Camelina sativa, (L.)

Razeq, Fakhria M., Kosma, Dylan K., D, França, ebora, Rowland, Owen & Molina, Isabel, 2021, Extracellular lipids of Camelina sativa: Characterization of cutin and suberin reveals typical polyester monomers and unusual dicarboxylic fatty acids, Phytochemistry (112665) 184, pp. 1-11 : 2-5

2.1. Amounts and ultrastructure of C. sativa extracellular lipid-based polymers

Cutin monomers were isolated after depolymerization by NaOMe-catalyzed methanolysis of solvent-extracted dry residues from whole leaf, stem and flower tissues ( Jenkin and Molina, 2015; Molina et al., 2006). Quantitative analyses by GC-FID, using internal standards, showed concentrations of ca. 2.2, 0.8 and 1.8 mg g 1 DW of total identified monomers in leaf, stem and flower, respectively ( Table 1 ). There is an apparent discrepancy with one published report on C. sativa leaf cutin, where the cutin concentration per leaf area unit was four times larger than the amount reported here ( Tomasi et al., 2017). However, that difference is mostly because of the large proportions of fatty acids derived from membranes, namely unsaturated fatty acids and 2-hydroxy fatty acids, which were thoroughly removed in our preparations. By comparison, the closely related species Arabidopsis thaliana had cutin monomer loads that were half and two thirds of the amounts found in C. sativa stem and flower depolymerizates, respectively (Li-Beisson et al., 2013), whereas the load of leaf cutin in Arabidopsis was comparable to that of C. sativa ( Franke et al., 2005) . However, C. sativa leaf adaxial cuticles determined by transmission electron microscopy (TEM) were about 70 nm-thick ( Fig. 1A ), which is more than two times thicker than A. thaliana cuticles ( Franke et al., 2005). Whereas the cutin amounts in leaves of both species are similar, Camelina sativa leaf wax coverage is at least four times larger than that of A. thaliana leaves ( Razeq et al., 2014); it is unclear whether differences in wax deposition between these species explains the differences observed in cuticle thickness. A larger proportion of cutan in C. sativa leaf cuticles would explain this difference, although we have not conducted a chemical characterization of such polymer to confirm this hypothesis. However, these comparisons among cuticles of different species should be taken with caution because the thickness of the cuticle varies with the experimental conditions, including plant growth conditions and fixation method employed (Guzm´an et al., 2014). Abaxial C. sativa leaf cuticles were slightly thinner (50 nm) than adaxial cuticles and showed one or two electron-translucent lamellae ( Fig. 1B ). The stem epidermis ( Fig. 1C and D ) was covered with a much thicker cuticle (about 220 nm), correlating with a cutin amount per area unit that was more than 10 times higher than the cutin coverage in leaves ( Table 1 ). Stem cuticles presented electron-translucent lamellae that were randomly oriented, reminiscent of those observed in cuticles of Cuscuta gronovii ( Heide-Jørgensen, 1991; Jeffree, 1996). Differences in chemical composition between both cutin polymers could influence the ultrastructural arrangement (discussed below). It has been also suggested that the lamellar structure in some species can be attributed to alternating waxes and cutin, or to a combination of soluble and saponifiable lipids associated with cutan (reviewed by Jeffree, 1996). Flower parts were not analyzed by TEM, and petal cuticles observed by scanning electron microscopy (SEM) presented characteristic nanoridge structures also observed in Arabidopsis petals ( Fig. 1E and F ) ( Li-Beisson et al., 2009).

Suberized cell walls were identified in root periderm and seed coat sections ( Fig. 2 ). The periderm of 5-week-old roots had two layers of cells presenting characteristic red staining upon treatment with sudan red 7B ( Fig. 2A ) as well as blue autofluorescence ( Fig. 2B ). Observation by transmission electron microscopy revealed a lamellar ultrastructure in both the root endodermis ( Fig. 2C ) and periderm ( Fig. 2D ). Suberization was also evident in the palisade cell walls of seed coat sections ( Fig. 2 E-F), but in these tissues the darker bands of the lamellae were almost imperceptible and reminiscent to the lamellae observed in the outer integument of Arabidopsis seed coats ( Yadav et al., 2014). It is unknown whether differences in ultrastructure may result from different monomer chemistries or the arrangement of these units in the polymer. In particular, we did not find substantial differences between the chemical composition of the major C. sativa root and seed coat suberin ester-bound monomers (section 2.2). The total amounts of suberin monomers were 9.1 and 2.7 mg g 1 root and whole seed dry cell wall residues, respectively ( Table 1 ). It should be noted that the seed suberin load includes a small contribution of cutin monomers from the embryo and internal seed coat cuticles ( Molina et al., 2006). The root suberin monomer yield from C. sativa was comparable to the amount reported for Arabidopsis (7.2 mg g 1 DW; Li et al., 2007) whereas seeds contained about one third of the amount of lipid polyesters reported for Arabidopsis (8.6 mg g 1 seed residue; Molina et al., 2006). Differences in seed lipid polyesters that are normalized to total cell wall delipidated tissue may lead to inaccurately concluding that C. sativa seeds have lower suberin content than Arabidopsis . Given the smaller size of Arabidopsis seeds, it is expected that lipid polyesters, which are mostly localized to the seed coat, will be more concentrated in a preparation that contains more seeds (and thus higher seed surface area) in a given mass. In fact, when the polyester monomer amounts are normalized to surface area, C. sativa seeds contain ca. 33% higher loads than Arabidopsis ( Table 1 ; Molina et al., 2006).

2.2. Characterization of C. sativa lipid polyesters

The sodium methoxide (NaOMe)-catalyzed methanolysis method employed in this study followed by solvent partition to recover monomers and their subsequent conversion to trimethylsilyl (TMSi) ether derivatives for GC-MS(EI) analysis, yields distinctive ions that allow for monomer identification. Upon transmethylation, ester-bound acids are converted to corresponding methyl esters and free hydroxyl acids. If epoxides are present in the polymer, the oxirane ring is opened during solvolysis in basic methanol giving a substitution product with methoxy and hydroxyl groups in adjacent carbons that is readily identifiable after derivatization ( Holloway, 1974; Kolattukudy and Agrawal, 1974). Mid-chain substituted fatty acids are also commonly found in cutins. These monomers undergo α- cleavage on either side of the substituent (e. g. – CH [OTMSi]-) giving diagnostic mass spectra ( Kolattukudy, 1984). C. sativa cutins depolymerized with this method yielded both typical, previously reported monomers, and monomers that were tentatively identified by their mass spectra (MS) and retention time and have not been described previously in the literature.

Leaf, stem and flower cutin. In C. sativa cutin from the three organs studied, 10,16-dihydroxy 10,16-dihydroxyhexadecanoic acid (16:0) (or 10,16-dihydroxypalmitate; DHP) was either the predominant or a major component of the transmethylation products, representing 17, 10 and 47% of the monomers from leaf, stem and flower tissues, respectively, with smaller proportions of the co-eluting 9-hydroxy positional isomer ( Table 2 ). This monomer was identified by comparison to published mass spectra (Eglinton and Hunneman, 1968; Eglinton et al., 1968; Holloway, 1982; Holloway and Deas, 1971) and by its retention time on a nonpolar column ( Fig. 3A ). In a given cutin, monomers are often classified according to the chain-length of the most predominant monomers belonging to the C 16 or C 18 families ( Holloway, 1982). Despite the predominance of DHP, leaf or stem C. sativa cutin can be classified as mixed C 16 /C 18 cutins, since the C 16 functionalized monomers accounted for 28% and 17% of the leaf and stem monomers, respectively, whereas C 18 and odd chain monomers altogether constituted 28% and 36% of the released monomers from leaf and stem cutins, respectively ( Table 2 ). Flowers, on the other hand, had a typical C 16 cutin with 60% of the monomers corresponding to functionalized 16:0 fatty acids. The flower cutin monomer profile overlapped only partially with that of the leaf and stem cutin. For example, hydroxycinnamates and several minor in-chain substituted fatty acids were clearly absent from these tissues and some mid-chain hydroxylated fatty acids were exclusively found in flower cutin.

Root and seed coat suberin. Whole C. sativa mature seeds and 5-week-old roots were exhaustively delipidated and chemical depolymerization was carried out on cell wall-enriched dry residues ( Molina et al., 2006). Seeds are complex organs with several tissues containing extracellular lipid polymers ( Molina et al., 2008; De Giorgi et al., 2015), and the monomers identified and quantified on whole seeds may derive from one or more of such tissues. Most of the monomers released from C. sativa seeds can be classified as typical of suberin components, because 1) dimethyl octadecene-1,18-dioate and methyl 18-hydroxy-octadecenoate (both of which are low in the cutins characterized in this study) were predominant in seed samples, and 2) the monomer profile was very similar to that of the root suberin samples ( Table 3 ). Although dimethyl octadecene-1,18-dioate and methyl 18-hydroxy-octadecenoate are typical constituents of many suberins, including potato periderm ( Kolattukudy and Dean, 1974) and root tissues of Zea mays and Ricinus communis ( Schreiber et al., 2005) , both monomers are also found in aerial cuticles of phylogenetically related species, namely Arabidopsis and Brassica napus ( Bonaventure et al., 2004) . Therefore, any classification as cutin or suberin solely based on chemical analyses should be taken cautiously. However, our microscopical analyses of root and seed sections confirmed the presence of suberin on cell walls of root periderm ( Fig. 2A and B,D ), root endodermis ( Fig. 2C ), and seed coat palisade cells ( Fig. 2 E-F). As a result, both seed and root monomers are reported as suberin components in Tables 1 and 3 .

In seeds and roots, the suberin monomer profiles were similar. A representative root suberin chromatogram is shown in Fig. 3B . Small differences between these organs can mainly be attributed to the fact that seeds had a mixture of cutin and suberin monomers ( Molina et al., 2008). To evaluate the contribution of embryo cutin to the total seed polyester composition, embryo and seed coat tissues were separated using a density gradient centrifugation method established for Arabidopsis ( Perry and Wang, 2003) . After transmethylation and GC-MS analysis of both fractions, it was determined that 50% of the HFAs, in particular the trihydroxy fatty acids, 18-hydroxy-octadecenoate, and DHP are largely contributed by the embryo-enriched fraction ( Fig. 4 A-B) whereas 1,ω- diols, 1-alkanols (with exception of eicosan-1-ol), and dicarboxylic acids (DCAs; except for dimethyl octadecene-1,18-dioate) were mainly found in the seed coat-enriched samples ( Fig. 4 A,C,D,E ). Similarly, polyester analyses of B. napus seed tissues showed that the trihydroxy 18:1 fatty acid fraction was largely found in embryo cutin, which also contained 18:1 and 18:2 DCAs, the major cutin monomers in this species ( Molina et al., 2006). Although coumarate was exclusively found in the seed coat fraction, ferulate was present in both fractions, indicating that it is a common component of embryo and seed coat lipid polyesters in C. sativa ( Fig. 4 F ). In spite of the high proportion of caffeate observed in leaf cutin, we did not detect this monomer in any of the seed fractions analyzed. Thus, in both B. napus and C. sativa , the embryo cutin composition seems to be different from that of leaf cutin with a predominance of 18:1 trihydroxy fatty acid. Conversely, most of the monomers that characterize suberin polyesters were found in the seed coat-enriched fraction.

Figures

Fig. 1. Ultrastructure of Camelina sativa cuticles. Transmission electron microscopy images of cross-sections of adaxial (A) and abaxial (B) leaves, and top (C) and bottom (D) stems. Scanning electron microscopy images of adaxial (E) and abaxial (F) petal surfaces. Scale bars: 500 nm (A), 200 nm (B, C, D), and 10 μm (E, F). C, cuticle; CW, cell wall.

Fig. 2. Suberin deposition in roots and seed coats of Camelina sativa. Root cross sections showing suberized root periderm stained with Sudan Red (A) or viewed via blue-yellow suberin autofluorescence (B). Transmission electron microscopy (TEM) image of root endodermis (C) and TEM image of root periderm (D). TEM image of seed coat showing suberized palisade cell walls (E, F). Scale bars: 100 μm (A, B), 100 nm (C, D), 5 μm (E), and 500 nm (F). CW, cell wall; P, palisade layer; S, suberin.

Fig. 3. Annotated chromatograms of TMSi derivatives of C. sativa leaf cutin (A) and root suberin (B) monomers. Peak numbers correspond to monomers listed in Table 2 (cutin monomers) and Table 3 (suberin monomers). Internal standard (IS): 17:0 fatty acid methyl ester (IS1) and 15-hydroxy 15:0 fatty acid methyl ester (IS2). Asterisks indicate peaks of residual unsaturated fatty acids from membranes, not considered part of the polyester.

Fig. 4. Lipid polyester monomer distribution in seed tissues. Comparison of transmethylation products from whole seeds, embryo-enriched and seed coatenriched delipidated residues. (A) Relative content of cutin monomer classes. (B–G) Detailed seed coat, embryo and whole seed monomer composition in each component class, namely hydroxy fatty acids (HFA; B), 1,ω-Diols (C), primary alcohols (PA; D), dicarboxylic acids (DCA; E) and hydroxycinnamic acids (HCA; F). Error bars represent SE; n =3. Fatty acids did not present any particular distribution between seed tissues and are not included in this figure.