,22dien-3b-ol (1) e 24-etil-24-metilcolest-5-en-3b-ol (2), isolados da esponja do caribe colombiano Topsentia ophiraphidites foram determinadas como R e S, respectivamente, com base na comparação dos seus dados de RMN com aqueles de amostras de configurações já definidas 24R e 24S, as quais foram sintetizadas em rotas envolvendo ortoésteres provenientes do rearranjo de Claisen de álcoois D-22 alílicos. Este é primeiro estudo de síntese, em que o rearranjo de Claisen é utilizado para introduzir um centro quaternário em C-24, de maneira estereoespecífica e com rendimento aceitável. Análises por difração de raio X de 1 confirmaram essas atribuições estereoquímicas.


Results and Discussion
In order to determine the C-24 configuration of 1 and 2, we decided to synthesize the stereochemically defined (24S)-and (24R)-(22E)-D 22 -24-ethyl-24methylcholesterols (1a and 1b), and (24R)-and (24S)-24-ethyl-24-methylcholesterols (2a and 2b).A method based on Claisen rearrangement appeared to be attractive for the stereoselective construction of the C-24 dialkyl side chains, although the application of this method to the construction of a C-24 quaternary center is unprecedented.0][11] We also utilized the orthoester Claisen rearrangement for the stereoselective synthesis of (24R)and (24S)-(22E)-24-isopropenyl-22-dehydrocholesterol and 24-isopropenylcholesterol, another multiply alkylated sterol of T. ophiraphidites. 12he synthetic route to 1a and 1b, and 2a and 2b is shown in Figure 2. The required starting materials, (22R)-allylic alcohol 3a and (22S)-allylic alcohol 3b, for the orthoester Claisen rearrangement were obtained stereoselectively from the corresponding enone [13][14][15] by reduction using DIBAL-H and L-selectride, respectively.14 Treatment of 3a with triethyl orthoacetate and propionic acid in refluxing xylene gave a mixture of the desired rearranged product 4a (not isolated as such) and by-products containing the 22-propionate ester of 3a.The C-24 epimer of 4a was not formed in the rearrangement reaction, since careful TLC and NMR analysis revealed that compounds 4a and 5a were free from 4b and 5b, respectively.The 24S configuration of 4a was assumed on the basis of the preferred transition state conformation for the rearrangement reaction (Figure S1) that is analogous to the conformation well established for the 3,3-sigmatropic reactions of steroidal 23-ene-22-allylic alcohols.16 The 24S configuration of 4a was ascertained by the eventual conversion of 4a to the (24S)-diastereomer 1a (vide infra).The product mixture containing 4a was reduced with LiAlH 4 and then the resulting alcohol mixture was acetylated to facilitate the separation.Hydride reduction of the purified acetate gave the primary alcohol 5a in 24% yield from 3a.Deoxygenation of the C-29 hydroxy group of 5a was achieved by super-hydride reduction of the corresponding mesylate 6a to give the i-methyl ether 7a.Removal of the i-methyl ether of compound 7a furnished the (24S)-diastereomer 1a.Hydrogenation of 7a gave the saturated (24R)-diastereomer 8a which was converted to the (24R)-diastereomer 2a upon deprotection of the i-methyl ether. Appliation of the same sequence of reactions to the (22S)-allylic alcohol 3b, furnished the (24R)-diastereomer 1b via the intermediates 4b, 5b, 6b and 7b.Hydrogenation of 7b gave 8b, which furnished the (24S)-diastereomer 2b by regeneration of the D 5 -3b-hydroxy system.
The 1 H and 13 C NMR spectroscopic data of the synthetic diastereomers 1a and 1b are listed in Tables 1 and 2, respectively, together with those of the sterol 1 isolated from T. ophiraphidites.The 1 H and 13  were different by more than 0.10 ppm, and this much difference would be sufficient for the identification of the diastereomers.The 13 C data of natural 1 met with those of the (24R)-diastereomer 1b.It was, therefore, concluded that the sterol 1 isolated from T. ophiraphidites has 24R configuration.The sterol 1 isolated from Pseudoaxinyssa sp. 3 was also identified as the (24R)-diastereomer 1b, since the published 1 H NMR data (adjusted by the addition of 0.005 ppm) was much closer to those of 1b, rather than those of 1a.The sterol 1 isolated from P. bulbosa could be assigned as the (24R)-diastereomer 1b, since the reported 13 C shifts of C-24, C-28 and C-29 resembled the values of 1b rather than those of 1a.
The 1 H and 13 C NMR spectroscopic data of the sterol 2 isolated from T. ophiraphidites and the synthetic 24-diasteromers 2a and 2b are listed in Tables 1 and 2, respectively.The 1 H and 13 C signals of 2a and 2b were unambiguously assigned as described for 1a and 1b.For example, in the HMBC spectrum of 2a,  S2 that the sterol 2 is identical with 2b, rather than 2a.The minor deviation observed in some signals of 2 and 2b could be within an error level in the NMR measurements.Hence, the 24S configuration of the sterol 2 isolated from T. ophiraphidites was unequivocally established.Comparison of melting points and [a] D values (see Experimental) also supported this stereochemical assignment.The C-24 configuration of 24-ethyl-24methylcholesterol isolated from Pseudoaxinyssa sp. 2 was also assigned as S, since the reported 1 H NMR data (recorded in benzene-d 6 ) met with those (recorded in benzene-d 6 ) of (24S)-2b.
Finally, we carried out X-ray analysis for a crystalline sample of the sterol 1 isolated from T. ophiraphidites.The molecular structure and the C-24 stereochemistry of 1 were established as shown in Figure S3.The 24R configuration of 1 was inferred by assuming the absolute configuration of natural sterols.The results proved the 24R configuration of the rearranged product 4b (therefore, 4a should have 24S configuration) and correctness of the transition state conformation (Figure S1) proposed for the orthoester Claisen rearrangement employed for the stereoselective creation of the C-24 quaternary center.

Conclusions
The present study established the 24R configuration of the marine sterol 1, isolated from T. ophiraphidites, Pseudoaxinyssa sp., and Psammocinia bulbosa, and the 24S configuration of the marine sterol 2, isolated from T. ophiraphidites and Pseudoaxinyssa sp.The C-24 configurations for the sterols 1 and 2, originally proposed by Djerassi's group, 3 was confirmed with full experimental details.The findings that sterols 1 and 2 possess the same C-24 orientation (24a F for the methyl group) supported the view that the sterol 2 is a biosynthetic precursor of the corresponding D 22 -sterol 1 in sponges.Djerassi and co-workers 3 suggested that stigmasta-5,24-dien-3b-ol could be a progenitor of the unique sterols, 1 and 2. Stereochemical studies on the other stereochemically undefined multiply-alkylated sterols isolated from T. ophiraphidites are in progress in our laboratory.

Experimental
Melting points were determined on a Yazawa BY-1 hotstage micro melting point apparatus and are uncorrected.NMR spectra were obtained on a JEOL JNM LA-400 (400 MHz for 1 H, 100 MHz for 13 C) spectrometer in CDCl 3 solution (4-10 mg sample per test tube) with tetramethylsilane as an internal reference at ca. 25 o C. 13 C chemical shifts are referred to the solvent signal (d 77.00).EI and FAB MS spectra were recorded with a JEOL JMS-AX700 spectrometer.Optical rotations were measured on a JASCO DIP-360 polarimeter.

X-ray study of natural sterol (1)
Crystals for X-ray analysis were obtained by allowing to stand a warmed MeOH solution of 1 with slow evaporation of the solvent through a small hole in the cap.C 30 H 52 O • 1/2MeOH • 1/2H 2 O, orthorhombic, P2 1 2 1 2 1 , Z = 8, a = 70.8949(17),b = 7.7590(2), c = 10.0415(2)Å, V = 5523.6(2)Å 3 , T = 123 K, Dx = 1.086 gcm -3 , l(CuKa) = 1.54184Å, m = 0.490 mm -1 .A number of 9298 reflections were collected by the oscillation photograph method using Rigaku R-AXIS RAPID Imaging Plate camera and Lp and semi-empirical absorption corrections were applied.The structure was solved by direct methods using SIR-97.The positional and anisotropic thermal parameters were refined by full-matrix least squares SHELXL-97.The positions of hydrogen atoms were calculated geometrically and refined using the riding model.Final R = 0.090 for 6737 unique reflections with |I o | > 2s(I o ).One methyl group of the terminal isopropyl group has a disordered structure in both independent molecules.The occupancy factors are 0.60:0.40 and 0.52:0.48,respectively.Sterol molecules are linked by hydrogen bonding directly or via methanol and water molecules.
H 3 -21 (d 0.923) was correlated with C-17, C-20 and C-22, H 3 -26/H 3 -27 (d 0.786 and 0.797) with C-24 and C-25, H 3 -29 (d 0.749) with C-24 and C-28, and H 3 -30 (d 0.674) with C-23, C-24, C-25 and C-28.The 1 H NMR signals, in particular H 3 -26 and H 3 -29, of 2 were in excellent agreement with those of the (24S)-diastereomer 2b, but not with those of the (24R)diastereomer 2a.This was confirmed by comparison of the 13 C NMR data.The C-30 signal showed the largest chemical shift difference (Dd 0.11) between the diastereomers, and the C-20 and C-28 signals exhibited 0.09 and 0.07 ppm difference, respectively.Figure S2 illustrates a graphic representation of the systematic 13 C shift comparison.It is obvious from Figure

Table 2 .
13C NMR spectroscopic data (100 MHz, in CDCl 3 ) for compounds 1, 1a, 1b, 2, 2a and 2b a An upfield resonance was tentatively assigned to C-26 (The C-26 and C-27 signals were not necessarily correlated to the H 3 -26 and H 3 -27 signals because of very close shifts of these signals).