Demonstration of the Intercellular Compartmentation of l -Menthone Metabolism in Peppermint ( Mentha piperita ) Leaves

Abstract

MATERIALS AND METHODS The metabolism of I-menthone, which is synthesized in the epidermal oil glands of peppermint (Mentha piperita L. cv. Black Mitcham) leaves, is compartmented; on leaf maturity, this ketone is converted to I-menthol and I-menthyl acetate in one compartment, and to d-neomenthol and d-neomen-thyl glucoside in a separate compartment. AU of the enzymes involved in these reactions are soluble when prepared from whole-leaf homogenates. Mechanical separation of epidermal fragments from the mesophyll, followed by preparation of the soluble enzyme fraction from each tissue, revealed that the neomenthol dehydrogenase and the glucosyl transferase resided specifically in the mesophyll layer, whereas the menthol dehydro-genase and substantial amounts of the acetyl transferase were located in the epidermis, presumably within the epidermal oil glands. These results suggest that the compartmentation of menthone metabolism in peppermint leaves is intercellular, not intracellular. Rapid turnover of monoterpenes in the epidermal oil glands of maturing peppermint leaves is accompanied by the reduction of 1-menthone, the major constituent of the essential oil, to i-menthol and d-neomenthol (3, 4, 7). Of the two diastereomeric alcohols, only i-menthol is partially converted to the corresponding acetate, whereas d-neomenthol is nearly all glucosylated and the resulting ,B-D-glucoside is transported to the rhizome (Refs. 7 and 8; Fig. 1). In vivo and in vitro studies (8, 9, 15) have shown that the observed specificity of the pathways of 1-menthone metabolism is a result of compartmentation of a menthol-specific dehydrogenase with a relatively nonselective acetyltransferase, and of a neomenthol-specific dehydrogenase with a relatively nonselective glucosyl-transferase. All of these dehydrogenases (menthone reductases) and transferases have been isolated from whole-leaf homogenates and characterized (6, 9, 15), and all are 'operationally soluble' (i.e. located almost exclusively in the 105,000g supernatant). Thus, the physical basis for the compartmentation of pathways that must be operative in peppermint was not readily apparent from the previous studies with whole leaves. In this communication, we provide evidence that the enzymes of these two distinct pathways (Fig. 1) are located in situ in different types of leaf tissue. 2Author to whom inquiries should be made. Peppermint (Menthapiperita L. cv. Black Mitcham) plants were grown from stolons as described previously (7), and leaves from the midstem of flowering plants were washed with 1.5 mm EDTA before use. Whole-leaf extracts were prepared (all operations carried out at 0-4°C) by grinding 2 g (fresh weight) of this tissue in a Ten-Broeck homogenizer with 0.1 M Na-phosphate buffer (pH 7.0) containing 0.25 M sucrose, 25 mm Na2S205, 5 mm Na-ascorbate, 5 mM MgC 12, 1 mm CaC12, I mm dithioerythritol, and 2 g of insoluble PVP (Polyclar AT; GAF Corp., New York, NY) (12). The homogenate was slurried briefly with 2 g of Amberlite XAD-4 polystyrene resin (Rohm and Haas Corp., Philadelphia, PA) (14), filtered through cheesecloth, and centrifuged at 105,000g (90 min) to provide a supematant fraction containing the soluble enzymes. Epidermis extracts were prepared by submerging leaves (2 g) in the extracting medium described above and gently brushing the upper and lower surface with a soft bristle toothbrush. The removal of the epidermis is evidenced by the sloughing of transparent fragments and the bright green appearance of the underlying tissue. The buffer containing the epidermal fragments was then homogenized and treated as above to obtain a 105,000g supernatant fraction. A comparable mesophyll supernatant fraction was obtained from the leaves remaining after removal of the epidermis. Each of the soluble preparations (whole leaf, epidermis, mesophyll) was brought to assay conditions by dialysis against 50 mm Na-phosphate buffer (pH 7.0) containing 12 ma mercaptoeth-anol, 6 mM MgCl2, 1 mM CaC 12, and 4 g/L each of insoluble PVP and XAD-4 resin. Particulate fractions were prepared with the same extraction buffer, but soluble PVP (Plasdone K-90; GAF Corp.) was substituted for the insoluble polymer, and the XAD-4 treatment and filtration were omitted. The membranes separated by standard differential centrifugation and sucrose density-gradient centrifu-gation techniques (10) did not contain significant levels of the relevant dehydrogenase or transferase activities, and these fractions were not examined further. The measurement of glucosyltransferase activity was based on the UDP-glucose-dependent glucosylation ofd-[G-3Hneomenthol with TLC isolation of the labeled product as described in detail elsewhere (15). The radiochemical assay for acetyltransferase was based on the acetyl CoA-dependent acetylation of l-[G-3H]men-thol and TLC isolation of the product (6). Assay for 1-menthone reduction was based on the conversion of the tritium-labeled ketone to d-neomenthol and i-menthol in the presence of NADPH and a regenerating system, with TLC isolation of the diastereo-meric products (9). All assays were conducted with linear kinetics and saturating levels of substrates and cofactors. Boiled controls were run for each experiment, and the identification of products confirmed by radio-GLC analysis and by preparation of crystalline derivatives using techniques described before (6, 9, 15). Activ-975 www.plantphysiol.org on April 11, 2020-Published by Downloaded from