Intracuticular wax fixes and restricts strain in leaf and fruit cuticles

Introduction

The primary surface of all terrestrial plants is covered by a lipoid cuticular membrane (CM), which forms the critical interface between the plant and its environment (Jeffree, 1996). Cutin, wax and polysaccharides are the major constituents of the CM. Cutin is a polyester formed by oxygenated C16- and C18-fatty acids, cross-linked by ester bonds (Heredia, 2003). In some species, cutan, a polymethylene polymer, may also be present as part of the matrix (Jeffree, 1996; Bargel et al., 2006). Waxes occur both epicuticularly as surface deposits on the cutin matrix and as intracuticular waxes embedded within it. They comprise a complex mixture mostly of long-chain C20–C40 alcohols, aldehydes, fatty acids, and alkanes (Kunst & Samuels, 2003; Dominguez et al., 2011a). In some species, significant amounts of triterpenoids and flavonoids are also present (Kolattukudy, 1996; Kunst & Samuels, 2003; Samuels et al., 2008). Polysaccharides are located on the cell wall side of the cuticle and represent epidermal cell wall constituents, such as cellulose, hemicelluloses, and pectins (Jeffree, 1996).

The cuticle functions as a barrier against water loss and uptake, and invasion by pathogens (Kerstiens, 1996; Riederer & Schreiber, 2001; Kunst & Samuels, 2003). Cuticles may also have mechanical functions (Matas et al., 2004; Bargel & Neinhuis, 2005; Dominguez et al., 2011a). Maintenance of these barrier functions requires an intact CM. This is challenging for a nonliving biopolymer deposited on an enlarging leaf surface (2–3 wk) and is even more of a challenge on a fruit surface where rapid and continuing expansion occurs throughout development (3–5 months). Nevertheless, cuticle deposition in fruit is often limited to their early development (e.g. grape, Becker & Knoche, 2012; plum, Knoche & Peschel, 2007 and sweet cherry Knoche et al., 2004). Failure of the CM to cope with excessive expansion (strain) results in a build-up of stress (Knoche et al., 2004), the formation of microscopic cracks (Peschel & Knoche, 2005), and ultimately in periderm formation in the surfaces of apple and pear (‘russeting’; Faust & Shear, 1972), increased rain cracking in sweet cherry (Peschel & Knoche, 2005) and grape (Becker & Knoche, 2012), and infections with pathogens (Borve et al., 2000). The situation in fruit contrasts with that in leaves where comparable phenomena are unknown. Presumably, this is because leaf expansion is usually limited to the early phase of development when the rate of cuticle deposition is high.

Interestingly, the development of fruit cuticles is sometimes very substantial and they often contain large amounts of wax compared with their leafy counterparts. As wax acts as a supporting filler (Petracek & Bukovac, 1995) that also increases the rigidity of the cutin matrix (Zlotnik-Mazori & Stark, 1988; Bargel et al., 2006; Dominguez et al., 2011a,b; Takahashi et al., 2012), it is plausible that the high wax content of fruit CM may have a function in mitigating cuticular failure.

The objective of our study was to test this hypothesis by recording the relationship between the amounts of wax present in different cuticles and their rheological properties. We did this in isolated leaf and fruit cuticles of a number of species, quantifying the effect of wax extraction from the cuticle on the release of strain and on mechanical properties such as the stiffness, the maximum force, and the maximum strain. We include the leaf cuticle of the xeromorph, agave, in our study as its CM is thick and contains a large amount of wax. In this respect, it is similar to a fruit cuticle.

Materials and Methods

Experiments

Release of biaxial strain

To quantify the release of biaxial strain (ɛ_biaxial, %) following heating or wax extraction, a square pattern of four holes (c. 3.25 mm × 3.25 mm, hole diameter c. 0.5 mm) was made in untreated CM discs (8 mm diameter) using a custom punch. Next, the CM discs were photographed (×1.25) under a dissecting microscope (MZ10F; Leica Microsysteme GmbH, Wetzlar, Germany; camera DP71, Olympus, Hamburg, Germany; Software Cell^P, Olympus Soft Imaging Solution, Münster, Germany). They were then dried and either heat treated or solvent extracted to remove wax. Next they were rehydrated in deionized water at 22°C for 16 h and photographed again. The areas enclosed by the square patterns of the four holes were quantified before (A_CM) and after heating (A_CM80) and before (A_CM) and after wax extraction (A_DCM; n = 13–15). The strain releases upon heating () and upon wax extraction () were calculated as:

(Eqn 1)

A possible effect of hydrating CM and DCM discs on the change in disc area was quantified. For this experiment, apple was selected because apple CMs had the highest amount of wax among the species investigated. Digital photographs were prepared before and after a 16 h hydration period and the area of the entire CM and DCM discs were quantified on an individual disc basis (n = 15).

Tensile tests

Cuticular membranes discs (24 mm diameter) were trimmed using parallel-mounted razor blades to obtain CM strips (5 mm wide). These strips were then mounted in a frame of paper and masking tape (TesaKrepp; tesaWerk Hamburg GmbH, Hamburg, Germany). Unless specified otherwise, all specimens were hydrated by incubating in deionized water at 22°C for a minimum of 16 h before testing. Subsequently, frames were mounted between the clamps of a universal material testing machine (Z 0.5; ZwickRoell, Ulm, Germany; clamping distance L₀ = 10 mm) equipped with a 10 N standard force transducer (KAP-Z; Zwick/Roell). The frames were cut open and uniaxial tensile forces were applied. Specimens were strained at a rate of 1 mm min⁻¹ until failure. Applied forces (F, in N) and corresponding specimen lengths (L) were continually recorded. The stiffness (S, in N) was calculated as the maximum slope of a linear regression line fitted through a plot of force (N) vs strain (%/100). This S differs from the commonly used modulus of elasticity in material science (E; in MPa) in that it reflects the properties of the specimens as present in different thicknesses on the respective leaf and fruit surfaces because it is not corrected for differences in cross-sectional area and, hence, in thickness between species. The S value allows an analysis of the change in stiffness (ΔS) upon extracting the CM simply by subtracting the S determined for DCMs from that of CMs. The strain at maximum force (ɛ_max, %) was obtained from

(Eqn 2)

For a representative data set, this ɛ_max was within 98% of the ɛ at fracture. The maximum force (F_max) corresponds to the maximum force recorded, usually just before fracture. Data for specimens that failed in, or adjacent to, clamps and/or that had irregular force–displacement curves were excluded from the analyses as these may have been damaged during handling or mounting. The remaining specimens represented 91% of the total population (n = 802) of strips investigated. Using this procedure, the stiffness (S) and failure thresholds (F_max, ɛ_max) were quantified for CMs and DCMs taken from all species. Replication (n) ranged from 10 to 19.

In apple, where the effect of wax on CM rheology was most pronounced, S, F_max, and ɛ_max were also recorded for CM – ECW. Untreated apple CM (with ECW) and extracted CM (DCM) served as controls (n = 11–12).

The effects of temperature were quantified in apple by oven-heating CMs and DCMs to temperatures between 20 and 140°C for 16 h. Tensile properties were quantified at 22°C and 50% RH on dry specimens within 12 h of removal from the oven (n = 9–14).

To address the possibility that wax recrystallization occurred after heating (16 h at 80°C), apple CMs were held at ambient conditions (22°C, 50% RH) for 0.04, 0.5, 2, 7, 14, 28, and 56 d. Thereafter, tensile properties were quantified as described earlier. Unheated CMs served as controls (n = 14–15).

Differential scanning calorimetry (DSC)

Unheated apple CMs, heated CMs (CM80) (80°C for 16 h), DCMs, and isolated wax were investigated by DSC. After heat treatment, CM80 were held at 22°C and 50% RH for up to 204 d to allow for possible wax recrystallization. Samples were weighed on aluminium pans (each c. 2.5 mg) and the pans crimped. Wax samples for DSC were prepared from a chloroform : methanol extract of apple CMs. Aliquots were taken to dryness in the aluminum pans and the pans crimped. Samples were loaded into the DSC system (Pyris 1 DSC; Perkin-Elmer Instruments, Waltham, MA, USA) and scanned at a heating rate of 10°C min^–1 from 20 to 140°C and a cooling rate of −10°C min^–1 from 140 to 20°C. Thereafter, the entire cycle was repeated on the same sample using the same settings. Empty pans were used as reference (n = 3).

Attenuated total reflectance Fourier transform infrared spectroscopy (ATR FT-IR) measurements

Attenuated total reflectance Fourier transform infrared spectroscopy spectra were recorded using an IFS 66 spectrometer (Bruker Optics, Ettlingen, Germany) equipped with a liquid N₂-cooled mercury cadmium telluride (MCT) detector. The CMs were placed with their outer surfaces facing a trapezoidal germanium crystal (50.7 mm × 10 mm × 3.9 mm). This geometry allowed five internal reflections at the sample surface with an incidence angle of 45°. The crystal was mounted in a custom-built aluminum sample holder that was held at a constant temperature by a computer-controlled circulating water bath (Haake C25P Phoenix II, Karlsruhe, Germany). For single-beam spectra, averages were taken of 128 scans with a spectral resolution of 4 cm⁻¹. Spectra were recorded at 2°C intervals between 28 and 74°C, with an equilibration time of 15 min at each temperature. Temperatures were measured inside the cover plate of the sample holder using a Pt100 resistor (Omega Newport, Deckenpfronn, Germany). Final absorbance spectra were calculated from the single beam spectra using the spectra of the Germanium crystal without CM (Ge–air interface) at each temperature as a reference. All absorbance spectra were baseline-shifted to zero in a spectral region where no vibrational peak occurred. To determine the peak position of the CH₂-vibrational bands in a certain wavenumber interval, the second derivatives of the spectra were calculated and the ‘peak picking’ function included in the Bruker OPUS software was used.

Data analysis and presentation

Data in tables and in Figs 1 and 3–5 are presented as means and SEMs. Where error bars are not shown in figures, they were smaller than data symbols. Data were subjected to ANOVA (Proc GLM) or linear regression analysis (Proc REG) using SAS (version 9.1.3; SAS Institute, Cary, NC, USA). Percentage strain data were arcsine-transformed before ANOVA. Means were compared using Tukey's studentized range test (P < 0.05). Significance of coefficients of determination (R²) and of coefficients of correlation (R) at the probability levels 0.05, 0.01 and 0.001 are indicated by *, ** and ***, respectively.

Results

Mass per unit area of leaf CMs ranged from the bush lily (lowest) to the xeromorphic agave (highest). Values for the fruit CMs of apple were just below those of agave leaves (Table 1). Apple fruit CMs had the highest masses of wax per unit area both in absolute terms and in percentage wax content. Wax mass per unit area and wax content were lowest in tomato fruit CMs (Table 1). Across all species, CM and wax mass per unit area were positively and significantly related (R² = 0.72; P = 0.016).

Extracting the wax from CMs caused a significant decrease in area of the DCM discs of apple, pear, agave, and holly, indicating that a release of biaxial strain was associated with wax removal. There was little or no release of biaxial strain in ivy or bush lily (leaves), or tomato (fruit) (Table 1). Across both organs and all species, strain release was a positive, linear and highly significant function of the amount of wax extracted (R² = 0.92, P = 0.001; Fig. 1a). In apple, where the release of strain was largest, an essentially identical linear and positive relationship was obtained between the amount of wax extracted and the release of strain (R² = 0.97, P = 0.002; Fig. 1b). Heating apple CM to 80°C released a small but significant amount of strain (2.8% ± 0.4%). Hydrating apple CMs and DCMs increased the surface area, on average, by 2.6 (± 0.1) and 5.7 (± 0.2)%.

Uniaxial force–strain diagrams obtained in tensile tests of apple CMs and DCMs were approximately linear up to about half the maximum strains (Fig. 2). The stiffness S, as indexed by the slope of the force–strain diagrams, decreased upon wax extraction (Fig. 2). Furthermore, F_max decreased and ɛ_max increased (Table 2). There was little or no difference in S, F_max, or ɛ_max between CMs with and without ECW, indicating that the effect of wax extraction on CM rheology resided largely with the embedded cuticular wax (Table 3; Fig. 2).

Across all organs and species, the decrease in S upon wax removal was consistent (Table 2). Furthermore, the F_max of apple, pear, holly, and ivy decreased upon wax extraction. There was no significant change for tomato, bush lily or agave. The ɛ_max generally increased, the only exceptions being tomato (fruit) and ivy (leaves) (Table 2). Across all species and organs, linear and significant relationships were obtained between S or the change in S upon wax extraction (ΔS) and the amount of wax extracted (Table 4, Fig. 3a). Further, ɛ_max and the change thereof upon wax extraction (Δɛ_max) were closely related to the amount of wax extracted (Table 4, Fig. 3b). Coefficients of determination obtained with CM mass per unit area as the independent variable were consistently lower than those with wax mass (Table 4). There was no relationship between F_max or the change thereof and either CM, DCM or wax mass per unit area (data not shown). Interestingly, the release of biaxial strain (; Table 1) and the increase in ɛ_max following wax extraction (Δɛ_max; Table 2) were closely related (Δɛ_max (%) = −1.06 (± 0.18) ×  (%), R² = 0.85, P = 0.001).

To address a possible effect of wax crystallization, apple CMs were subjected to heat treatments before the tensile tests. Heating of apple CMs significantly decreased S and increased ɛ_max (Fig. 4), but had no effect on F_max (data not shown). The effect of heating increased with rising temperature up to 80°C and remained essentially constant between 80 and 140°C (Fig. 4). Higher temperatures discoloured the apple CM and caused significant mass loss (B. P. Khanal, unpublished). There was no effect of heating to 80°C on either S or ɛ_max, or on F_max of the DCMs (Fig. 4; F_max data not shown).

Holding apple CM80 samples at ambient temperature and humidity for several days gradually and partially restored their extensibility as indicated by an increase in S and a decrease in ɛ_max within 2–7 d (Fig. 5). Holding them beyond a 7 d period produced little additional effect. There was no significant effect on F_max (data not shown).

Differential scanning calorimetry thermograms of apple fruit CM showed endothermic peaks at 53.2 and 62.3°C during a first heating cycle. After cooling, a second heating cycle was imposed. Although the two critical peak temperatures were unchanged, the peak heights were markedly reduced (Fig. 6a). The peaks are reasonably ascribed to the melting of cuticular wax and the differences in peak areas between heating cycles to the presence of differing amounts of crystalline wax (Fig. 6a). The area beneath the melting peaks decreased when heat-treated CMs (80°C for 16 h) were subjected to DSC after 0.2 d of heating (Fig. 6b). However, the peak height of heat-treated CM recovered if held at laboratory temperature for a period of 14 or 204 d (Fig. 6c,d). The melting peaks became even more pronounced after extracted wax had been subjected to DSC (Fig. 6e). There were no endothermic peaks in thermograms of DCM (Fig. 6f).

The wavenumber (cm⁻¹) of the maximum absorbance of CH₂ symmetric (v_s) and asymmetric (v_as) stretching bands increased with increasing temperature (Fig. 7a). The maxima of the absorption bands of the v_s (CH₂) and v_as (CH₂) stretching vibrations showed sigmoidal increases in wavenumber at c. 50 and 58°C, respectively. These indicate melting of the long-chain aliphatic fraction of cuticular wax (Fig. 7a; Merk et al., 1998). In the subsequent cooling cycle, the absorption maxima shifted downwards in wavenumber, indicating recrystallization upon cooling (Fig. 7b). The temperatures determined from the sigmoidal increase in wavenumber coincided with the peak maxima in the DSC thermograms (Fig. 6). The peak of the CH₂ scissoring absorption band was split into two components at low temperature (c. 30°C). The splitting disappeared as the temperature increased to c. 50°C when the two absorption bands merged to a single band. This temperature corresponded to the first sigmoidal increase in wavenumber of the v_s (CH₂) and v_as (CH₂) stretching bands and the first DSC peak. A further sigmoidal decrease in the wavenumber of the CH₂ scissoring band was observed at c. 58°C, which is consistent with the second endothermic peak of the DSC and the largest increase in wavenumber for the CH₂ stretching bands. In the subsequent cooling cycle, the scissoring peak reappeared (Fig. 8b), but disappeared again when CMs were reheated (data not shown). The higher absorption intensity in the cooling cycle (Fig. 8b) or the second heating cycle (data not shown) as compared with the first heating cycle (Fig. 8a) was the result of improved contact between the CM and ATR crystal as a result of the heating. The increase in absorption intensity occurred during the first heating cycle exactly at the melting temperature of the wax, indicating that heating must have improved the contact between CM and the crystal.

Discussion

Our results demonstrate that wax fixes the strain in CMs and decreases their extensibility. They also show that the magnitude of this effect is a linear function of the amount of wax in the cuticle per unit area. In other words, expansion growth of the leaf or fruit epidermis creates a reversible strain in the overlying cutin matrix. The presence of wax in this matrix effectively ‘fixes’ the strain, converting a reversible strain into a strain that, under in vivo conditions in the plant, is irreversible and, hence, plastic. This conclusion is based on the following arguments. First, the release of biaxial strain in the CMs upon wax extraction was a highly significant, linear function of the amount of wax present across a wide range of leaf and fruit cuticles (Fig. 1a). Furthermore, when varying the amount of wax extracted from apple CMs by varying extraction time (1–1000 min) and temperature (22 vs 50°C), strain release was again a linear function of the amount of wax extracted (Fig. 1b). In fact, the regression lines fitted in both figures were essentially superimposed (Fig. 1b). Secondly, the decrease in S (ΔS) obtained in tensile tests was linearly related to the amount of wax extracted (Fig. 3a). Thirdly, the release of biaxial strain from CM discs and the increase in ɛ_max that occurred upon wax extraction bore a 1 : 1 relationship that was highly significant across the seven CM sources investigated. Fourthly, the ɛ_max increased following wax extraction (Table 3), with the increase being positively and linearly related to wax mass per unit area (Fig. 3b). Fifthly, heating CMs beyond the melting point of their low-melting wax constituents yielded effects on strain release and on the tensile properties S and ɛ_max that were qualitatively identical to those of wax extraction (Fig. 4). Artefacts arising from differences in hydration between CMs and DCMs because of a change in polarity of the specimens as a result of extraction of wax can be excluded. The change in surface area as a result of swelling of apple CMs and DCMs upon hydration was similar and small compared with the amount of strain released after extraction. Thus, it is reasonable to suppose that the stresses and strains that occur in the CMs during fruit development as the surface expands, and which are fixed by the concurrent deposition of wax in the cutin matrix, are released suddenly upon either the melting or the extraction of the wax. These data are consistent with a role for wax as a filler in the flexible network of the cutin matrix (Zlotnik-Mazori & Stark, 1988; Petracek & Bukovac, 1995; Dominguez et al., 2011a,b).

Because the removal of epicuticular wax had little or no effect on biaxial strain release, or on the tensile characteristics of the CM, we can infer that the CM's decreased extensibility is wholly the result of the wax embedded within the cutin matrix – the epicuticular wax is not involved.

Cuticular wax occurs in both amorphous and crystalline states (Reynhardt & Riederer, 1994; Bargel et al., 2006). The observation that heating the CM to 80°C produced an effect qualitatively identical to, but quantitatively smaller than, that of wax extraction on the release of biaxial stain and on tensile properties leads to the conclusion that the crystalline wax fraction accounts, in part, for the effect on tensile properties. Apparently, the component of elastic strain in the CM fixed by the cuticular wax was released either when the wax was melted or when it was completely extracted. Consistent with this conclusion is the observation that extending the heating period beyond 2 h did not produce any additional effects (data not shown). Also, the gradual recrystallization that occurred when holding previously heated CMs at ambient temperature for up to 56 d partly restored the effect of heating on CM rheology. This is also consistent with a decrease in water permeability of citrus cuticles during storage that has been attributed to a time-dependent recrystallization of cuticular wax (Geyer & Schönherr, 1990). Direct support for recrystallization comes from the increase with storage time in the endothermic peak of heat-treated CMs (Fig. 6b–d). Further, the disappearance upon heating and reappearance upon cooling of the splitting of the CH₂ scissoring bands observed in the FT-IR spectra may be attributed to a breakdown and re-formation of an orthorhombic chain packing (Merk et al., 1998; Fig. 8a–c). This change in crystal structure occurs at c. 50°C, which corresponds to the first endothermic peak in the DSC thermograms. The second endothermic peak and the shift of the absorption maxima of the v_s (CH₂) and v_as (CH₂) stretching bands in the FT-IR spectra at c. 58°C must be attributed to the melting of a major constituent of wax. The most likely candidate for this is the alkane fraction that constitutes c. 30% of total apple wax (Belding et al., 1998). The dominant alkane representative is nonacosane which, in the pure form, has transition and melting temperatures of 58.2 and 63.4°C, respectively (Schaerer et al., 1955; Belding et al., 1998).

That heating the CM did not produce the same quantitative effect on biaxial strain release as wax extraction is probably a result of the restricted temperature range we investigated. We would expect the heating effect to be equivalent to the wax extraction effect if all wax constituents had been melted during the heat treatment. Under these conditions, the elastic (and reversible) strain of the cutin matrix would have been released quantitatively. Unfortunately, exposing apple CMs to the melting temperature range of the ursolic acids (290–300°C; Ritter et al., 2001), which are another major (up to 70%) component of apple fruit wax (Belding et al., 1998), results in severe mass loss and discoloration (charring) of the cuticle (B. P. Khanal, unpublished).

The effect of heating on the increase in ɛ_max of CM is surprisingly large compared with the release of biaxial strain upon heating. Although this result was consistent among several experiments, the reason for it is unknown. It is speculated that an embedded wax crystal deforms the cutin network in its immediate vicinity, but less so further away. In this way it causes a localized stress and strain in the matrix. This phenomenon is recognized in filled rubbers where it is referred to as ‘strain amplification’ (Vincent, 1990; Westermann et al., 1997). Strain amplification by embedded crystalline wax is consistent with the observation that unheated CM fails at lower strains than heated CM. In this case, the melting wax crystals will partly release the strain in their vicinity.

Our results provide the first clear evidence that cuticular wax ‘fixes’ strain in the cutin matrix of leaf and fruit cuticles. This strain results from the expansion of the underlying surface during growth. The strain is ‘fixed’ because neither excision nor isolation of the CM results in significant strain release. However, upon wax extraction, this strain is released. Therefore, wax deposition in the cutin polymer of the cuticle on the expanding surface has effectively converted a reversible elastic strain into a strain that, in vivo in the plant, is irreversible and plastic. From an ecological point of view, this modification of the mechanical properties of the cuticle should be seen as a mechanism that operates to reduce cuticular failure in the expanding surface of a leaf or a fruit. However, a negative-going consequence of such ‘strain fixation’ by wax deposition is an increase in the stiffness of the cutin matrix that could well restrict further expansion.