Limonene epoxides

Van der Werf Ml, KM Overkamp, JAM de Bont (1998) Limonene-1,2-epoxide hydrolase imm Rhodococcus erythropolis DCL14 belongs to a novel class of epoxide hydrolases. J Bacteriol 180 5052-5057. 

Limonene (10.128) is an analogue of 4-vinylcyclohexene, and, like the latter, it undergoes epoxidation of both the C(1)=C(2) and C(8)=C(9) bonds. Like in the dioxide 10.127, the two epoxide groups are hydrated at different rates by EH. Indeed, incubations in rat liver microsomes showed that hydrolysis of limonene 1,2-epoxide was 70 times slower than that of the 8,9-epoxide, a much larger difference than that observed for the dioxide 10.127 [192], Comparison of EH-catalyzed hydration of the four epoxy groups in 4-vinyl-cyclohexene and limonene confirmed that the relative rates decreased with increasing steric hindrance at these groups. 

C. Samples were withdrawn every two days from the samples stored at 45 C and every three days from the samples stored at 37 C. Pulled samples were stored in screw cap vials at 0 C until analsis by gas chromatography (GC). The products were monitored for the formation of limonene-1,2-epoxide and L-carvone, both oxidation products of d-limonene (3). 

Gas Chromatographic Analysis. The contribution of limonene-1, 2-epoxides and carvone to the development of oxidized flavor of encapsulated orange oil has been investigated (5). The concentrations of these two compounds were reported to provide a reliable index of the stability of the encapsulated orange oil. 

The identification of limonene-1,2-epoxides and carvone was accomplished by retention indices and mass spectra, obtained separately by GC-MS, in comparison to authentic reference compounds. 

Overall, powder B with intermediate particle diameter was found to provide best protection against the oxidation of orange oil. The rate of limonene-1,2-epoxide formation of powder A was 4.04 mg/g oil/day. This rate was decreased to 60JIJ and 80JIJ for powders B and C, respectively. The decrease in protective effect of powder C could have been caused by the increase in particle surface imperfections as discussed earlier. 

Cold pressed orange oil contains over 95 by weight monoterpene hydrocarbons. The principal constituent, d-limonene, ranges from 83-97 percent (9), Limonene degradation has been well docu-meted in the literature (LO, 1 1). Anandaraman identified limonene-1,2-epoxide and carvone as two of the earliest degradation products of d-limonene (12, 13). 

Little work has been published comparing the shelf life of flavors encapsulated by different processes. The purpose of this investigation was to determine the shelf life of cold pressed orange oil encapsulated by spray drying, extrusion, and molecular inclusion. Limonene-1,2-epoxide concentration was used to monitor oxidation. 

Previous work established that similar concentrations of limonene-1,2-epoxide could be used as an index for measuring oil oxidation, although the direct contribution of this compound to off-flavor development was marginal (12). 

The graphs of limonene-1,2-epoxide as a function of storage time at 25, 37 and 50 C are presented in Figures 3, 4 and 5. Limoene-1,2-epoxide appeared to have an induction period where little or no epoxide formation occurs. Beyond the induction period, limonene-1,2-epoxide formation followed first order kinetics for the spray dried products. Gum arabic, M250, and Amiogum consistently showed the fastest rates of epoxide formation at the three storage temperatures. 

The two extrusion products, Reg and SF Duraromes, exhibited the lowest rates of limonene-1,2-epoxide formation. After 6 months storage at temperatures of 25 and 37 C, these products contained epoxide levels of < 0.20 mg/g orange oil. At 50 C, the Reg Durarome showed levels 0.50 mg/g orange oil. Epoxide degradation may also have been a factor similar to that seen with beta-cyclodextrin. 

Overall, extrusion encapsulation provided superior protection of orange peel oil as measured by epoxide formation. The molecular inclusion of orange oil via beta-cyclodextrin also provided very good protection although the limonene-1,2-epoxide concentrations were consistently higher than for the extrusion products. 

Very recently, the purification and characterisation of an epoxide hydrolase, catalysing the conversion of limonene-1,2-epoxide to limonene-1,2-diol has been described [90]. The enzyme was isolated from Rhodococcus erythropolis DCL14 and is induced when the microorganism is grown on monoterpenes. The authors found evidence that the enzyme, limonene-1,2-epoxide hydrolase is the first member of a new class (the third class) of epoxide hydrolases [91]. 

Two new preparations of 1,8-cineole (553) [the biogenesis of which from geraniol (25) in Rosmarinus officinalis has been elucidated ] have been recorded. A Diels-Alder adduct 625 (R = Me) of methyl vinyl ketone and isopropenyl methyl ketone was converted to the diazoketone (R = CH = N2) with diethyl oxalate/base, then toluenesulfonyl azide, and treatment of the latter with [Rh(OAc)2]2 in methylene chloride at room temperature for five minutes converted it in very high yield to the tricyclic compound 626. Lithium dimethylcuprate then yielded the ketone 627, conversion of which to 1,8-cineole (553) was known.The other 1,8-cineole synthesis was a by-product of an observation which enabled the two stereoisomers of limonene 1,2-epoxide to be separated. The cjs-epoxide 549 was brominated to stereoisomers of a dibromocineole 628, under conditions when the rran.s-epoxide did not react, and could be distilled pure afterward. The dibromo compound 628 yielded 1,8-cineole (553) with tributyltin hydride. 

The products of peracid oxidation of limonene have also been re-examined. Wylde and Teulon have shown that the best method for making pure cis-or trans- limonene 1,2-epoxides, a mixture of which is obtained by direct peracid oxidation in chlorinated hydrocarbon solvents, is to treat this mixture with hydrogen chloride in ether, when the two diaxial chlorohydrins (122) and (123) are obtained with practically no equatorially substituted isomers. Of these two isomers only (122) forms a p-nitrobenzoate, allowing (123) to be distilled from the residue. Treatment of the nitrobenzoate of (122) with methanolic potassium hydroxide now leads to the c/s-epoxide (120) similar treatment of... 

The reaction of limonene-1,2-epoxides, either pure or as a mixture, with bases like aluminum alkoxides to give allyl alcohols has been examined, and the similar reaction with propyl-lithium of 1,2-epoxy-trans-p-menthane and the menthane-2,3-epoxide (126), yielding (127) and (128) has also been discussed. ... 

An alternative process for the production of (-)-carvone has recently been elaborated. Starting from (-i-)-limonene 1,2-epoxide, a regioselective rearrangement of the epoxide leads to (-)-carveol (trans- -.[2102-58-1] cis- -.[2102-59-2]). The reaction is effected by the use of a catalyst consisting of a combination of metal salts and phenolic compounds. 

LEH originates from the bacterium Rhodococcus erythropolis DCL14." ° LEH is part of a limonene degradation pathway where it catalyzes the conversion of limonene-1,2-epoxide to limonene-1,2-diol (Scheme... 

LEH displays relatively narrow substrate specificity and accepts only few substrates. These include both enantiomers of 1-methylcyclohexene oxide (1 and 2, Scheme 2) and all four stereoisomers of the natural substrate limonene-1,2-epoxide (3-6, Scheme 2). The substrates are converted with different enantioselec-tivities and regioselectivities. The four stereoisomers of limonene-1,2-epoxide are hydrolyzed in an enantioconvergent fashion. Conversion of the diastereomeric mixture of 3 and 4 leads to enantioconvergent formation of (li ,2i ,4i)-limonene-l,2-diol, whereas conversion of 5 and 6 leads to enantioconvergent... 

Scheme 2 Conversion of the two enantiomers of 1 -methylcyclohexene oxide and the four stereoisomers of limonene-1,2-epoxide by LEH. Adapted from K. H. Hopmann B. M. Hallberg F. Himo, J. Am. Chem. Soc. 2005, 127, 14339-14347.

The regioselectivity of LEH has been studied experimentally with different substrates. Studies with the enantiomers of 1-methylcyclohexene oxide (1 and 2, Scheme 2) revealed preferred attack at the methyl-substituted oxirane carbon, Cl, with a regioselectivity of 85(C1) 15(C2). This indicated an acid-catalyzed mechanism, which would result in preferred attack at the more substituted carbon. However, conversion of limonene-1,2-epoxide did not support this conclusion and showed somewhat intriguing results. Exclusive attack at the more substituted carbon (Cl) is seen for the stereoisomers 4 and 5, while exclusive attack at the less substituted carbon (C2) is observed for stereoisomers 3 and 6 (Scheme 2). Interestingly, the two limonene-1,2-epoxide stereoisomers with the same stereochemistry at the oxirane ring, (IR,2S) for 3 and 5 and IS,2R) for 4 and 6, exhibit attack at opposite carbons (Scheme 2). A suggested explanation for the differences was differential binding of the substrates in the active site, which would lead to attack at different carbons. ... 

Scheme 4 (a) Cyclohexene oxide adopts a half-chair conformation that exists in two different helicities. (b) The isopropyl substituent of limonene-1,2-epoxide makes one of the two half-chair conformations energetically more favorable. 

H) Regioselectivity of limonene-1,2-epoxide hydrolysis For limonene-1,2-epoxide, the natural substrate of LEH, the situation is less complex than for 1-methylcyclohexene oxide. The isopropyl substituent at C4 of limonene-1,2-epoxide determines the preferred helicity of the half-chair, and for each stereoisomer, only the helicity with the substituent in an equatorial position will be observed (Scheme 4). For this helicity, attack is preferred at the carbon that leads to a chair-like transition state. The transition states for attack at either Cl or C2 of 5 are shown in Figure 3. Attack at Cl of 5 leads to a chair-like transition state and exhibits a barrier of 14.9 kcal mol. Attack at C2 of 5 results in a twist-boat transition state and exhibits a barrier of... 

Ozone will therefore selectively cleave the ring double bond leaving the other untouched, provided of course, that no more than one molar equivalent of ozone is used. Similarly, one molar equivalent of m-chloroperbenzoic acid will selectively give only limonene 1,2-epoxide... 

---