Bicyclic Triterpenoids

Most triterpenoids are 6-6-6-5 tetracycles, 6-6-6-6-6 pentacycles, or 6-6-6-6-5 pentacycles, but mono-, bi-, tri- and acyclic triterpenoids have also been isolated from native sources or generated by mutant forms.

The known bicyclic triterpenoids which are supposed to be derivatives of 2,3-oxidosqualene are polypoda-8(26),13,17,21-tetraen-3-ol (5)^[59], polypoda-7,13,17,21- tetraen-3-ol (6)^[60] and polypoda-13,17,21-trien-3,8-diol (7)^[61], which were found in Cratoxylum cochinchinense (5 and 6) and Pistacia resins (7). They are thought to be products that result from quenching chair-chair bicyclic C-8 carbocation by deprotonation of H26 (5) or H7 (6), or by water addition (7) (Fig. 1-11). These compounds have not been experimentally verified to be products of oxidosqualene cyclase, but no plausible alternative origin has been proposed. Interestingly, although B-ring boat folding also appears in oxidosqualene cyclase, the known bicyclic C30H50O triterpenoids are only chair-chair derivatives.

Figure 1-11 Cyclization of oxidosqualene with chair-chair conformation to form bicyclic 5

6

8 7

Some bicyclic triterpenoids, 7,13,17,21-polypodatetraene (γ-polypodatetraene)(9)^[62], 8(26),13,17,21-polypodatetraene (α-polypodatetraene)(10)^[62], 13,17,21-polypodatrien-8-ol (11)^[63], and neopolypodatetraene A (12)^[14,64] are considered to be squalene cyclization products which were derived from the chair-chair bicyclic cation intermediate via immediate deprotonation or water addition, or through several 1,2- shifts of hydrides and a methyl group before deprotonation (Fig. 1-12). The α-(10) and γ-polypodatetraene(9) were first isolated from the fresh leaves of Polypodium fauriei and Lemmaphyllum microphyllum for 9, and Polystichum ovatopaleaceum and P. polyblephalum for 10, respectively.^[62] 9 was also been found in moss Floribundaria aurea subsp^[65]; and 11 was first isolated from the fern Polypodiodes; whereas 12 has been merely generated by A. acidocaldarius SHC mutant F365A^[64]. Compounds 9 and 10 have also been obtained from A. acidocaldarius SHC mutants.^[14] In addition, (+)-α-polypodatetraene (10) and (+)-γ-polypodatetraene (9) have also been synthesized successfully. ^[66]

-Figure 1-12 Cyclization of squalene to form bicyclic triterpene compounds from bicyclic cation (13).

In the triterpene cyclization, whether the mechanism is a “concerted” or “stepwise”

process still confused most scientists. The cyclization of squalene and (3S)-2,3-oxidosqualene 9

12 11

13 10

otherwise detect any intermediates.^[16] The isolation of these abortive cyclization products may provide indirect evidence that the polycyclization is a stepwise process which proceeds via a series of partially cyclized carbocationic intermediates.^[38] In 1984, Bohr et al. first isolated a bicyclic triterpenoid (7) that the structure and absolute stereochemistry of which are fully consistent with its formation by interception of the bicyclic carbocation on OSC postulated as an intermediate in the cyclization of the chair-chair-chair conformation of (3S)-2,3-oxidosqualene.^[61] It provided an indirect evidence that the cyclization of B-ring may not concert with C-ring. In 1985, Nishizawa’s experiments also supported the stepwise mechanism of a biomimetic olefin cyclization by trapping of mono-, bi-, and tri-cyclic cationic intermediates.^[67]

However, Kronja et al. argued that the product composition is not very indicative of the reaction mechanism; either a stepwise or a concerted polycyclization can yield acyclic and polycyclic products. He further suggested an extended participation involving at least two double bonds in a biomimetic reaction of a squalene derivative according their kinetic measurements, and indicated a conceted biomimetic polycyclization.^[68] In contrast, recently, Hess et al., by using a density functional computational study, found two intermediates located in the oxidosqualene-lanosterol cyclase reaction pathway during the formation of the B ring, provided an evidence that the formation of the A and B rings was not concerted, although they were be found to be very shallow and perhaps would not lead to formation of a viable intermediate on the enzymatic pathway.^[40]

In the field of mutagenesis studies, the abortive products coupled with the structure information and kinetic analysis can give a detailed description of the functional role of a specific residue. Bicyclic products can be obtained from various mutants of A. acidocaldarius SHC (Fig. 1-13).^[14] There are significant accumulations of two bicyclic products in F365A mutant, indicating that Phe365 is critical to the completion of the polycyclization and that the

This finding also suggested that the Phe365 residue is close to the C-8 cation in the enzyme cavity. It is noteworthy that the formation of 12 may be achieved by a series of 1,2-shifts of hydrides and a methyl group in an antiparallel fashion that trigger the deprotonation at C-6.^[14,64]

In the mutants of Tyr609, bicyclic 9 and 10 were also appeared. However, the yields of the bicyclic compounds were different. The higher yield of 96% for 9 and 12 by F365A, than that of ca. 50% by Y609A and less than 10% of 10 by Y609F, suggesting that the bicyclic carbocation stabilization may have been achieved mainly by the π-electrons of Phe365 with the aid of Tyr609.^[54,69,70] Further, the mutants of Tyr612 produced bicyclic product 9, but only with a little amount (6%), indicating that Tyr612 may work to place Phe365 at the most favorable positions for cyclization catalysis and enrich the π-electrons of Phe365.^[54]

On the other hand, some mutants of Tyr420 afforded bicyclic 9 and 10 in a significant yield. Tyr420 and Tyr609 are located on the wall of the catalytic cavity in a mirror-image position, thereby, pointing to ring B of hopene.^[12] The accumulation of bicyclic products in Tyr420 mutants suggested that the major function of Tyr420 is to stabilize the bicyclic cation intermediate formed during the cyclization cascade.^[70,71,72] In addition, the Leu607 mutants gave significant amounts of bicyclic 9 and 10, indicating that the steric bulk size at position 607 is critical to the optimal folding of thechair form for the B-ring; the most appropriate bulk size of Leu gives rise to perfect contact around the B-ring formation site.^[72,73] (Note: there may some other products besides the bycyclic products in these mutants, but they are not described herein.)

Figure 1-13 Bicyclic products obtained by the mutated A. acidocaldarius SHCs.

H

H

H 9 10 12

F365A, Y420A/S, L607K, Y609A/C/L/S, Y612A/L

Y420A/G/F, Y609F F365A