Leaf cuticular waxes are arranged in chemically and mechanically distinct layers: evidence from Prunus laurocerasus L.

Authors


Dr R. Jetter Julius-von-Sachs-Institut für Biowissenschaften, Julius-von-Sachs-Platz 3, D-97082 Würzburg, Germany. E-mail: jetter@botanik.uni-wuerzburg.de

ABSTRACT

The composition and spatial arrangement of cuticular waxes on the leaves of Prunus laurocerasus were investigated. In the wax mixture, the triterpenoids ursolic acid and oleanolic acid as well as alkanes, fatty acids, aldehydes, primary alcohols and alcohol acetates were identified. The surface extraction of upper and lower leaf surfaces yielded 280 mg m2 and 830 mg m2, respectively. Protocols for the mechanical removal of waxes from the outermost layers of the cuticle were devised and evaluated. With the most selective of these methods, 130 mg m2 of cuticular waxes could be removed from the adaxial surface before a sharp, physically resistant boundary was reached. Compounds thus obtained are interpreted as ‘epicuticular waxes’ with respect to their localization in a distinct layer on the surface of the cutin matrix. The epicuticular wax film can be transferred onto glass and visualized by scanning electron microscopy. Prunus laurocerasus epicuticular waxes consisted entirely of aliphatic compounds, whereas the remaining intracuticular waxes comprised 63% of triterpenoids. The ecological relevance of this layered structure for recognition by phytotrophic fungi and herbivorous insects that probe the surface composition for sign stimuli is discussed.

INTRODUCTION

The interface between primary plant tissues and the atmosphere is provided by cuticles, i.e. extracellular membranes consisting of a polymeric cutin matrix and soluble cuticular lipids (SCL) ( Walton 1990). Plant cuticles contain SCL, also called ‘cuticular waxes’ ( Fig. 1), not only embedded within the cutin (intracuticular waxes) but also on their surface (epicuticular waxes) ( Baker 1982; Holloway 1982). It is widely accepted that all plant cuticles carry a thin film of epicuticular waxes on the surface of their cutin matrix ( Fig. 1) ( de Bary 1871; Baker 1982; Holloway 1982; Walton 1990). On particular plant species and organs epicuticular SCL can additionally form microscopic aggregates (epicuticular wax crystals) protruding from the film of epicuticular waxes ( Barthlott et al. 1998 ). Although the crystals are readily visualized by scanning electron microscopy (SEM) there is only circumstantial evidence for epicuticular wax films. To date, neither the dimensions nor the chemical composition of the epicuticular wax film could be assessed.

Figure 1.

Schematic cross-section through the outer parts of a generalized plant epidermis devoid of wax crystals (CP, cuticle proper; CL, cuticular layer; CW, cell wall; EWF, epicuticular wax film; IW, intracuticular waxes; P, pectinaceous layer and middle lamella; PL, plasmalemma). (Modified after Jeffree 1986)

For many plant systems it has been shown that the chemical composition of epicuticular wax crystals differs from the composition of bulk cuticular waxes. Evidence for this distinction largely exploited the characteristic shapes and the high mass portions of epicuticular crystals ( Baker 1982; Gülz et al. 1992 ; Jetter & Riederer 1994). Unfortunately, on those plants that are covered only by an epicuticular wax film similar approaches cannot be taken because such films (necessarily) lack characteristic micromorphology and consist of relatively small quantities. Instead, a distinction between intracuticular waxes and the epicuticular wax film was attempted by either mechanically stripping ( Haas & Rentschler 1984) or chemically washing ( Silva Fernandes, Baker & Martin 1964; Holloway 1974; Baker & Procopiou 1975; Baker, Procopiou & Hunt 1975; Svenningsson 1988; Garrec, Henry & LeMaout 1995) the film off the plant surface. The remaining SCL were subsequently extracted and regarded as intracuticular compounds. None of these methods had sufficient selectivity to exclude cross-contamination of both fractions and hence the individual constituents could not be localized unambigously. However, in all cases the transverse gradients in the percentages of individual compounds suggested chemical differences between intracuticular SCL and the epicuticular film.

Further evidence for a heterogenous arrangement of waxes within the cuticle came from Fourier Transform Infared Spectroscopy (FTIR) studies on the phase behaviour and the crystallinity of aliphatic SCL constituents. For some plant species the investigation of reconstituted cuticular waxes and those still embedded in the cuticle produced comparable results ( Merk, Blume & Riederer 1998). However, for those SCL mixtures comprising high proportions of triterpenoids these structural parameters changed markedly after wax recrystallization ( Merk 1998). These findings indicated a (partial) separation of aliphatic and triterpenoid constituents, possibly a layered structure, within the native cuticles that was lost during wax isolation. Due to the principal differences in the chemical properties of both compound classes a demixing of their solid phases in the cuticle seems plausible.

It is now an intriguing objective to (1) verify the uneven distribution of triterpenoids by analytical methods and (2) to use it for delineating differences between epi- and intracuticular SCL in respective cuticles. In the present inves-tigation procedures for the selective preparation of epicuticular waxes were used to assess the spatial organization of leaf waxes containing high percentages of triterpenoids. As a model system we chose leaves of Prunus laurocerasus. The results were evaluated by comparison with protocols that had previously been used for the preparation and chemical analysis of SCL.

MATERIALS AND METHODS

Plant materials

Plants of Prunus laurocerasus were continuously grown in greenhouses in the Botanical Garden of the University of Würzburg. Mature leaves were harvested randomly from three individual trees in autumn and winter 1998 and pooled for wax removal and analysis. Surface areas were determined gravimetrically using photocopies of the extracted leaves. They were found to have average total (adaxial + abaxial) surface areas of 1·58 × 10−2 m2. For analyses of immature leaves three shoots were marked at the time of budbreak and inspected regularly. Pairs of fresh leaves were harvested when they had fully expanded but still appeared light green.

Mechanical wax removal

For cryo-adhesive-based mechanical wax removal a droplet of approximately 100 μL glycerol was mounted on a metal spatula. Then a leaf disc, 12 mm in diameter, was gently pressed onto the droplet until the liquid covered the entire disc area. The whole arrangement was frozen by dipping the spatula into liquid nitrogen. Finally, the still-frozen leaf disc was removed and the glycerol was transferred into a vial containing 5 mL each of chloroform and water. Preliminary experiments had shown that this procedure assured reproducible SCL yields in the organic phase and separation from the glycerol which accumulated in the aqueous phase. The glycerol droplets with adhering surface waxes from 15 individual leaf discs were pooled into the same two-phase system. Tetracosane was added as internal standard; after vigorous agitation and phase separation the organic solution was removed and the solvent was evaporated under reduced pressure.

In one series of experiments, the same protocol was carried out using water instead of glycerol. In another series, 20–100 μL of chloroform were applied to the surface of leaf discs, 12 mm in diameter, in 10 μL portions using a syringe. The solvent was allowed to evaporate between application of consecutive portions. Waxes were subsequently removed with glycerol as cryo-adhesive as described above. Alternatively, chloroform was applied in single 50, 100, 150 or 200 μL portions in a glass cylinder and the solvent was evaporated during 1, 2, 3, or 4 min, respectively, under a stream of N2. The surface waxes were subsequently removed using glycerol as cryo-adhesive as described above.

For polymer-based mechanical wax removal a protocol first published by Haas & Rentschler (1984) was used. Collodion (4%, Merck, Darmstadt, Germany), i.e. a solution of nitrocellulose in ether : ethanol (3 : 1, v/v), was applied to leaf discs, 12 mm in diameter, using a small paint brush. After evaporation of the solvent a polymer film could be stripped off the surface together with adhering waxes. The polymer films from 15 leaf discs were pooled and extracted with 10 mL chloroform at room temperature. As before, tetracosane was added and the solution was concentrated under reduced pressure. After treatment with either glycerol, water or collodion the leaf discs were still intact and could be used in repeated mechanical removal experiments or for superficial solvent extraction of the remaining SCL.

Duplicate experiments using the protocol for wax removal with glycerol were carried out to test the reproducibility of the method. In these parallels all the steps from leaf harvest through wax preparation to chromatographic analysis were included. The overall yields of SCL resulting from single treatments were found to vary by ± 13%, whereas the percentages of the main individual compound classes varied by ± 2 to ± 11%. In view of the largely qualitative interpretation of the data the results were considered as sufficiently reproducible. Therefore, in all the subsequent experiments involving mechanical removal of surface waxes pooled samples were analysed instead of parallel treatment of independent replicas.

Wax extraction

For superficial extraction of SCL four leaf blades were immersed twice for 30 s in approximately 250 mL chloroform at room temperature. This procedure had routinely been applied in most of the wax analyses reported in the literature ( Walton 1990). Selective extraction of SCL from either the adaxial or the abaxial surface was achieved by placing the intact leaf onto a flexible rubber mat, gently pressing a glass cylinder, 10 mm in diameter, onto the exposed surface and filling the cylinder with approximately 1·5 mL of chloroform. The solvent was agitated for 30 s by pumping with a Pasteur pipette and removed. When any solvent leaked between cylinder and leaf surface the sample was discarded. Extracts from 10 individual leaves were pooled for further analysis. Tetracosane was immediately added to all the extracts of cuticular waxes as internal standard and the solvent removed under reduced pressure.

All extraction experiments were carried out as entirely independent duplicates, results are given as mean values. Overall extraction yields varied by ± 0·1 to ± 1% and coverages of individual aliphatic compound classes were reproducible within ± 0·3 to ± 6%. Respective values for triterpenoid acids varied by ± 4 to ± 16%.

Chemical analysis

Prior to gas chromatography (GC) analysis hydroxyl-containing compounds in all samples were transformed to the corresponding trimethylsilyl derivatives by reaction with bis-N,O-trimethylsilyltrifluoroacetamide (Macherey-Nagel, Düren, FRG) in pyridine (30 min at 70 °C). The composition of the mixtures was studied by capillary GC (Hewlett Packard 5890 II; Hewlett Packard, Avondale, Pennsylvania, USA) with on-column-injection (30 m OV-1 WCOT i.d. 320 μm, Chrompack, Middelburg, The Netherlands) and mass spectrometric detector (70 eV, m/z 50–650, hp 5971). GC was carried out with temperature-programmed injection at 50 °C, oven 2 min at 50 °C, increasing by 40 °C min−1 to 200 °C, 2 min at 200 °C, increasing by 3 °C min−1 to 320 °C, 30 min at 320 °C and He carrier gas inlet pressures programmed 5 min at 5 kPa, increasing by 3 kPa min−1 to 50 kPa, 30 min at 50 kPa. SCL components were identified by comparison of their mass spectra with those of authenticated standards and literature data. For quantification of individual compounds GC was used under conditions as described above, but with carrier gas H2 (5 min at 50 kPa, increasing by 3 kPa min−1 to 150 kPa, 30 min at 150 kPa) and flame ionization detector.

Scanning electron microscopy

The method of Ensikat, Neinhuis & Barthlott (2000) was used to mechanically remove wax films from the surface of P. laurocerasus leaves. The treated leaves and the glass slides carrying the wax preparations were mounted on aluminium holders, sputtered with 20 nm of gold (Balzers Union Sputtering Device; 25 mA, 300 s; Balzers, Switzerland) and investigated by SEM (Zeiss DSM 15 kV, 6 mm; Zeiss, Oberkochen, Germany).

RESULTS

The chemical composition of bulk cuticular waxes from mature Prunus laurocerasus leaves was analysed using standard methods for extraction. Immersing whole leaf blades into chloroform yielded 450 mg m−2 of cuticular waxes. Two triterpenoids, oleanolic and ursolic acid, were found to be among the most prominent individual compounds with 2 and 9% (10 and 40 mg m−2), respectively. The most abundant compound class (72 mass-% of the mixture, 320 mg m−2) consisted of alkanes with chain lengths C25– C33 ( Table 1, Fig. 2a). They showed a strong predominance of odd carbon numbers with a maximum for C29. In addition, the bulk cuticular wax contained 9% primary alcohols (40 mg m−2), 1% aldehydes (6 mg m−2) and 2% fatty acids (10 mg m−2) as well as traces of alcohol acetates ( Table 1, Fig. 2a). All these fractions were dominated by even-numbered homologues, especially C30.

Table 1.  Homologous composition (% of the fraction) and relative amounts (% of the wax mixture) of aliphatic compound classes in Prunus laurocerasus leaf cuticular waxes
Chain lengthAlkanesAlcoholsAldehydesFatty acids
  1. atraces, i.e. less than 0·5% detectable.

201
21tra
22tr2
23tr1
24127
25trtr2
26tr335
275144
283132419
2947488
304333131
31384
322282821
3323
347
Relative amounts72912
Figure 2.

Yields and compositions of SCL extracted from mature P. laurocerasus leaves using chloroform. Coverages of individual compound classes are compared for extracts (a) of whole leaves and (b) of intact abaxial and adaxial surfaces.

In order to obtain adaxial and abaxial leaf-surface waxes separately, new extraction techniques were devised. The solvent was applied in a glass cylinder that was gently pressed onto the leaf surface, thus allowing extraction of well-defined areas. Care was taken that conditions (e.g. solvent volume to leaf area ratio, temperature, exposure time) were similar when either whole leaves were immersed or individual leaf sides were extracted. Adaxial and abaxial leaf surfaces yielded 280 mg m−2 and 830 mg m−2 of cuticular wax, respectively ( Fig. 2b). The SCL yields achieved by whole leaf immersion hence represent the mean between the two largely differing adaxial and abaxial coverages. Wax extracts from adaxial and abaxial leaf surfaces contained similar proportions of compound classes. In particular, the ratios between aliphatic and triterpenoid constituents were constant. The most pronounced difference was the higher relative amount of alkanes in the waxes from abaxial surfaces that was compensated by higher relative contributions from other aliphatic compound classes on the adaxial surfaces. Typical constituents of internal lipids were not detected either in the abaxial or in the adaxial surface extract. In accordance with previous reports on whole leaf extractions it can therefore be assumed that the organic solvent does not gain access to internal tissues through stomatal cavities.

The spatial arrangement of cuticular waxes on adaxial leaf surfaces of P. laurocerasus was investigated by mechanical removal of the outermost wax layers and subsequent chromatographic analysis of the resulting samples. The fraction of cuticular wax removed with glycerol as cryo-adhesive consisted of alkanes, fatty acids, aldehydes, alcohols and alcohol acetates in relative amounts similar to those in the chloroform extract of the same surface ( Fig. 3a). However, in sharp contrast to solvent extractions the mechanical removal did not yield any triterpenoids. Two further applications of cryo-adhesive to the same leaf discs produced only minute amounts of the same aliphatic compounds and triterpenoids were again absent ( Fig. 3a). Only when the glycerol-treated leaf discs were subsequently extracted with chloroform were the triterpenoids released quantitatively. These extracts also contained aliphatic compounds although in significantly reduced proportions in comparison with those present in mechanically removed surface films ( Fig. 3b). The total quantities of aliphatic components and triterpenoids detected in the four consecutive treatments were similar to those achieved by chloroform extraction of the intact adaxial leaf surface.

Figure 3.

Figure 3.

Yields and compositions of SCL removed mechanically from adaxial surfaces of mature P. laurocerasus leaves. (a) and (b) Frozen glycerol was used in three consecutive treatments of a given surface area and the remaining SCL were finally extracted with chloroform. (c) and (d) Frozen water was used in three consecutive treatments of a given surface area and the remaining SCL were finally extracted with chloroform. (e) and (f) Collodion film was used in three consecutive treatments of a given surface area and the remaining SCL were finally extracted with chloroform.

Figure 3.

Figure 3.

Yields and compositions of SCL removed mechanically from adaxial surfaces of mature P. laurocerasus leaves. (a) and (b) Frozen glycerol was used in three consecutive treatments of a given surface area and the remaining SCL were finally extracted with chloroform. (c) and (d) Frozen water was used in three consecutive treatments of a given surface area and the remaining SCL were finally extracted with chloroform. (e) and (f) Collodion film was used in three consecutive treatments of a given surface area and the remaining SCL were finally extracted with chloroform.

Figure 3.

Figure 3.

Yields and compositions of SCL removed mechanically from adaxial surfaces of mature P. laurocerasus leaves. (a) and (b) Frozen glycerol was used in three consecutive treatments of a given surface area and the remaining SCL were finally extracted with chloroform. (c) and (d) Frozen water was used in three consecutive treatments of a given surface area and the remaining SCL were finally extracted with chloroform. (e) and (f) Collodion film was used in three consecutive treatments of a given surface area and the remaining SCL were finally extracted with chloroform.

Treatment with frozen glycerol might, in addition to the mechanical removal of waxes, cause dissolution of these highly lipophilic compounds. In order to check this possibility, the experiments were repeated using distilled water instead of glycerol as cryo-adhesive to remove the SCL from the plant surface. Due to its high polarity, water is not a suitable solvent for waxes and thus, dissolution effects can be excluded. The composition of the wax mixture removed by frozen water was very similar to that obtained with glycerol as cryo-adhesive ( Fig. 3c). In contrast to glycerol, the mechanical removal of surface layers with frozen water yielded lower amounts of SCL in the first treatment and more of the same compounds in consecutive ones ( Fig. 3c). The total coverages of aliphatic SCL removed in all three treatments were identical in both experiments. Triterpenoids could not be detected in either of the three water-generated samples but were again released quantitatively in a subsequent chloroform extraction ( Fig. 3d). These findings demonstrate that dissolution of internal wax fractions did not occur either with glycerol or with water whereas glycerol proved to be the more effective adhesive.

These results could now be compared with data obtained with a traditional and widely used method for the mechanical removal of SCL from leaf surfaces using collodion ( Haas & Rentschler 1984). This method employs films of nitrocellulose that are deposited from an organic solution on the plant surface and, after evaporation of the solvent, peeled off together with adhering cuticular waxes. When this method was applied to adaxial leaf surfaces of P. laurocerasus the relative composition of the aliphatic compound classes removed resembled that found for the treatments with water or glycerol ( Fig. 3e). However, in contrast to the previous experiments, the collodion treatments additionally yielded considerable amounts of triterpenoids. Due to this effect, the total amount of material removed in consecutive collodion treatments was higher than in the experiments with cryo-adhesives. On the other hand, the repeated application of collodion left relatively high quantities of the aliphatic components on the surface that could subsequently be extracted with chloroform ( Fig. 3f).

The additive effect of solvent extraction and mechanical removal of waxes was tested using a combined application of chloroform and frozen glycerol. Defined areas of individual P. laurocerasus leaves were pre-treated with varying amounts of chloroform and then the surface waxes were removed using glycerol as cryo-adhesive. Other areas of the same leaves were treated either with glycerol as cryo-adhesive or with collodion. Relative amounts of SCL compound classes in these reference samples were identical to the results shown in Fig. 3. The consecutive application of chloroform and frozen glycerol yielded SCL amounts intermediate between those of both reference treatments. The percentage of compound classes removed with frozen glycerol was independent of the previously applied volume of chloroform. Most remarkably, chloroform pre-treatment in all cases increased the triterpenoid portion in SCL amenable to cryo-adhesive treatment from 0 to 2–3% whereas a single collodion treatment yielded SCL containing 6% triterpenoids.

The mechanical removal of superficial wax films from leaves of P. laurocerasus was visualized by SEM. The native adaxial leaf surface had an irregularly granulated appearance and was (probably due to the underlying epidermal cells) slightly undulated ( Fig. 4a). After mechanical removal of SCL with glycerol as cryo-adhesive, the cuticular surface became much smoother ( Fig. 4b, c). The circular contact area of a droplet of cryo-adhesive could easily be detected. It was clearly delineated from the surrounding area where the granular suface structure was still intact ( Fig. 4c, d). At higher magnification the boundary between both areas appeared as a rough line. When the film of epicuticular waxes was removed and deposited onto glass slides its granular surface structure was indistinguishable from its native state ( Fig. 4e). Along the edge and in small holes and cracks in the deposited film, platelets were bent up and, hence, the cross-section of the surface film was exposed ( Fig. 4f). The corresponding film thickness was estimated to be ≤ 200 nm.

Figure 4.

Scanning electron micrographs of adaxial P. laurocerasus leaf surfaces. (a) Untreated control (bar 20 μm). (b), (c) and (d) Outer segment of the leaf area that had been treated with frozen glycerol (lower half) and the adjacent area that had been frozen but not treated with glycerol (upper half) (bars 200 μm, 20 μm, and 2 μm, respectively). (e) Film of epicuticular waxes (lower half) transferred onto glass (upper half) using frozen glycerol (bar 20 μm). (f) Detail of (e) showing bent platelets near the rim of the deposited film (bar 2 μm).

The above-mentioned experiments showed striking differences in the chemical composition of waxes at the outermost surface and in the interior of adaxial leaf cuticles of P. laurocerasus. The progressive development of this layered wax structure during leaf ontogeny was exemplified by analysing the spatial arrangement of SCL on immature leaves. As this involved the mechanical removal of surface waxes only leaves with a certain thickness and surface area could be used. Hence, P. laurocerasus leaves were harvested in their final stage of blade expansion and treated with frozen glycerol. In the resulting wax mixtures only small amounts of triterpenoids were detected ( Fig. 5a). Conversely, they constituted most of the SCL in lower parts of the cuticle that were accessible only by chloroform extraction ( Fig. 5b). Thus, the localization of triterpenoids within the cuticles of immature and mature leaves was very similar. On the other hand, both developmental stages differed markedly in the amounts of SCL released by mechanical removal and solvent extraction. It should also be noted that younger leaves predominantly contained alcohol acetates instead of alkanes in the mechanically amenable outer parts of their adaxial cuticles.

Figure 5.

Yields and compositions of SCL removed mechanically from adaxial surfaces of immature P. laurocerasus leaves. (a) Series of three consecutive treatments of a given surface area using frozen glycerol. (b) Chloroform extraction of the remaining lipids and total extraction of reference areas.

DISCUSSION

In the present investigation a method that had recently been developed for the non-destructive removal of wax crystals from cuticular surfaces ( Ensikat, Neinhuis & Barthlott (2000) was modified in order to allow the chemical analysis of the removed lipophilic components. Preliminary experiments had shown that single application of the procedure on a defined surface area gave reproducible yields (measured as mg m−2) and that individual parts of the surface could be processed repeatedly without mechanical damage of the underlying tissue. Hence, this method enabled the stepwise removal of SCL, starting at the surface of the cuticle, and the chemical analysis of all the corresponding preparations. The yields achieved in respective steps could be used to evaluate the relative thickness of the removed layer(s).

The method was to be applied in investigations of plant cuticles that lack epicuticular wax crystals. Under these circumstances the surface film of epicuticular waxes should be directly accessible to mechanical removal. Leaves of Prunus laurocerasus were chosen as a model system because they have glossy adaxial and abaxial surfaces, i.e. they are devoid of epicuticular wax crystals. Cuticular membranes can easily be isolated from both sides of the leaves and the adaxial cuticles of this hypostomatic species have widely been used in permeability studies ( Schreiber et al. 1995 ; Kirsch et al. 1997 ). It was therefore of special interest to analyse the layered structure of the SCL in this model system, hence to verify the presence of epicuticular waxes, to quantify them and to differentiate them from intracuticular SCL.

Both glycerol and water were used as cryo-adhesives for removing cuticular waxes from the adaxial leaf surface of P. laurocerasus. Although both methods yielded different amounts of SCL in the first treatment, they gave very similar results for the total quantities removable by repeated treatments. In individual treatments the efficiency of the adhesive depends on its polarity whereas the total yield after multiple treatments is independent of the chemical nature of the removing agent. The constant overall yield must be due to mechanical instead of chemical effects thus proving the presence of a sharp, physically resistant boundary. It is very likely that the physical boundary thus characterized in the present experiments is identical with the outer limits of the cutin matrix. Hence, the results of this investigation show for the first time that the definition of epi- and intracuticular components is meaningful as referring to their localization on and in the cutin matrix, respectively.

Adaxial cuticles of P. laurocerasus leaves are covered by approximately 130 mg m−2 of epicuticular waxes whereas standard extraction procedures yielded 280 mg m−2 of SCL. Although the epicuticular waxes consisted entirely of aliphatic compounds, the intracuticular SCL were largely made up of triterpenoids (63%). Epicuticular waxes accounted for less than 4% of the cuticular material of 3330 mg m−2 ( Schreiber & Riederer 1996). Based on crystallographic data of individual long-chain compounds ( Small 1984) and on density values of alkane blends ( Le Roux 1969) it can be supposed that the mixture of aliphatic P. laurocerasus SCL has a solid-state specific weight between 0·8 and 1·0·106 g m−3. A yield of 130 mg m−2 of epicuticular waxes hence corresponds to a layer thickness of 130–160 nm. Assuming that the epicuticular film consists of all cis-configured wax molecules which are packed perpendicular to the leaf surface ( Sitte & Rennier 1963) it can be only 35–45 molecules thick ( Small 1984).

SEM showed that the granular appearance of P. laurocerasus leaves is not altered during treatment with liquid nitrogen. When the epicuticular wax film was transferred to glass slides using glycerol as cryo-adhesive the granular appearance was retained. At the same time, on the leaf a smooth surface was generated that was lined by a circular rim. Hence, the presence and the approximate dimensions of the epicuticular wax film could be visualized both as positive and negative imprints ≤ 200 nm high. This value agrees well with the thickness deduced from epicuticular wax yields.

Consecutive mechanical and extractive removal of SCL from P. laurocerasus leaves enabled separate analyses of epi- and intracuticular compounds, respectively. The qualitative and quantitative results can now be compared with those from solvent extraction of the same plant material using standard methods. The latter yielded mixtures representing a qualitative and quantitative blend of epi- and intracuticular compounds. Hence, chloroform molecules had access to both epi- and intracuticular SCL. It is therefore inappropriate to designate solvent extracts as ‘epicuticular waxes’, as has been common practice in the past literature.

Haas & Rentschler (1984) had devised a method for discriminating between epi- and intracuticular SCL by mechanical removal of surface waxes using collodion. In the present investigation their protocol was applied to P. laurocerasus leaves to acquire reference data. As a result, triterpenoids as well as aliphatic compounds, i.e. markers for intra- and epicuticular waxes, respectively, were found to be both associated with the nitrocellulose films removed and in the remaining waxes. Epi- and intracuticular waxes appeared to be cross-contaminated although they could clearly be separated using cryo-adhesive techniques. Hence, the partial mixing of epi- and intracuticular lipids must be an artifact generated by the application of collodion. Our results nonetheless indicated that the collodion method yields wax samples that are enriched in compounds originating from the leaf surface. Collodion films show a selectivity for the removal of epicuticular SCL similar to that of mild solvent extraction of intact surfaces.

The partial mixing of wax layers upon collodion application could be due to the presence of organic solvents (ether : ethanol 3 : 1) during the formation of the nitrocellulose film. In order to test this hypothesis we simulated combined solvent extraction and mechanical removal of waxes by pre-treatment with chloroform and subsequent treatment with glycerol as cryo-adhesive. Incubation with the solvent rendered triterpenoids amenable to mechanical removal, i.e. chloroform washed part of the intracuticular compounds to the surface. Consequently, a similar solvent effect is likely to act when collodion is applied to leaf surfaces. It should also be noted that collodion film samples contained more triterpenoids than chloroform/glycerol-treated samples, irrespective of the chloroform volume applied. Possibly the synchronous presence of both a solvent and the removing agent or the relatively high polarity of the nitrocellulose film had enhanced the yield of intracuticular compounds.

In numerous investigations a discrimination between epi- and intracuticular SCL had been attempted by variation of solvent extraction methods ( Silva Fernandes et al. 1964 ; Holloway 1974; Baker & Procopiou 1975; Baker et al. 1975 ; Svenningson 1988; Garrec et al. 1995 ). SCL released by mild (surface) extraction of the intact tissue were considered as epicuticular components whereas thorough extraction of isolated cuticular membranes was assumed to yield intracuticular waxes. Our results for P. laurocerasus cuticles showed that differential extraction can indicate chemical differences between epi- and intracuticular SCL but is not selective enough for their accurate quantification. Instead, the methods described here allow for the first time a sharp distinction between epi- and intracuticular SCL.

The layered structure of cuticular waxes in adaxial cuticles of P. laurocerasus could be due to the sequential formation and/or deposition of the different compounds. In this case the aliphatic compounds found at the surface of mature leaves should appear first, i.e. already in immature leaves, whereas the triterpenoids would subsequently be deposited in a lower layer. This hypothesis was tested by comparing the spatial arrangement of SCL in adaxial cuticles of immature and mature leaves. Young (but fully expanded) leaves gave higher yields of extractable SCL than mature leaves. This could be due to the immature state of the cutin matrix in respective cuticles that would give solvent molecules relatively easy access for the (mild) extraction of waxes. It was also found that leaves of both developmental stages had similar distributions of triterpenoids and aliphatic compounds, but the composition of the aliphatic compound classes differed. This implies that intracuticular SCL remained unaltered whereas the chemical composition of outer parts changed. Consequently, chemical differences between SCL located in inner and outer parts of the cuticle cannot be explained by ontogenetic programmes but must arise from spontaneous separation and diffusion of the compounds.

Cuticles play a pivotal role in limiting the loss of water across the large surface that plants expose to the atmosphere ( Schreiber, Kirsch & Riederer 1996). It has been shown that SCL alone constitute the cuticular transpiration barrier ( Schönherr 1976), that the permeability of the barrier is independent of the overall thickness of the cuticle ( Becker, Kerstiens & Schönherr 1986), and that the most effective part of the barrier resides at or near the cuticle surface ( Schönherr & Riederer 1988). To which degree the transpiration barrier is located in the epicuticular wax film can now be tested by comparing cuticular transpiration rates before and after mechanically removing the epicuticular layer.

Fungi as well as herbivorous insects are known to recognize host plants by probing the chemical nature of their surfaces ( Städler & Buser 1984; Carver & Thomas 1990; Carver et al. 1990 ; Podila, Rogers & Kolattukudy 1993; Flaishman, Hwang & Kolattukudy 1995; Kolattukudy et al. 1995 ; Hegde & Kolattukudy 1997). Obviously, only the outermost cuticular layers will be relevant for chemical recognition. When fungal development or insect behaviour after probing is related to specific plant constituents, then it has to be verified that candidate compounds are exposed on the very surface of the plant, i.e. that they are present in the epicuticular layer. The parts of the cuticular compounds that were shown to influence insects and fungi upon in vitro contact might not be exposed in vivo and would consequently not serve as contact allelochemicals. Previously, only mixtures of both intra- and epicuticular waxes could be analysed and hence the precise location of potential signalling compounds could not be tested. The methods developed here allow for the first time a specific analysis of those compounds residing at or near the plant surface.

For P. laurocerasus leaves large differences between the chemical compositions of epi- and intracuticular SCL were found. At first sight, it is surprising that the triterpenoid acids, known to possess insecticidal and fungicidal properties, were not located at the surface where they could serve as a first line of defence. On the other hand it ought to be advantageous for the plant to sequester compounds that could serve as specific clues to insects and fungi within the cuticle. Consequently, the film of epicuticular SCL might be an efficient camouflage because it consists of aliphatic wax compounds that are present on virtually all plant surfaces. According to this model, on leaves of P. laurocerasus the triterpenoids in the intracuticular SCL might serve as defence against test biting insects whereas the aliphatic compounds in the epicuticular layer would serve as a defence against contact-probing insects and fungi.

In the current investigation the presence of a thin film of epicuticular waxes on the leaves of Prunus laurocerasus was demonstrated and its thickness as well as its chemical composition were assessed. It remains to be determined how far these findings are specific for this plant system or whether they can be repeated for other species. First results from on-going studies show that surface waxes of other plant systems, for example, in the Rosaceae, Moraceae, Liliaceae and Brassicaceae, are arranged in chemically and mechanically distinct layers comparable to those on P. laurocerasus. Respective analytical results will help to reach a better understanding of the chemical surface characteristics that insects and fungi use as sign stimuli. In further experiments the methods presented here can be used to manipulate epicuticular wax films in order to test their ecological and physiological functions.

ACKNOWLEDGMENTS

The authors are indebted to Professor Dr W. Barthlott, Dr C. Neinhuis, Dr I. Meusel and H.-J. Ensikat (Universität Bonn) for technical advice on one of the methods used. Technical assistance by the Department for Electron Microscopy and by the Botanical Garden of the Universität Würzburg is gratefully acknowledged. We would also like to thank Dr C. E. Jeffree (University of Edinburgh) for critical discussions on an earlier version of this manuscript. Financial and instrumental support for this work was granted by the Sonderforschungsbereich 251 ‘Ökologie, Physiologie und Biochemie pflanzlicher und tierischer Leistung unter Stress’ and by the Fonds der Chemischen Industrie.

Ancillary