Biosynthesis and secretion of plant cuticular wax

Progress in Lipid Research 42 (2003) 51–80
www.elsevier.com/locate/plipres
Review
Biosynthesis and secretion of plant cuticular wax
L. Kunst, A.L. Samuels*
Department of Botany, UBC, 6270 University Boulevard, Vancouver, BC, Canada V6T 1Z4
Abstract
The cuticle covers the aerial portions of land plants. It consists of amorphous intracuticular wax
embedded in cutin polymer, and epicuticular wax crystalloids that coat the outer plant surface and impart
a whitish appearance. Cuticular wax is mainly composed of long-chain aliphatic compounds derived from
very long chain fatty acids. Wax biosynthesis begins with fatty acid synthesis in the plastid. Here we focus
on fatty acid elongation (FAE) to very long chains (C24–C34), and the subsequent processing of these
elongated products into alkanes, secondary alcohols, ketones, primary alcohols and wax esters. The identity of the gene products involved in these processes is starting to emerge. Other areas of this field remain
enigmatic. For example, it is not known how the hydrophobic wax components are moved intracellularly,
how they are exported out of the cell, or translocated through the hydrophilic cell wall. Two hypotheses
are presented for intracellular wax transport: direct transfer of lipids from the endoplasmic reticulum to the
plasma membrane, and Golgi mediated exocytosis. The potential roles of ABC transporters and non-specific lipid transfer proteins in wax export are also discussed. Biochemical-genetic and genomic approaches in
Arabidopsis thaliana promise to be particularly useful in identifying and characterizing gene products
involved in wax biosynthesis, secretion and function. The current review will, therefore, focus on Arabidopsis
as a model for studying these processes.
# 2002 Elsevier Science Ltd. All rights reserved.
Contents
1. Introduction ............................................................................................................................................................. 53
2. Wax biosynthesis...................................................................................................................................................... 55
2.1. Contributions from biochemical and genetic approaches ............................................................................... 55
2.2. Biosynthesis of very long chain fatty acids...................................................................................................... 60
2.2.1. Pathways for elongation of the fatty acyl chains................................................................................. 60
2.2.2. Enzymes of the extraplastidial FAE and their functional organization .............................................. 61
2.3. Pathways for cuticular wax biosynthesis ......................................................................................................... 64
2.3.1. The acyl reduction pathway ................................................................................................................ 64
2.3.2. The decarbonylation pathway ............................................................................................................. 65
* Corresponding author.
0163-7827/03/$ - see front matter # 2002 Elsevier Science Ltd. All rights reserved.
PII: S0163-7827(02)00045-0
52
L. Kunst, A.L. Samuels / Progress in Lipid Research 42 (2003) 51–80
3. Wax secretion ........................................................................................................................................................... 66
3.1. Considerations of wax precursors as secretory products................................................................................. 66
3.2. Wax biosynthesis in the context of the epidermal cell..................................................................................... 67
3.2.1. FAS and thioesterase activities in the plastid produce long chain fatty acid wax precursors ............. 67
3.2.2. FAE to very long chain fatty acids is associated with the endoplasmic reticulum.............................. 68
3.2.3. Transport from the ER to the plasma membrane and export from the cell ....................................... 72
3.3. Transport of wax precursors across the cell wall to the cuticle ....................................................................... 74
3.3.1. Lipid transfer proteins......................................................................................................................... 74
3.3.2. The physical path of wax components through the epidermal cell wall.............................................. 75
4. Conclusions .............................................................................................................................................................. 76
Acknowledgements........................................................................................................................................................ 77
References ..................................................................................................................................................................... 77
Nomenclature
ACP
ACBP
CER
CoA
ER
FABP
FAE
FAR
FAS
FAT
KAS
KCS
LTP
NBD-DAG
NBD-PC
PM
SER
TAG
TEM
VLCFA
WS
X:Y
acyl carrier protein
acyl-CoA binding proteins
eceriferum; gene involved in wax production in Arabidopsis
coenzyme A
endoplasmic reticulum
fatty acid binding proteins
fatty acid elongase
fatty acyl-CoA reductase
fatty acid synthase
fatty acyl-ACP thioesterase
plastidial condensing enzyme
b-ketoacyl-CoA synthase
lipid transfer protein
1-acyl-2-(N-4-nitrobenzo-2-oxa-1,3-diazole)aminoacyldiacylglycerol
1-acyl-2-(N-4-nitrobenzo-2-oxa-1,3-diazole)aminoacylphosphatidylcholine
plasma membrane
smooth endoplasmic reticulum
triacylglycerol
transmission electron microscopy
very long chain fatty acids (fatty acids 520 carbons long)
wax synthase
A fatty acyl group containing X carbon atoms and Y double bonds
L. Kunst, A.L. Samuels / Progress in Lipid Research 42 (2003) 51–80
53
1. Introduction
Plant wax is the collective term used to describe lipid components of the cuticle which covers
the outer surface of aerial plant tissues. In addition, wax is associated with the suberin matrix in
underground and wound tissues [1], as well as lipids of the pollen and seed coat [2–6]. In plant
cuticles, wax is predominantly comprised of very long chain aliphatic lipids, but it also includes
triterpenoids and minor secondary metabolites, such as sterols and flavonoids. The physical and
chemical properties of cuticular wax determine functions vital for plant life. It restricts non-stomatal water loss, protects plants against ultraviolet radiation [7,8] and reduces water retention on
the surface of the plant thus minimizing deposition of dust, pollen and air pollutants [9,10]. In
addition, surface wax is believed to play important roles in plant defense against bacterial and
fungal pathogens [11] and has been shown to participate in a variety of plant-insect interactions
[12].
The terminology describing plant surface wax has been first defined by the early morphological
work (reviewed by Martin and Juniper [13] and Baker [14]) where the entire waxy coating is called
the ‘‘cuticular membrane’’. It consists of (from exterior inwards) epicuticular wax crystalloids or
films, the cuticle proper, and a cuticular layer which interweaves with the epidermal cell wall.
More recently, a chemically oriented definition of plant surface lipids has predominated, with the
cuticle described as a polyester matrix of hydroxy- and hydroxy epoxy fatty acids C16 and C18
long (cutin) embedded and overlayed with cuticular wax. For molecular genetic analysis of waxrelated genes, the wax has been defined as the ‘‘lipids which are removed from plant surfaces after
brief immersion in an organic [nonpolar] solvent’’ [15]. Phenotypically wax is considered to be the
glaucous or whitish bloom on plant shoots, which corresponds to crystalloid epicuticular wax
with light scattering capacity [16].
As our goal in this review is to integrate morphological, biochemical and molecular data, our
definition of wax includes aspects of each of the above. A working definition of the total plant
cuticle includes the cuticle proper, intracuticular wax and epicuticular wax (Fig. 1). The cuticle
proper consists of cutin, the insoluble, covalently cross-linked polymer that forms an electron
dense layer over the epidermal cells. The intracuticular wax is defined as the amorphous mixture
of lipids embedded in the cutin which links the cuticle with the cell wall matrix, and epicuticular
wax refers to the surface lipids forming the crystalloids or smooth film exterior to the cuticle.
Distinct chemical compositions of epicuticular vs. intracuticular wax have been demonstrated by
selective physical removal and analysis of epicuticular wax, followed by analysis of the remaining
intracuticular wax [17,18].
The focus of this review is the biosynthesis and secretion of the aliphatic very long-chain components of plant wax by epidermal cells. They include aldehydes, primary and secondary alcohols, alkanes, ketones and esters, which are derived from saturated very-long-chain fatty acids
(VLCFAs) with predominant chain lengths from 20 to 34 carbons (Table 1). Other lipids present
in plant cuticular wax, such as terpenoids, flavonoids and sterols will not be considered in this
review. For information on these wax components, readers are referred to a review by Walton
[19]. The biochemistry of b-diketones, and their derivatives, will not be covered either, as this
topic has been extensively reviewed by von Wettstein-Knowles [1,20], and Walton [19]. Instead,
recent advances in identification and functional characterization of genes involved in wax synthesis, and new developments in the cell biology of wax secretion will be discussed. Molecular-
54
L. Kunst, A.L. Samuels / Progress in Lipid Research 42 (2003) 51–80
genetic approaches in Arabidopsis thaliana, coupled with thorough biochemical analysis of
mutant phenotypes, have contributed key information to this field of research. Therefore, the
current review will emphasize the utility and potential of Arabidopsis for studies of wax
biosynthesis and function.
Fig. 1. Generic representation of transverse views of wax secreting epidermal cells, showing the components of the
cuticle, cell wall domains, and the nonphotosynthetic epidermal cell.
Table 1
Major classes of aliphatic cuticular wax components
Wax class
Chain length
range in plantsa
Predominant chain
lengths in Arabidopsis
stemsb
Content in
Arabidopsis
stemsc (%)
n-Alkanes
Secondary alcohols
Ketones
Fatty alcohols
Fatty acids
Aldehydes
Wax esters
C21–C35
C21–C35
C21–C35
C22–C34
C16–C34
C21–C35
C32–C64
C29,
C29,
C29
C28,
C30,
C30,
NA
38
10
30
12
3
6
1
C31, C27
C31, C27
C30, C26
C28
C28
NA, data not available.
a
Adapted from Kolattukudy [23] and Eigenbrode and Espelie [12].
b
Data are from Millar et al. [38] for the Columbia ecotype and Rashotte et al. [166] for the Landsberg erecta ecotype, listed in order of abundance.
c
Data are averages from two sources: Hannoufa et al. [28] and Jenks et al. [33] for the Landsberg erecta ecotype.
L. Kunst, A.L. Samuels / Progress in Lipid Research 42 (2003) 51–80
55
2. Wax biosynthesis
2.1. Contributions from biochemical and genetic approaches
Wax synthesis requires coordinated activities of a large number of enzymes believed to be
organized into multienzyme complexes that carry out the elongation of fatty acid wax precursors and catalyze the formation of a multitude of aliphatic compounds. The reactions of wax
production are localized in the epidermal cells [21]. The elucidation of wax biosynthetic pathways in a number of plant species has been possible due to the development of complex techniques, such as gas chromatography/mass spectrometry, that allowed the identification of wax
components (reviewed by Walton [19]), and elegant genetic and biochemical studies using radiolabeled tracers [1,22,23] (see Table 2 for reviews). Our knowledge of the enzymes involved in
wax biosynthesis, however, remains extremely limited for a number of reasons. Wax biosynthetic enzymes are associated with membranes, and conventional biochemical approaches
involving solubilization and purification are not well-suited for membrane-bound enzymes.
Furthermore, sufficient amounts of epidermal tissue for enzyme purification are difficult to
obtain, and substrates for enzyme activity assays are insoluble in aqueous buffers and not
commercially available.
Similarly, genetic approaches alone were initially not successful in revealing the identity of loci
involved in wax production. Wax-deficient mutants have been isolated in a number of plant species, including barley (Hordeum vulgare), Arabidopsis, maize (Zea mays), and Brassica napus. The
mutant loci in barley and Arabidopsis are termed eceriferum (cer), whereas loci identified in maize
and B. napus are referred to as glossy. Barley is the most extensively studied species, with 85
identified cer loci [24]. In Arabidopsis, there are currently 22 known cer mutants displaying varying degrees of waxlessness: cer1- cer10 are very glossy and cer 11- cer22 are less so [16,25]. Scanning electron microscopy has shown that the epicuticular wax crystals on many of these mutants
are absent or exhibit particular morphological changes (Fig. 2). Chemical analyses demonstrated
that mutants have a lower wax load than wild-type plants and/or contain wax with a dramatically
altered chemical composition. Most of the mutations, however, do not cause an accumulation of
metabolic intermediates due to a specific block in the pathway. Instead, there seems to be an
increased flux of precursors into an alternative branch pathway, resulting in more complex
changes in wax composition [15,22,26]. Thus, biochemical functions of the mutated gene products
are difficult to predict from the alterations in wax compositions in the mutants.
Complex changes in wax composition in the majority of wax-deficient mutants may be a consequence of the visual screens employed to identify lines with reduced wax loads. Only mutants
with significant decreases in epicuticular crystalloid accumulation would be isolated in such a
screen, most likely those with lesions in regulatory genes, or genes affecting epidermal development. Mutants affecting individual biosynthetic enzymes would be difficult to find due to their
more subtle biochemical phenotypes. Nevertheless, attempts to assign tentative functions to
genetic loci that affect wax accumulation based on their chemical phenotype have been made for
almost every species in which wax-deficient mutants have been identified. Biochemical phenotypes and suggested functions of the gene products of a small subset of Arabidopsis cer mutants
are shown in Table 3. Subsequent cloning of genes identified by mutations in these cer lines, and
sequence comparisons with sequences in the databases in most cases did not reveal significant
56
L. Kunst, A.L. Samuels / Progress in Lipid Research 42 (2003) 51–80
Table 2
Reviews of the wax literature including their subject emphases
Authors, reference
Subjects
W. Barthlott et al. (1998) [167]
Classification and terminology of epicuticular wax
crystalloid morphology with respect to taxonomy.
D. Post-Beittenmiller (1996) [66]
Branch points of wax biosynthesis with other pathways,
mutant analyses summary.
D. Post-Beittenmiller (1998) [15]
Cloned eceriferum genes of Arabidopsis and glossy
genes of maize.
M. Jenks and E.N. Ashworth (1999) [168]
Epicuticular wax with emphasis on horticultural aspects.
B. Lemieux (1996) [35]
Molecular genetics of epicuticular wax, emphasis on
Arabidopsis mutants.
M. Reiderer and L. Schreiber (1995) [169]
Wax as a transport barrier in cuticular membranes.
P. von Wettstein-Knowles (1995) [1]
Genetics and biochemistry of epicuticular wax, emphasis on
evidence supporting different pathways.
T.J. Walton (1990) [19]
Methods of chemical analyses of wax, cutin and suberin;
wax variation.
similarity to any gene with a known function. Alternatively, it resulted in different functional
assignments altogether of the CER gene products, as described below.
Reduced levels of alkanes, secondary alcohols and ketones and high aldehyde levels on the
surface of cer1 mutants were interpreted as being caused by a mutation in the gene encoding an
aldehyde decarbonylase [25], a key wax biosynthetic enzyme catalyzing the conversion of aldehydes to alkanes (Fig. 3). This prediction has not been confirmed after cloning of the CER1 gene
[27]. Moreover, CER1 has sequence similarity with EPI 23, a Kleinia odora epidermis-specific
protein [23] and GLOSSY1 (GL1) protein of maize (Zea mays) [28]. As these proteins contain five to
seven predicted transmembrane domains, it was suggested that they may be membrane transporters
involved in wax secretion [29].
The cer2 wax composition suggested a block in the elongation of C26 fatty acids [28]. When the
CER2 gene was isolated [30,31], the deduced amino acid sequence had no homology to any previously
known gene product, providing no clues of its biochemical function. A more recent search of the
nucleic acid and protein databases, however, revealed that CER2 may be a member of a large family
of coenzyme A-dependent acyltransferases [32], which operate by a catalytic mechanism related to
chloramphenicol O-acyltransferase (CAT). This prediction is based on two highly conserved consensus motifs, one of which is found in the catalytic site of CAT. If CER2 indeed encodes an acyltransferase, the specific role for such an enzyme in wax biosynthesis would have to be determined.
Attempts to deduce the biochemical function of CER3 were equally challenging. The surface
wax of the cer3 mutant is characterized by a pronounced increase in the chain length of primary
alcohols and alkanes, which suggested a defect in the release of fatty acids from fatty acid elongation complexes [28,23]. The isolation of the CER3 gene showed that while its deduced amino
acid sequence had no similarities with proteins of known function, it contained a putative nuclear
localization sequence (NLS) [34]. In addition, it had two predicted phosphorylation sites, known
L. Kunst, A.L. Samuels / Progress in Lipid Research 42 (2003) 51–80
57
to be involved in transport of proteins to the nucleus [35]. The presence of the NLS, together with
evidence that CER3 is expressed in all tissues examined, including the roots, led to a proposal
that CER3 may encode a novel regulatory protein. Recent database searches, using the predicted
amino acid sequence of CER3, identified significant matches with E3-ubiquitin ligases involved in
the N-end rule pathway from a wide range of organisms: Schizosaccharomyces pombe, Homo
sapiens, Mus musculus, Saccharomyces cerevisiae, Kluyveromyces lactis andCoenorhabditis elegans
Fig. 2. Scanning electron micrographs of the surface of Arabidopsis thaliana stems. Epicuticular wax crystalloids project from the surface, abundantly in wildtype (Landsberg erecta ecotype), and are either altered or sparse in mutants
cer1–cer6 (bars, 10 mm).
58
Table 3
Summary of characterized eceriferum mutants of Arabidopsis
Sequence/gene similarity
Described chemical phenotype
of stem waxb
% Of wildtype
stem wax load
Leaf wax
Wild type
–
High C29 alkane, C29 2 alcohol
and ketone levels
100
Normal leaf wax
cer1 (BRL3, BRL11)
72 kDa protein with 3 His-rich
motifs; sequence similarity
to maize GL1; putative
decarbonylase/membrane
receptor.
C30 aldehyde accumulation;
Reduced levels of both
decarbonylation and acyl
reduction pathway products
13
Normal leaf wax
cer2 (BRL5, BRL7, BRL9)
47 kDa protein; sequence
similarity to GL2 and acetyl
CoA-deacetylvindoline
4–0-acetyl transferase; putative
acyl-CoA transferase.
Accumulation of C26/28 acyl groups,
1 OH and wax esters; Reduced
levels of decarbonylation pathway
products
25
Normal leaf wax
cer3 (BRL1, BRL6)
90 kDa protein; nuclear
localization signal; putative
E3-ubiquitin ligase.
Accumulation of C30 1 OH; Reduced
levels of C29 alkane, 2 alcohol and
ketone
31
Decreased leaf wax
cer4 (BRL2, BRL10)
Not cloned;putative alcoholforming acyl-CoA reductase
Reduced levels of 1 OH and wax
esters; aldehyde accumulation; No
change in decarbonylation product
levels
69
Normal leaf wax
cer6 (BRL4)
56 kDa protein; Putative
VLCFA KCS involved in
elongation of wax precursors
Accumulation of C24-C26 products;
Reduced levels of both decarbonylation
and acyl reduction pathway products
19
Decreased leaf wax
a
b
Koorneef line in Landsberg erecta ecotype; additional alleles-Lemieux lines with T-DNA insertions in WS ecotype.
Based on Hannoufa et al. [28], Jenks et al. [33], Rashotte et al. [166].
L. Kunst, A.L. Samuels / Progress in Lipid Research 42 (2003) 51–80
Eceriferum mutant
(additional alleles)a
L. Kunst, A.L. Samuels / Progress in Lipid Research 42 (2003) 51–80
59
[36]. E3-ubiquitin ligases mediate the critical step of substrate recognition that results in substrate
ubiquitination and its targeting for degradation by proteasomes [37].
The only case in which the alteration of wax composition gave a relatively accurate prediction
of biochemical function is the cer6 mutant, where accumulation of C24 fatty acid suggested a
mutation affecting microsomal fatty acid elongation [33]. Similarly, in the CER6-suppressed
plants, the levels of C24 acyl groups increased by more than seven fold over wild type levels,
suggesting that the product of the CER6 gene (originally called CUT1) is involved in the elongation of acyl chains longer than 24 carbons [38]. Cloning of the CER6 gene [38,39] demonstrated
that CER6 encodes an elongase condensing enzyme required for the synthesis of very long chain
fatty acid precursors for stem and pollen wax production (see Section 2.2.2).
Fig. 3. Proposed metabolic pathways for wax biosynthesis in Arabidopsis stems. VLCFAs in the epidermal cells are
converted through two biosynthetic pathways to all of the aliphatic wax components. The decarbonylation pathway
produces aldehydes, alkanes, secondary (2 ) alcohols, and ketones, whereas the acyl reduction pathway yields primary
(1 ) alcohols and esters. Widths of the arrows indicate relative metabolic fluxes through different reactions. Reproduced, with permission, from Ref. [38].
60
L. Kunst, A.L. Samuels / Progress in Lipid Research 42 (2003) 51–80
The functions of Arabidopsis CER genes, which were first identified on the basis of their waxless
phenotypes by M. Koornneef and colleagues over a decade ago [16], can now be reassessed using
the large amounts of sequence information deposited in expressed sequence tag (EST) and genomic sequence databases. Access to this wealth of information, together with greatly expanded
range of methods to isolate and identify new plant genes, has created new opportunities to
examine genetic mechanisms that control wax production. The following section will summarize
progress in characterization of genes and gene products involved in fatty acid elongation and wax
biosynthesis.
2.2. Biosynthesis of very long chain fatty acids
2.2.1. Pathways for elongation of the fatty acyl chains
Aliphatic components of cuticular wax are synthesized in the epidermal cells from saturated
very long chain fatty acids (VLCFA) (Fig. 3). Formation of VLCFA wax precursors is a complex
process that is accomplished in two stages in different cellular compartments. The first stage, the
de novo fatty acid synthesis of C16 and C18 acyl chains, occurs in the stroma of plastids by
soluble enzymes forming the fatty acid synthase complex (FAS) [40,41]. During synthesis, the
growing acyl chain is attached to acyl carrier protein (ACP), an essential protein cofactor usually
considered a component of FAS. Fatty acid synthesis proceeds through a series of four reactions:
condensation of a C2 moiety originating from malonyl-ACP to acetyl-CoA, followed by the
reduction of b-ketoacyl-ACP, the dehydration of b-hydroxyacyl-ACP, and the reduction of trans2-enoyl-ACP. These reactions result in a fully reduced acyl chain extended by two carbons.
Three different types of FAS complexes are required for the synthesis of 18-carbon fatty acids in
the plastid. They differ in their condensing enzymes which have strict acyl chain length specificities: KASIII (C2–C4) [42], KASI (C4–C16) and KASII (C16–C18) [43]. In contrast, the two
reductases and the dehydratase apparently have no particular acyl chain length specificity and are
shared by all three plastidial elongation complexes [44].
The second stage of fatty acid elongation, the extension of the ubiquitous C16 and C18 fatty
acids to VLCFA chains that are used for the production of aliphatic wax components, is catalyzed by extra-plastidial membrane-associated multienzyme complexes, known as fatty acid
elongases (FAE) [45], (see Sections 2.2.2 and 3.2.2). Similar to the de novo fatty acid synthesis in
the plastid, a series of four consecutive enzymatic reactions results in a two carbon extension of
the chain: condensation between a CoA-esterified fatty acyl substrate and malonyl-CoA, followed
by a b-keto reduction, dehydration and a final enoyl reduction [46]. Since chain lengths of aliphatic wax components are typically in the range of 20–34 carbons, multiple rounds of elongation
are needed to extend the acyl chain to the desired length. The differential effects of inhibitors on
incorporation of radiolabeled precursors into wax components of various chain lengths, and
analyses of mutants with defects in fatty acid elongation, demonstrated that sequential acyl chain
extensions are carried out by several distinct elongases with unique substrate chain length specificities [20]. Specificity of each elongation reaction has been attributed to the selectivity of the
elongase condensing enzyme [47] (see Section 2.2.3). Consistent with the requirement for fatty
acyl precursors of various chain lengths for the synthesis of cuticular wax lipids, a large family of
21 elongase condensing enzyme-like sequences has been identified in the Arabidopsis genome. It is
not known how many of these putative condensing enzymes participate in wax production and
L. Kunst, A.L. Samuels / Progress in Lipid Research 42 (2003) 51–80
61
how many different condensing enzymes are needed for the elongation of a C18 to C34 fatty acid.
Since elongase condensing enzymes can catalyze multiple elongation steps [48], and because wax
biosynthesis takes place in several different plant tissues at different stages of plant development,
a detailed analysis of the substrate specificity and expression patterns of individual condensing
enzymes will be necessary to determine their exact metabolic roles.
2.2.2. Enzymes of the extraplastidial FAE and their functional organization
The membrane-bound nature of the fatty acid elongation enzymes has hindered both their
biochemical characterization, and studies into their functional organization and subcellular
localization by cellular fractionation. Partial purification of elongases from several plants [49–51],
resulted in each case in an enrichment of several protein bands. This observation suggested that
elongases do not consist of a single high molecular weight multifunctional polypeptide as proposed by Agrawal and Stumpf [52], but rather of several individual protein components.
Extensive mutant screens for changes in very long chain fatty acid composition in Arabidopsis
seeds resulted in isolation of a single class of mutants with a lesion in the gene designated FATTY
ACID ELONGATION1 (FAE1) [48,53,54]. A mutation in FAE1 affected only seed VLCFA
content, suggesting that FAE1 was specifically involved in VLCFA elongation reactions in the
seed [48]. Cloning [55] and subsequent characterization of the FAE1 gene [47] indicated that it,
indeed, encoded only a condensing enzyme and not all four activities of the elongase. Ectopic
expression of FAE1 in Arabidopsis and tobacco, and expression in a heterologous yeast system,
demonstrated that the condensation reaction is both the substrate-specific step and the rate-limiting step of FAE [47]. In addition, these experiments revealed that the introduction of FAE1
alone in these systems is sufficient for the production of VLCFAs, suggesting that the other three
components of the elongase (two reductases and a dehydrase) are found ubiquitously in all cells
[47].
The isolation of this first gene encoding an elongase condensing enzyme (b-keto acyl-CoA
synthase, abbreviated KCS) led to a rapid identification of additional sequences with a high
degree of sequence similarity to FAE1 in the Arabidopsis EST database. To date, three condensing enzymes: KCS1 [56], FIDDLEHEAD (FDH [57,58]) and CER6 [38,39] have been implicated
in the synthesis of VLCFA precursors for wax production in shoots. Interestingly, mutations at
the KCS1 and FDH loci did not result in visible surface glossiness on any organ, indicating that
surface wax load was not severely reduced. In contrast a major reduction of CER6 activity in cer6
mutants and sense-suppressed CER6 plants nearly abolished stem wax accumulation, and resulted in conditional male sterility [38]. The fact that KCS1 and FDH cannot ‘‘rescue’’ the severe
waxless phenotype and male sterility of cer6 mutants and CER6-suppressed plants indicates that
these condensing enzymes do not functionally overlap with CER6 in the stem and anther of
Arabidopsis. Furthermore, a recently identified VLCFA condensing enzyme CER60, with high
amino acid sequence identity to CER6 does not appear to significantly contribute to the synthesis
of stem and pollen surface lipids [39]. This may be due to a very low level of CER60 expression in
mature Arabidopsis tissues [59]. Taken together, these data suggest that CER6 is the major condensing enzyme for stem wax and pollen coat lipid biosynthesis. Accordingly, CER6 is highly
transcribed throughout development, exclusively in the epidermal cells of the Arabidopsis shoot
[59]. The only exception to the epidermal expression was observed in the anthers nearing maturity, in which CER6 transcript was localized in the tapetum.
62
L. Kunst, A.L. Samuels / Progress in Lipid Research 42 (2003) 51–80
Until very recently virtually nothing was known about the two reductases and the dehydratase
activities of the fatty acid elongase. Mutant screens in Arabidopsis were not successful in finding
mutants defective in these three enzyme activities, implying that these enzymes are either essential
or functionally redundant [47]. Although the dehydratase remains uncharacterized, our understanding of the reductases has benefited from recent work on the VLCFA biosynthetic pathway
in yeast and animal cells. In Saccharomyces cerevisiae, the b-keto acyl reductase and enoyl
reductase genes involved in sphingolipid fatty acid elongation have been identified and characterized, providing sequences for searching the Arabidopsis genome, and a heterologous mutant
system for testing candidate Arabidopsis genes.
A putative b-ketoacyl reductase of the FAE in S. cerevisiae was identified by searching the
genome for oxidoreductases or sequences containing a diagnostic NADH binding motif, predicted transmembrane topology and a dilysine ER-retention motif. Elongation activity assays of
mutants with deletions in identified genes demonstrated that the yeast gene YBR159w is required
for the activity of the microsomal elongase, and that it encodes a b-ketoacyl reductase [60].
Despite a pronounced reduction in elongation activity, Ybr159 loss-of-function mutants are
viable. This is surprising, because there are no functionally redundant sequences found in the
yeast genome. By contrast, two sequences homologous to YBR159w are present in Arabidopsis.
One of these sequences designated At-YBR159 (MIPS Accession number Table 4) functionally
complements the ybr159 mutant [60]. A second related sequence (MIPS Accession number
Table 4) which is 44% identical, 68% similar to the first polypeptide, detected in a BLAST query
of the Arabidopsis genome database, has not yet been characterized. A truncated Arabidopsis
cDNA corresponding to the At-YBR159 gene has been previously identified by Xu et al. [61] as
an ortholog of the maize GLOSSY8 (GL8) gene. In maize, a mutation at the GL8 locus results in
decreased levels of wax ester components longer than C24 [24,62,63,64]. The GL8 gene was isolated by transposon tagging and suggested to encode a reductase involved in fatty acid elongation
[61]. Based on evidence presented by Beaudoin et al. [60], it is clear that the maize GL8 functions
as a b-ketoacyl-reductase of the FAE involved in wax production.
The availability of these genes should allow us to determine the structures of the b-ketoacyl
reductase enzymes, the specific elongases they are associated with, and their site of action. Furthermore, experimental data of Beaudoin et al. [60] suggest that overexpression of YBR159w and
At-YBR159 results in moderate increases in the overall elongation activity. It will be interesting to
determine if this is also the case in Arabidopsis, or whether the condensing enzyme is the only
rate-limiting step of fatty acid elongation.
The TSC13 gene encoding the yeast enoyl reductase was isolated from a temperature sensitive
mutant defective in sphingolipid biosynthesis [65]. TSC13 is essential for viability, and as there
are no TSC13 homologous sequences in the S. cerevisiae genome, it is likely that Tsc13p catalyzes
a step in the fatty acid elongation cycle for acyl-CoA substrates of all chain lengths. Co-immunoprecipitation of the TSC13 gene product with the elongase condensing enzymes Elo2p and
Elo3p demonstrated that Tsc13p physically interacts with the other components of the elongation
machinery, supporting the hypothesis that FAE enzymes function as a large multimeric protein
complex in yeast. Tsc13p was localized to the endoplasmic reticulum, specifically at the nuclearvacuolar junction [65]. A similarity search using the deduced amino acid sequence encoded by
TSC13 identified a single sequence in the Arabidopsis genome with significant homology over its
entire length (30% identity, 46% similarity). Functional complementation of the yeast mutant
63
L. Kunst, A.L. Samuels / Progress in Lipid Research 42 (2003) 51–80
Table 4
Summary of database accession numbers for genes encoding wax biosynthetic enzymes
Known genes
GenBank accession
number
MIPS accession
number
References
KCS1
FIDDLEHEAD
CER6=CUT1
CER60
AF053345
AF337910
AF129511
At1g01120
At2g26250
At1g68530
At1g25450
[56]
[57,58]
[38,39,59]
[39,59]
AtYBR159
AtYBR159-like
NM_105441
NM_102292
At1g67730
At1g24470
[60]
At3g55360
[65]
17979129
At4g33790
[70]
MS2
15293283
15229989
15229991
15229993
15242233
15228993
12322016
At5g22500
At3g44540
At3g44550
At3g44560
At5g22420
At3g56700
At3g11980
[71,72]
LTP1
LTP2
LTP3
LTP4
LTP5
LTP6
LTP7
LTP10
15231093
15218659,
15218662,
15240494,
15240495,
15240496,
15240497,
15240498,
15240499,
15240500,
15218667,
15242082
AC005499
AF159799
AF159800
AF159801
AF159802
AF159803
AC005957
AI998609
At3g51970
At1g34490
At1g34500
At5g55320
At5g55330
At5g55340
At5g55350
At5g55360
At5g55370
At5g55380
At1g34520
At5g51420
At2g38540
At2g38530
At5g59320
At5g59310
At3g51600
At3g08770
At2g15050
At5g01870
Fatty acid elongation
b-keto acyl-CoA
synthases (KCS)
b-Ketoacyl reductases
Enoyl reductase
AtTSC13
Acyl reduction pathway
Fatty acyl-CoA
reductase (FAR)
At FAR-like
sequences
Wax synthase
Lipid transfer
proteins
[78]
[142]
64
L. Kunst, A.L. Samuels / Progress in Lipid Research 42 (2003) 51–80
with the Arabidopsis sequence and reverse genetic approaches in Arabidopsis should establish its
role in fatty acid elongation.
2.3. Pathways for cuticular wax biosynthesis
VLCFA products of the elongases generated in the epidermis are used for the synthesis of other
wax components. In most plants, including Arabidopsis, there are two principal wax biosynthetic
pathways: an acyl reduction pathway, which gives rise to primary alcohols and wax esters, and a
decarbonylation pathway, leading to the formation of aldehydes, alkanes, secondary alcohols and
ketones (Fig. 3). Comprehensive reviews describing the pathways of wax biosynthesis, including
the less widespread b-diketone pathway, are available [1,66] (see Table 2—summary of wax
reviews). Therefore, we will limit our discussion to the recent results which shed new light on our
understanding of the wax biosynthetic process, and point out where additional research is needed.
2.3.1. The acyl reduction pathway
The fatty acyl reduction of VLCFacyl-CoA esters to primary alcohols has been examined in a
variety of plants [67–70]. Partial purification of reducing activities from B. oleracea initially led to
a proposal that primary alcohol production is a two-step process carried out by two separate
enzymes, an NADH-dependent acyl-CoA reductase required for a reduction of VLCFAs to
aldehydes, and an NADPH-dependent aldehyde reductase required for a further reduction of
aldehydes to primary alcohols [68]. Subsequent work of Vioque and Kolattukudy [69] demonstrated that, at least in pea (Pisum sativum), this is not the case, and that alcohol formation from
VLCFA precursors is carried out by a single fatty acyl-CoA reductase (FAR). Furthermore, even
though alcohol biosynthesis likely proceeds via an aldehyde intermediate, the free aldehyde is not
released. Solubilization and biochemical characterization of a fatty acyl-CoA reductase involved
in wax ester formation in developing embryos of jojoba (Simmondsia chinensis), further substantiated this conclusion [70]. Cloning of the cDNA encoding the jojoba enzyme and expression
in E. coli and B. napus verified its biochemical role and resulted in the accumulation of fatty
alcohols. The pea and the jojoba alcohol-forming fatty acyl reductases are both integral membrane proteins, which appear to be associated with the ER [70] and have similar molecular masses
in the range of 56–58 kDa. FAR-related sequences from corn (Z. mays), rice (Oryza sativa), cotton (Gossypium hirsutum) and B. napus found in the public databases, and a family of eight FARlike proteins in Arabidopsis (MIPS accession numbers Table 4) suggest that alcohol-generating
reductases are ubiquitous in plants. One of these FAR-related Arabidopsis genes, MS2 (MIPS
accession number Table 4), encodes a tapetum-specific protein essential for pollen fertility [71,72].
A unique type of plastids found in tapetal cells called elaioplasts is known to accumulate neutral
lipids, such as TAGs, wax and sterol esters [73,74]. After lysis of the tapetal cells these wax and
sterol esters get deposited onto the surface of mature pollen. It is, therefore, attractive to propose
that MS2 protein is involved in the formation of fatty alcohols required for wax ester production
in the tapetum of the anther.
The final step of the acyl reduction pathway is the synthesis of wax esters, a reaction catalyzed
by a fatty acyl-CoA: fatty alcohol acyltransferase (wax synthase, WS). WS activity from several
different sources has been characterized biochemically [75–77], and described to be membranebound. However, very little was known about the proteins associated with the WS activity and
L. Kunst, A.L. Samuels / Progress in Lipid Research 42 (2003) 51–80
65
the genes encoding them. A major breakthrough was a recent report of a partial purification of a
WS polypeptide from jojoba embryos [78], which also resulted in identification of a cDNA
encoding the WS enzyme. The identity of the cloned cDNA was demonstrated by expression in
Arabidopsis, which resulted in high levels of WS activity in seeds of transgenic plants. Hydropathy analysis of the deduced protein sequence revealed seven to nine transmembrane domains
indicating that jojoba WS is an integral membrane protein. Twelve Arabidopsis sequences (MIPS
accession numbers Table 4) have considerable sequence similarity to the jojoba WS protein.
Several of these genes are differentially expressed in inflorescences and seeds of Arabidopsis [78].
One can anticipate that WS enzymes in these tissues would be involved in the formation of wax
esters for pollen coats, and synthesis of seed storage lipids, respectively. In addition, several of the
WS genes are probably required for cuticular wax production in Arabidopsis shoots. However,
the exact biochemical roles of these enzymes are yet to be determined.
Of the 22 wax-deficient cer mutants in Arabidopsis identified to date, only a mutation in the
CER4 gene has been suggested to affect the synthesis of products via the acyl reduction pathway.
A major reduction in primary alcohol and wax ester levels in the cer4 mutant, partially compensated for by an increase in aldehydes, has been attributed to a lesion in an aldehyde reductase of
the acyl-reduction pathway [28,33]. If, however, primary alcohols are generated from fatty acids
without release of an aldehyde intermediate, as suggested by Vioque and Kolattukudy [69], there
would be no aldehyde reductase enzyme in the acyl reduction pathway. Therefore, the cer4 phenotype can only be explained by a mutation in the alcohol-forming reductase of the acyl reduction pathway. The observed increase in aldehyde levels in the cer4 mutant may then be a result of
an increased flux of fatty acyl precursors into the decarbonylation pathway.
2.3.2. The decarbonylation pathway
The decarbonylation pathway is initiated by the production of aldehydes from VLCFA precursors by a membrane-bound fatty acyl-CoA reductase. This aldehyde-forming reductase has
been resolved and purified to homogeneity from pea leaves, and shown to be an enzyme with
distinct molecular weight (28 kDa) and biochemical properties from the reductase involved in the
production of primary alcohols [69]. However, a gene encoding this enzyme has not yet been
isolated. The generated aldehydes are then decarbonylated to odd-chain alkanes with a release of
carbon monoxide. This reaction is catalyzed by an aldehyde decarbonylase [79]. Two plant aldehyde decarbonylases have been studied in some detail, one from pea, and one from a green
colonial alga Botryococcus braunii [79–81]. Both enzymes need metal ions for their function and
are integral membrane proteins, with the pea decarbonylase proposed to be associated with the
cuticle (however, see Section 3.2.2) and the Botryococcus decarbonylase activity located in the
microsomes. However, technical difficulties encountered during solubilization of decarbonylases
have delayed molecular characterization of these enzymes and elucidation of the catalytic
mechanism of decarbonylation.
A molecular genetic approach to identifying and characterizing a plant decarbonylase is an
attractive alternative to conventional biochemical approaches. In Arabidopsis stems, products of
the decarbonylation pathway account for 90% of the total wax [38]. Thus, a lesion in a key
biosynthetic step of this pathway, such as decarbonylation, is likely to result in a significantly
reduced wax load, and such a mutant could be easily identified on the basis of its increased glossiness.
Surprisingly, of the 22 different genetic loci identified by mutation to date, only CER1 was reported
66
L. Kunst, A.L. Samuels / Progress in Lipid Research 42 (2003) 51–80
to specifically affect the decarbonylation pathway. Stem wax of cer1 mutants shows a dramatic
reduction in alkane, secondary alcohol and ketone levels, accompanied by an increase in aldehydes.
Based on this phenotype CER1 was suggested to encode an aldehyde decarbonylase [25,26,28].
Aarts et al. [27] cloned the CER1 gene, but because aldehyde decarbonylase genes have not been
cloned previously, the CER1 sequence was not useful in assigning a biochemical function to its
gene product. Moreover, as described in Section 2.1, CER1 shares considerable sequence similarity with the products of the maize GL1 gene and the epi23 gene from Kleinia odora, that have
been proposed to belong to a family of membrane-bound receptors [29]. Sequence similarity,
however, between GL1 and CER1 is restricted to the NH2-terminal region, and the CER1 protein
is much larger than the GL1 polypeptide. Furthermore, mutations in these two genes confer different phenotypic changes on the cuticular wax compositions of Arabidopsis and maize. Taken
together, these data suggest that, despite their similar structures, CER1 and GL1 proteins may
perform different roles in cuticular wax biosynthesis. Further work is clearly needed to resolve the
functional identities of these proteins.
Subsequent steps of the decarbonylation pathway, the hydroxylation of alkanes to secondary
alcohols and oxidation of secondary alcohols to ketones, have been demonstrated in broccoli
(Brassica oleracea) using radiolabeled alkanes [82,83]. An absence of secondary alcohols and
ketones in Arabidopsis cer1 mutant blocked in alkane synthesis supports the conclusion that these
wax components are derived from alkanes [25]. Our knowledge of the enzymes catalyzing these
reactions, however, is extremely limited. In broccoli, the introduction of the hydroxyl group
required oxygen and was inhibited by phenanthroline. Because this inhibition was reversed by
Fe2+, it was suggested that a mixed function oxidase was involved in this hydroxylation [83].
However, whether the alkane is indeed the substrate for hydroxylation is not known with certainty. No mutants affecting this part of the decarbonylation pathway have been identified in
Arabidopsis to date, and only a single gene in barley, cer-soh is known that affects the synthesis of
secondary alcohols [22,84]. Progress in this area awaits identification and characterization of
genes encoding the enzymes of this pathway.
3. Wax secretion
3.1. Considerations of wax precursors as secretory products
Wax components are unusual secretory products because of their hydrophobicity that sets them
apart from the ‘‘conventional’’ products of Golgi-mediated secretion, such as polysaccharides
and proteins. Hamilton [85] has reviewed the physical processes governing fatty acid transport
across lipid bilayers in each step of the transport including adsorption, transmembrane ‘‘flipflop’’, and desorption. Measurements of pure fatty acids crossing small unilamellar vesicles
(SUV) show that long chain fatty acids spontaneously partition into bilayers and flip-flop across
the bilayer on a time scale of seconds [86]. The tendency of fatty acids to partition into a nonpolar
environment has been quantified as a large negative drop in free energy, which increases logarithmically with chain length [87]. Thus, the longer chain fatty acids found during wax synthesis
will strongly favour a nonpolar environment. Conversely, measurements of the rate of fatty acid
removal or desorption from a model phospholipid bilayer showed a logarithmic decline in deso-
L. Kunst, A.L. Samuels / Progress in Lipid Research 42 (2003) 51–80
67
rption rate for each additional (–CH2)2 unit on the aliphatic chain [86]. As wax precursors mature
from long chains produced in the plastid to very long chain fatty acids and their derivatives, they
become increasingly hydrophobic and more difficult to remove from a nonpolar environment. It
is clear that very long chain, saturated hydrocarbons will not be freely soluble in an aqueous
environment, yet a typical epidermal wax-secreting cell consists of a collection of aqueous spaces
separated by membrane bilayers.
The nonpolar environments available to the very long chain fatty acid wax precursors are the
interior of organelle membrane bilayers or the interior of carrier protein hydrophobic pockets.
The cell must control the partitioning of fatty acids into membranes to avoid deleterious effects
on bilayer structure and to maintain the organelle’s unique lipid milieu. In mammalian cells,
abundant fatty acid-binding proteins (FABP) in the cytoplasm control the intracellular pool of
free long chain fatty acids, acting as a buffer to constrain fatty acid interactions to those required
by the cell [88–90]. FABP can be ‘‘membrane active’’, removing fatty acids from membranes
during transient collisions and delivering fatty acids to specific membranes and target enzymes
[91]. The mammalian cytosolic 14 kD FABP gene family is made up of a suite of tissue-specific
isoforms, but homologs of FABP have not been found in the S. cerevisiae or the Arabidopsis genomes.
A possible alternative to FABPs for binding hydrophobic VLCFA wax precursors is acyl CoAbinding proteins (ACBP). ACBPs have been characterized in a variety of organisms and demonstrated to bind cytoplasmic acyl-CoAs [92–94]. In Arabidopsis, both cytoplasmic acyl CoA-binding proteins [95,96], and membrane-associated forms of acyl CoA-binding proteins [97] have been
described. The latter were detected in developing siliques associated with the plasma membrane
and vesicles of epidermal cells of the embryos, which was interpreted as indicating that this acylCoA binding protein plays a role in embryo cuticle development. The role of acyl CoA-binding
proteins in binding CoA esters of very long chain fatty acids during wax synthesis has not been
experimentally tested.
3.2. Wax biosynthesis in the context of the epidermal cell
3.2.1. FAS and thioesterase activities in the plastid produce long chain fatty acid wax precursors
The typical epidermal cell that is active in wax secretion contains a large central vacuole surrounded by a very thin layer of cortical cytoplasm. The cytoplasm contains small leukoplasts,
smooth and rough endoplasmic reticulum (ER), Golgi, mitochondria, and cytoskeletal elements.
What cellular structures support the biosynthesis and transport of hydrophobic wax precursors?
The first cellular structure to support wax biosynthesis is the plastid, where fatty acid synthesis
occurs. Here, elongating acyl chains are esterified to acyl carrier proteins (ACP) in the stroma
[98]. Once they reach C16 or C18 in length, the partitioning of fatty acids for membrane glycerolipid synthesis from the fatty acids destined for wax biosynthesis must occur. This is an important control point in epidermal cells resulting in a massive flow of saturated fatty acids into wax
production [66]. For use in membrane lipids, acyl chains can either be transferred directly from
the ACP to glycerol-3-phosphate in the plastid (‘‘prokaryotic pathway’’) or they can be released
from the ACP by a thioesterase and exported from the plastid to the ER (‘‘eukaryotic pathway’’)
[41]. For wax production, the 16:0 or 18:0 acyl groups are released from ACP by an acyl-ACP
thioesterase and move to the ER for further elongation. The thioesterase activity involved in this
68
L. Kunst, A.L. Samuels / Progress in Lipid Research 42 (2003) 51–80
process has not been identified. This is partly due to the fact that it is not known with certainty
whether 16:0 or 18:0 are the precursor fatty acids exported from the plastid to be used for wax biosynthesis. One possibility is that 16:0 is released by the thioesterase in the plastid, with subsequent
elongation steps occurring in the cytosol by elongases anchored in the ER membranes. Alternatively, 16:0 could be elongated to 18:0 by 3-ketoacyl-ACP synthase II (KASII) activity in the
plastid [43], in which case 18:0 would be exported for extraplastidial fatty acid elongation, as
postulated by Post-Beittenmiller [66]. A single acyl-ACP thioesterase that preferred 18:0 was
purified from the wax producing epidermis of leek (Allium porrum) [99].
Acyl-ACP thioesterases have been classified into FatA and FatB groups based on gene
sequence analysis and substrate preferences [100,101]. FatA thioesterases are ubiquitously
expressed housekeeping genes with oleoyl-ACP (18:1-ACP) as their preferred substrate and with
minor activity towards 16:0 and 18:0. FatB thioesterases prefer saturated acyl-ACP substrates of
various chain lengths, and are further subdivided into a seed specific subgroup with specificity
towards medium chain (C10-C14) fatty acyl-ACPs and a widely expressed subgroup with specificity towards saturated C14–18 acyl-ACPs [101]. It is this latter subgroup, FatB1, which would
logically be expected to be involved in cleaving C16:0-ACP or C18:0-ACP for wax biosynthesis.
When the Arabidopsis FatB1 gene was expressed in E. coli, the preferred substrate determined in
vitro was 16:0> >18:1 >18:0>14:0-ACP [102]. Arabidopsis FatB1 is highly expressed in flowers,
with lesser expression in leaves, roots, and stems. FatB1 is responsible for generating the membrane pool of 16:0 found in flower membrane lipids, and overexpression of the FatB1 gene in
Arabidopsis led to accumulation of 16:0 in seeds [103].
It is important to note that the acyl-ACP thioesterases in the plastid are competing with the
acyltransferases and an 18:0 desaturase for substrates by mechanisms that are not well understood and, therefore, the in vitro activity of the thioesterase alone may not accurately predict
what its in vivo substrate will be. If the wax producing cell has relatively low desaturase activity,
high KASII activity and a typical FatB1 thioesterase, then 18:0 could still be the primary export
product of its plastids.
Following liberation from ACP, the long chain fatty acids (C16 or C18) are thiol-esterified to
CoA by acyl-CoA synthetases in the chloroplast envelope [104]. Trapping of fatty acids in an
intracellular pool by addition of CoA has been termed ‘‘vectorial acylation’’ since the presence of
the CoA esterified to the carboxyl group of the fatty acid prevents ‘‘flip-flop’’ of fatty acids back
across the bilayer [85].
3.2.2. FAE to very long chain fatty acids is associated with the endoplasmic reticulum
Long chain fatty acid wax precursors are further elongated by the microsomal elongase complexes to very long chain fatty acids (Section 2.2.1). The close physical association between plastids and endoplasmic reticulum may facilitate fatty acyl transfer from the plastid envelope to the
ER [105]. Because some domains of the ER are sensitive to fragmentation by aldehyde fixatives
used to prepare cells for electron microscopy, the most reliable reports of the plastid-ER relationship come from studies using cryofixation [106]. For example, cryofixed epidermal cells from Arabidopsis stems contain plastids that are often wrapped in a domain of ER (Fig. 4). The mode of transfer
of fatty acids between the plastid and the ER has not been established. Fatty acids can spontaneously
partition into lipid bilayers, so acyl-CoA esters destined for further elongation in the ER could partition into the endomembrane system. Alternatively, a protein carrier, such as acyl-CoA binding
L. Kunst, A.L. Samuels / Progress in Lipid Research 42 (2003) 51–80
69
protein, may be involved. In S. cerevisiae, deletion of the gene encoding ACBP led to accumulation of
C18:0, and dramatically reduced levels of C26:0, suggesting that ACBP is involved in transport of
acyl-CoAs to the fatty acid elongation system associated with sphingolipid biosynthesis [107].
There are several lines of evidence that FAE occurs in the endoplasmic reticulum (Fig. 5).
When the Arabidopsis condensing enzyme CER6, tagged with GFP, was transiently expressed in
the epidermis of Nicotiana benthamiana, the fluorescence appeared in a pattern which is characteristic of the endoplasmic reticulum (Fig. 6). In the cer6 mutant, a proliferation of the SER in
epidermal cells suggests a buildup of blocked substrate at this site (Samuels and Kunst, unpublished observations). Elongation activity has also been demonstrated in the microsomal fractions
carrying ER marker enzymes from epidermal cells in a variety of plants [49,50]. Jenks et al.
[108] quantitatively demonstrated the proliferation of ER membranes when wax synthesis was
Fig. 4. (A) The close physical relationship between the endoplasmic reticulum (ER) and plastid (arrow) is seen in this
transmission electron micrograph of cryofixed Arabidopsis epidermis. In the epidermal cells, the non-green plastids,
leukoplasts, are small and not photosynthetically active. CW, cell wall; v, vacuole; PM, plasma membrane. (B) Freeze
fracture replica of a tobacco (var. BY-2) suspension cultured cell showing the close association (arrow) between the ER
and the plasma membrane (PM). At these sites, the ER presses against the cytoplasmic face of the PM but does not
fuse with it. (bars, 0.25 mm). (Samuels and Staehelin, unpublished data).
70
L. Kunst, A.L. Samuels / Progress in Lipid Research 42 (2003) 51–80
Fig. 5. Wax biosynthetic pathways placed in the context of the epidermal cell. Fatty acid synthesis to C18 occurs in the
plastid, and fatty acid elongation to VLCFA takes place in association with the membranes of the endoplasmic reticulum (ER). Four enzymes working sequentially carry out each 2C unit addition in the plastid fatty acid synthase and
the ER fatty acid elongase. Acyl reduction and decarbonylation reactions may occur on domains of the ER closely
associated with the plasma membrane and linked to the cell wall (CW).
L. Kunst, A.L. Samuels / Progress in Lipid Research 42 (2003) 51–80
71
Fig. 6. Transient expression of green fluorescent protein (GFP) tagged proteins in epidermis of tobacco (Nicotiana
benthamiana) leaves. (a) HDEL-GFP, ER resident protein retention signal fused to green fluorescent protein gives
control characteristic pattern for ER. (b) Subcellular localization of CER6-GFP, showing the distribution of b-ketoacyl-CoA synthase (condensing enzyme of fatty acid elongation producing very long chain fatty acids for aliphatic wax
components) in the ER (Zhang and Kunst, unpublished data).
72
L. Kunst, A.L. Samuels / Progress in Lipid Research 42 (2003) 51–80
switched on by light in sorghum cork cells. Further support for the ER localization of elongase
enzymes comes from reports in S. cerevisiae. The VLCFA enoyl reductase in this organism has
been shown to reside in the ER membrane, in a novel ER subdomain between the nucleus and the
vacuole [65]. There is also good evidence that the b-ketoacyl reductase required for VLCFA
production is an ER membrane protein [60].
Once elongation is complete, VLCFAs are reduced to fatty alcohols by an acyl-CoA reductase
of the acyl-reduction pathway, which appears to be associated with the ER [70]. The primary
alcohols can then be linked to fatty acids to give rise to wax-esters. Wax synthase, catalyzing this
reaction, is an integral membrane protein [78], but its site of action is not known.
Alternatively, VLCFA precursors are reduced to aldehydes by a membrane-bound acyl-CoA
reductase before they are used by the enzymes of the decarbonylation pathway. The decarbonylation reaction has been localized by subcellular fractionation to a heavy ‘‘cell wall-cuticle
region’’. This is surprising, because the pH optimum of the enzymatic activity showed a sharp
peak near 7, suggesting a cytoplasmic location [79]. It is possible that the decarbonylase exists on
the cytoplasmic face of the ER or plasma membrane (PM) in a domain that is physically associated with the cell wall. Strong physical links between the ER, PM, and cell wall have been
demonstrated [109,110]. This physical interaction can be resolved after plasmolysis of cells. In one
type of association, shrinkage of the protoplast from the wall reveals thin strands of cytoplasm
that remain attached to the cell wall. These strands, first described by Hecht in 1912 [111], are
long thin extensions of the cytoplasm that attach to the wall in discrete spots and include endoplasmic reticulum, ribosomes, and cytoskeletal elements [112,113]. Localization of the decarbonylase using a decarbonylase-GFP/confocal microscopy and immuno-gold/TEM will be
informative, especially in cells that have been plasmolyzed.
Alkanes can be further processed into secondary alcohols and ketones by enzymes that are
entirely uncharacterized either biochemically or molecularly.
3.2.3. Transport from the ER to the plasma membrane and export from the cell
It is currently not known how molecules such as very long chain fatty acids, alcohols, aldehydes
or alkanes are moved from the ER to the plasma membrane. It is even more difficult to envision
how they are exported from their ‘‘comfortable’’ nonpolar environment to the aqueous apoplast.
In the absence of fatty acid binding proteins in plants, it appears that molecules this hydrophobic
would partition into membrane bilayers. Based on our current knowledge, there are at least two
nonexclusive mechanisms which could move wax components from the ER to the PM. It must be
stressed that there is no experimental evidence supporting any of these two potential mechanisms,
and they are presented as hypotheses to be tested, rather than well-established fact.
Hypothesis 1. Wax components could be transferred directly to the PM via sites of close apposition of ER domains with the protoplasmic face of the plasma membrane. In rapidly frozen, freeze
fractured preparations of plant cells, plasma membrane domains in tight association with the
cortical ER are observed [114,115] (Fig. 4, Samuels and Staehelin, unpublished observations). No
membrane fusion occurs in these sites but the ER membrane is in close (10 nm) proximity to the
PM. These sites have been postulated to represent regions where molecular recycling of lipids
could occur following addition of lipids to the plasma membrane during exocytosis [114]. During
conventional chemical fixation for electron microscopy, the osmotic disruptions created by
L. Kunst, A.L. Samuels / Progress in Lipid Research 42 (2003) 51–80
73
inactivation of plasma membrane pumps would disrupt these ER-PM associations, so cryofixation is necessary to preserve them. Likewise, thin sectioning for TEM makes it difficult to visualize these associations as the 70 nm thick sections typically only contain disjointed strands of ER.
If direct transfer of lipid molecules from the ER to the PM exists, it must be carefully controlled
to maintain the unique lipid composition of each compartment. The membrane lipids of the PM
contain more saturated fatty acids and there are more sterols in the PM than the ER [116]. Evidence that lipids can move in the opposite direction, from the PM to the ER without a vesicular
intermediate comes from observations in suspension cultured soy cells treated with fluorescent
membrane lipids. Application of fluorescent phosphatidylcholine (NBD-PC) led to rapid conversion of NBD-PC to NBD-diacylglycerol (NBD-DAG), and translocation directly from the
PM to the ER [117]. During wax biosynthesis, very long chain fatty acyl-CoA molecules could be
converted to alkanes, secondary alcohols, ketones or primary alcohols at ER sites adjacent to the
plasma membrane followed by direct translocation into the plasma membrane.
Hypothesis 2. Wax components could be carried by vesicular traffic from the ER to the Golgi and
finally PM. Following elongation in the ER, acyl reduction or decarbonylation reactions could also
occur in the ER, or anywhere along the secretory pathway, and the wax products could be exported
by exocytosis of Golgi derived vesicles. The ‘‘cell wall/cuticle’’ fraction containing the decarbonylase
could just as easily be a plasma membrane fraction tightly associated with the cell wall, as suggested
above, except in this scenario decarbonylation could be occurring in the plasma membrane [22].
Sphingolipids, which carry very long chain fatty acyl tails, have been shown to follow the classical
ER-Golgi-PM secretory route to the plasma membrane in mammals, S. cerevisiae, and plants
(reviewed by Moreau et al. [118]). The acyl chains found in sphingolipids are typically C24:0 and may
associate with sterols to form ‘‘lipid rafts’’ [119], transient iceberg-like microdomains in the PM. In
yeast, there is evidence that lipid rafts assemble in the ER, and that they are involved in ER-Golgi-PM
traffic [120]. VLCFAs in membrane bilayer can cause strong perturbations in the acyl chain environment within the bilayer [121,122]. Association of the very long chain acyl groups into lipid rafts with
sterols could be a mechanism for preventing disruption of the lipid bilayer structure by VLCFA. If
wax precursors follow the ER-Golgi-PM route, a physical sorting via lipid rafts can be a mechanism to
separate these potentially disruptive molecules from the housekeeping domains of the ER.
As stated above, these are hypothetical routes of very long chain acyl groups leaving the
endomembrane system. There is little evidence to support either a direct ER-PM or the Golgimediated route. Jenks et al. [108] looked for ultrastructural changes in sorghum cork cells when
wax production was stimulated by light. They showed that, while there was a proliferation of the
ER, the Golgi and the associated vesicles did not change. It should be noted, however, that the
Golgi associated vesicles are difficult to quantify using the conventional chemical fixation electron
microscopy techniques employed in this study. In the analysis of the Golgi and the ER of waxless
Arabidopsis mutants cer1–cer10, there were no mutants with accumulations of Golgi vesicles or
altered Golgi morphology [123].
Once the wax components are in the plasma membrane, they must be removed from the bilayer
and moved into the aqueous apoplast. Moving VLCFacyl derivatives from the hydrophobic core of the
bilayer to the aqueous apoplast would require an input of energy. Members of the ‘‘ABC transporter’’
protein family pump a variety of lipophilic molecules out of yeast and mammalian cells [124,125]. The
74
L. Kunst, A.L. Samuels / Progress in Lipid Research 42 (2003) 51–80
characteristics of ABC transporters that make them potential candidates for wax exporters are their
homology to known lipid transporters, location of one subset of ABC transporters in the plasma
membrane, and their ability to hydrolyze ATP to provide energy for energetically unfavourable
processes. In the Arabidopsis genome, 129 sequences have been identified as putative ABC transporters [126,127]. Many of these have strong sequence similarity to drug resistance proteins and
pigment transporters of mammalian and yeast systems, and they can be classed into six ’families’ of
related sequences (summarized at www.arabidopsisABC.net). A role for ABC transporters in the
export of secondary metabolites has been postulated, based on ATP dependent export of a diterpene
[128]. If ABC transporters are involved in diterpene export, it is logical to extrapolate a role for ABC
transporters in export of triterpenoids in cuticles rich in these compounds, such as Prunus laurocerasus [18]. With T-DNA insertion lines of Arabidopsis available, it should be possible to examine the
lines with disruptions to their ABC transporters for waxless phenotypes.
3.3. Transport of wax precursors across the cell wall to the cuticle
3.3.1. Lipid transfer proteins
Hydrophobic wax components must move through the hydrophilic cell wall matrix to reach the
cuticle. Lipid transfer proteins (LTPs) have been suggested to mediate this movement during wax
biosynthesis [66,129,130]. They may be specific for one lipid class, such as phosphoinositol, or,
more typically for plants, have broad substrate binding specificity (nonspecific lipid transfer proteins; nsLTP). The nsLTPs are abundant in plant wax. For example, the broccoli LTP WAX9
(73% identical to Arabidopsis LTP1) represented more than 90% of the wax-associated protein
on broccoli (B. oleracea) leaves [131]. LTPs were originally defined by their ability to transfer
phospholipids between membranes in vitro, leading to a hypothesized role in lipid transfer
within the cell (reviewed by Kader [132]). However, this hypothesis is not supported in plants
where nsLTP have hydrophobic N-terminal signal sequences, suggesting that LTPs are targeted
to the secretory pathway [133] and ultimately to extracellular locations in the cell wall and
cuticle [131,134].
LTPs are small (9 kd), basic proteins with eight conserved cysteines which are joined by disulfide bonds to form a hydrophobic tunnel [135–138]. High resolution x-ray crystallography of
maize nsLTP has revealed that the hydrophobic tunnel accomodates a variety of ligands from
C10-C18 by allowing some flexibility in the ligand binding cavity [139]. This would be consistent
with nsLTP carrying C16 and C18 cutin precursors to the plant surface. However, the modeling
to date suggests that the very long chain fatty acid wax precursors would not fit in the hydrophobic cavity of the LTP [140]. Furthermore, in direct competition experiments, maize LTP
showed substrate preference for C18 over C20, C22 or hydroxylated C18 acyl chains [141].
Although there has been no direct demonstration of a role for the LTPs in transferring wax
components or cutin across the apoplast, there is circumstantial evidence for their involvement in
cuticular lipid synthesis. Data from RNA blotting experiments, in situ hybridizations, promoterGUS fusions and EST database projects demonstrate that LTPs are expressed in an epidermisspecific manner, especially in organs active in cuticle production such as developing flowers,
leaves and stems [131,132,134,142,143]. In addition, positive correlations have been reported
between conditions that stimulate wax synthesis, such as drought [144], light [145], or heavy metal
treatment [146], and lipid transfer protein expression.
L. Kunst, A.L. Samuels / Progress in Lipid Research 42 (2003) 51–80
75
If LTPs do, indeed, carry cuticular lipid components to the plant surface, they could be loaded
with their ‘‘cargo’’ in the lumen of the ER and follow the secretory pathway out into the apoplast. Immuno-gold localization and cell fractionation studies, however, suggest that the lipid
transfer proteins are primarily present in the apoplast, and not the endomembrane system
[131,134]. It is possible that these wax associated proteins represent ‘‘spent’’ carriers, although if
each LTP is capable of carrying one or two fatty acids on a one-way trip, then massive amounts
of protein would be required. Alternatively, lipid transfer proteins could interact with the plasma
membrane lipid bilayer and extract lipids directly into their hydrophobic cavities. An intriguing
preliminary study has shown that LTP from wheat (T. aestivum) binds to plasma membrane
vesicles with saturation kinetics, suggesting the presence of a plasma membrane receptor for the
LTP [147].
3.3.2. The physical path of wax components through the epidermal cell wall
The unique cell wall structure of the typical epidermal cell may give clues as to the path that
wax must take through the cell wall (reviewed by Martin and Juniper [13]). Between the symplasm and the epicuticular surface, there are distinct subdomains of cell wall around the waxsecreting cell (Fig. 1). The innermost zone is a typical primary cell wall of about 0.25–0.5 mm,
which encases the epidermal cell. The current models depict the primary cell wall as a hydrophilic,
reinforced gel: 2–3 layers of cellulose microfibril layers, complexed with hemicelluloses, enclosed
in a matrix of pectins and proteins [148–150]. On the periclinal surface facing the plant surface is
a cap of secondary cell wall layer over the cell, which may play a mechanical role in preventing
deformation of the epidermis. Overlying this, and interacting directly with the cuticular lipids, is a
pectin rich layer with cellulose microfibrils branching through it, which is in continuity with the
pectin-rich middle lamella of the anticlinal walls between the adjacent cells. The porosity of the
wall seems to be determined by the properties of the pectin matrix [151], a mixture of homogalacturonans and rhamnogalacturonans variably methylesterifed or crosslinked by calcium
through their carboxyl groups.
So how the hydrophobic wax components make their way through the layers of the epidermal
cell wall to the plant surface? Epidermal cells do not show strong structural polarity in their
organelle distribution, i.e. accumulation of Golgi and secretory vesicles on the periclinal surface
of the cell. If Golgi vesicles are carrying wax components, there must be targeting information
which is specific to the apical surface of the cell to facilitate vesicle fusion in this area. Following
release, the wax components could move through the primary cell wall and into the middle
lamella which is continuous with the apical pectic layer. It is conceivable that they ‘travel’ within
the hydrophobic cavity of LTPs, which are highly abundant in the apoplast and below the size
exclusion limit of the cell wall. Once at the cuticle/pectin interface, hydrophobic wax components
could partition into the nonpolar phase of the cuticular layer. When the concentration of a certain wax component reaches a critical threshold, epicuticular crystalloids of that component
could form on the aerial surface from the amorphous wax mixture embedded in the crosslinked
cutin [152–154].
An alternative possibility is that the precursors move through the cell wall in direct contact with
the cell wall matrix. Cell wall proteins or polysaccharides may create relatively hydrophobic
subdomains within the cell wall where passage of free hydrophobic components could be facilitated and wax components could self-associate into a narrow ’river’ of hydrophobic compounds
76
L. Kunst, A.L. Samuels / Progress in Lipid Research 42 (2003) 51–80
moving to the cell surface. Although the cell wall is predominantly hydrophilic, hydrophobic
domains could be formed by cell wall proteins such as glycine rich proteins (GRP), whose
secondary structure has been predicted to form b-pleated sheets with a large hydrophobic surface [155,156]. GRPs have been localized in vascular tissues but also in the epidermis of maize
[157] and Arabidopsis [158]. Potential polysaccharide components that could contribute to
hydrophobic domain formation are highly methyl-esterified pectins. There are cases where polysaccharides bind hydrophobic compounds. The cyclic oligosaccharide from a bacterium, cyclodextrin, is used as a tool in animal cell biology to selectively remove and bind cholesterol from the
plasma membrane [159]. The a-(1-4) glycosidic ring of cyclodextrin has the hydroxyl groups
projecting outward, creating a hydrophobic pocket for lipid binding [160]. A homologous cell
wall polysaccharide held under tension to project the hydrophobic domain along a passage could
contribute to an area of the cell wall where wax constituents could pass. The movement through
the aqueous apoplast has been difficult to examine experimentally. The process cannot be
observed with conventional electron microscopy because sample preparation procedures such as
dehydration and resin embedding extract lipidic components. As well, saturated long chain fatty
acid derivatives such as alkanes lack chemical groups that would react with heavy metal stains.
Historically, debate has centred on whether pores exist in the cuticle that could be responsible
for extrusion of wax crystals [14,161]. More recent descriptions of the phase behaviour and crystallinity characteristics of the cuticle studied by Fourier transform infrared spectroscopy [162]
confirm earlier nuclear magnetic resonance results [163] showing that plant cuticular wax components organize themselves into crystalline and amorphous zones. Amorphous regions might
allow freer movement of epicuticular wax through the cuticle, resulting in regions which act as
functional ‘‘pores’’ without a physical opening occurring. Once in the hydrophobic realm of the
cuticle, wax behaves as a self-organizing system [164]. Based on movement of wax through isolated cuticles, a mechanism for movement of epicuticular wax components in the cuticular water
vapour stream has been proposed [165].
4. Conclusions
The attention devoted to the plant cuticular wax is a reflection of its importance to plant biology. The wax on the outside of a plant is the fossil record of the metabolic activity of that epidermis. Its examination has contributed valuable information about the chemical diversity of wax
composition within a single species and amongst taxa. The pioneering contributions of Kolattukudy, von Wettstein-Knowles, and others have shown us the basic wax biochemical pathways.
The isolation of mutants in Arabidopsis and other plant species, combined with biochemical
analyses of mutant phenotypes, resulted in identification gene products involved in wax production, and cloning of a number of wax biosynthetic genes. However, not all genes involved in the
biochemistry, regulation and transport of wax to the plant surface have been identified by
genetic approaches. The challenge now is to take the genomic tools available to find additional
genes whose functions are essential, redundant, or more subtle, and investigate their specific
contributions to cuticular wax production during the life cycle of the plant. The path to greater
understanding of how wax is synthesized and secreted will come from delving into the active
plant cell.
L. Kunst, A.L. Samuels / Progress in Lipid Research 42 (2003) 51–80
77
Acknowledgements
Thanks to O. Rowland, T. Hooker, T. Western, H. Zhang, G. Chowrira, R. Jetter and A.
Staehelin for helpful discussions. This work was supported by the Natural Sciences and
Engineering Research Council of Canada grants to L.K. and A.L.S.
References
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]
[21]
[22]
[23]
[24]
[25]
[26]
[27]
[28]
[29]
[30]
[31]
[32]
[33]
[34]
von Wettstein-Knowles P. In: Hamilton RJ, editor. Waxes: chemistry, molecular biology and functions. Dundee: Oily Press; 1995. p. 91–129.
Bianchi G, Murelli C, Ottaviano E. Phytochemistry 1990;29:739–44.
Preuss D, Lemieux B, Yen G, Davis RW. Genes Dev 1993;7:974–85.
Liu H, Przybylski R, Dawson K, Eskin NAM, Biliaderis CG. J Am Oil Chem Soc 1996;73:493–8.
Isbell TA, Carlson KD, Abbott TP, Phillips BS, Erhan SM, Kleiman R. Ind Crop Prod 1996;5:239–43.
Reiter B, Lechner M, Lorbeer E, Aichholz R. J High Res Chromatogr 1990;22:514–20.
Reicosky DA, Hanover JW. Plant Physiol 1978;62:101–4.
Barnes JD, Percy KE, Paul ND, Jones P, McLaughlin CK, Mullineaux PM, et al. J Exp Bot 1996;47:99–109.
Kerstiens G. Trends Plant Sci 1996;1:125–9.
Barthlott W, Neinhaus C. Planta 1997;202:1–8.
Jenks MA, Joly RJ, Peters PJ, Rich PJ, Axtell JD, Ashworth EN. Plant Physiol 1994;105:1239–45.
Eigenbrode SD, Espelie KE. Annu Rev Entomol 1995;40:171–94.
Martin JT, Juniper BE. The cuticles of plants. London: Edward Arnold; 1970.
Baker EA. In: Cutler DJ, Alvin KL, Price CE, editors. The plant cuticle. London: Academic Press; 1982. p.
139–65.
Post-Beittenmiller D. Plant Physiol Biochem 1998;36:157–66.
Koornneef M, Hanhart CJ, Thiel F. J Hered 1989;80:118–22.
Jetter R, Schäffer S, Riederer M. Plant Cell Environ 2000;23:619–28.
Jetter R, Schäffer S. Plant Physiol 2001;126:1725–37.
Walton T. In: Harwood JL, Bowyer JR, editors. Methods in plant biochemistry: lipids, membranes and aspects
of photobiology. San Diego: Academic Press, 1990; p. 4:105–58.
von Wettstein-Knowles P. In: Moore TS, editor. Lipid metabolism in plants. Boca Raton: CRC Press, 1993; p.
127–66.
Kolattukudy PE. Plant Physiol 1968;43:375–83.
von Wettstein-Knowles P. In: Appelqvist L-A, Liljenberg L, editors. Advances in the biochemistry and physiology of plant lipids. Amsterdam: Elsevier/North Holland Biomedical Press, 1979; p. 1–26.
Kolattukudy PE. In: Kerstiens G, editor. Plant cuticles. Oxford: BIOS Scientific Publishers; 1996. p. 83–108.
von Wettstein-Knowles P. In: von Wettstein D, Chua NH, editors. Plant molecular biology. New York: Plenum, 1987. p. 305–14.
McNevin JP, Woodward W, Hannoufa A, Feldmann KA, Lemieux B. Genome 1993;36:610–8.
Lemieux B, Koornneef M, Feldmann KA. In: Meyerowitz EM, Somerville CR, editors. Arabidopsis. New
York: Cold Spring Harbor Press; 1994. p. 1031–47.
Aarts MGM, Keijzer CJ, Stiekema WJ, Pereira A. Plant Cell 1995;7:2115–27.
Hannoufa A, McNevin J, Lemieux B. Phytochemistry 1993;33:851–5.
Hansen JD, Pyee J, Xia Y, Wen T-J, Robertoson DS, Kolattukudy PE, et al. Plant Physiol 1997;113:1091–100.
Xia Y, Nikolau BJ, Schnable PS. Plant Cell 1996;8:1291–304.
Negruk V, Yang P, Subramanian M, McNevin JP, Lemieux B. Plant J 1996;9:137–45.
St-Pierre B, Laflamme P, Alarco A-M, De Luca V. Plant J 1998;14:703–13.
Jenks MA, Tuttle HA, Eigenbrode SD, Feldman KA. Plant Physiol 1995;108:369–77.
Hannoufa A, Negruk V, Eisner G, Lemieux B. Plant J 1996;10:459–67.
78
[35]
[36]
[37]
[38]
[39]
[40]
[41]
[42]
[43]
[44]
[45]
[46]
[47]
[48]
[49]
[50]
[51]
[52]
[53]
[54]
[55]
[56]
[57]
[58]
[59]
[60]
[61]
[62]
[63]
[64]
[65]
[66]
[67]
[68]
[69]
[70]
[71]
[72]
[73]
[74]
[75]
[76]
[77]
[78]
[79]
L. Kunst, A.L. Samuels / Progress in Lipid Research 42 (2003) 51–80
Lemieux B. Trends Plant Sci 1996;1:312–8.
Kwon YT, Reiss Y, Fried VA, Hershko A, Yoon JK, Gonda DK, et al. Proc Natl Acad Sci USA 1998;
95:7898–903.
Bonifacino JS, Weissman AM. Annu Rev Cell Dev Biol 1998;14:19–57.
Millar AA, Clemens S, Zachgo S, Giblin EM, Taylor DC, Kunst L. Plant Cell 1999;11:825–38.
Fiebig A, Mayfield JA, Miley NL, Chau S, Fischer RL, Preuss D. Plant Cell 2000;12:2001–8.
Ohlrogge JB, Jaworski JG, Post-Beittenmiller D. In: Moore TS Jr., editor. Lipid metabolism in plants. Boca
Raton: CRC Press, 1993. p. 3–32.
Browse J, Ohlrogge J. Plant Cell 1995;7:957–70.
Clough RC, Matthis AL, Barnum SR, Jaworski JG. J Biol Chem 1992;267:20992–8.
Shimakata T, Stumpf PK. Proc Natl Acad Sci USA 1982;79:5808–12.
Stumpf PK. In: Numa S, editor. Fatty acid metabolism and its regulation. Amsterdam: Elsevier Science, 1984.
p. 155–79.
von Wettstein-Knowles P. Physiol Veg 1982;20:797–809.
Fehling E, Mukherjee KD. Biochim Biophys Acta 1991;1082:239–46.
Millar AA, Kunst L. Plant, J 1997;12:121–31.
Kunst L, Taylor DC, Underhill EW. Plant Physiol Biochem 1992;30:425–34.
Lessire R, Bessoule J-J, Cassagne C. FEBS Lett 1985;187:314–20.
Bessoule J-J, Lessire R, Cassagne C. Arch Biochem Biophys 1989;268:475–84.
Murphy DJ, Mukherjee KD. Z Naturforsch 1989;44c:629–34.
Agrawal VP, Stumpf PK. Lipids 1985;20:361–6.
James Jr. DW, Dooner HK. Theor Appl Genet 1990;80:241–5.
Lemieux B, Miquel M, Somerville C, Browse J. Theor Appl Genet 1990;80:234–40.
James Jr. DW, Lim E, Keller J, Plooy I, Ralston E, Dooner HK. Plant Cell 1995;7:309–19.
Todd J, Post-Beittenmiller D, Jaworski JG. Plant J 1999;17:119–30.
Yephremov A, Wisman E, Huijser P, Huijser C, Wellesen K, Saedler H. Plant Cell 1999;11:2187–201.
Pruitt RE, Vielle-Calzada JP, Ploense SE, Grossniklaus U, Lolle SJ. Proc Natl Acad Sci USA 2000;97:1311–6.
Hooker TS, Millar AA, Kunst L. Plant Physiol 2002;129:1568–80.
Beaudoin F, Gable K, Sayanova O, Dunn T, Napier JA. J Biol Chem 2002;277:11481–8.
Xu X, Dietrich CR, Delledonne M, Xia Y, Wen T-J, Robertson DS, et al. Plant Physiol 1997;115:501–10.
Emerson RA. In: Cornell University Agricultural Experimental Station Memoir, no. 180. Ithaca (NY): Cornell
University; 1935. p. 180.
Bianchi G, Avato P, Salamini F. Heredity 1979;42:391–5.
Bianchi A, Bianchi G, Avato P, Salamini F. Maydica 1985;30:179–98.
Kohlwein SD, Eder S, Oh CS, Martin CE, Gable K, Bacikova D. Mol Cell Biol 2001;21:109–25.
Post-Beittenmiller D. Ann Rev Plant Physiol Plant Mol Biol 1996;47:405–30.
Kolattukudy PE. Biochemistry 1970;9:1095–102.
Kolattukudy PE. Arch Biochem Biophys 1971;142:701–9.
Vioque J, Kolattukudy PE. Arch Biochem Biophys 1997;340:64–72.
Metz JG, Pollard MR, Anderson L, Hayes TR, Lassner MW. Plant Physiol 2000;122:635–44.
Aarts MGM, Dirkse WG, Stiekema WJ, Pereira A. Nature 1993;363:715–7.
Aarts MGM, Hodge R, Kalantidis K, Florack D, Wilson ZA, Mulligan BJ, et al. Plant J 1997;12:615–23.
Wu SSH, Platt KA, Ratnayake C, Wang T-W, Ting JTL, Huang AHC. Proc Natl Acad Sci USA 1997;94:12711–
6.
Hernandez-Pinzon I, Ross JHE, Barnes KA, Damant AP, Murphy DJ. Planta 1999;208:588–98.
Kolattukudy PE. Biochemistry 1967;6:2705–17.
Khan AA, Kolattukudy PE. Biochemistry 1973;12:1939–48.
Wu W-Y, Moreau RA, Stumpf PK. Lipids 1981;6:897–902.
Lardizabal KD, Metz JG, Sakamoto T, Hutton WC, Pollard MR, Lassner MW. Plant Physiol 2000;122:645–
55.
Cheesbrough TM, Kolattukudy PE. Proc Natl Acad Sci USA 1984;81:6613–7.
L. Kunst, A.L. Samuels / Progress in Lipid Research 42 (2003) 51–80
[80]
[81]
[82]
[83]
[84]
[85]
[86]
[87]
[88]
[89]
[90]
[91]
[92]
[93]
[94]
[95]
[96]
[97]
[98]
[99]
[100]
[101]
[102]
[103]
[104]
[105]
[106]
[107]
[108]
[109]
[110]
[111]
[112]
[113]
[114]
[115]
[116]
[117]
[118]
[119]
[120]
[121]
[122]
[123]
[124]
[125]
[126]
79
Dennis M, Kolattukudy PE. Proc Natl Acad Sci USA 1992;89:5306–10.
Schneider-Belhaddad F, Kolattukudy PE. Arch Biochem Biophys 2000;377:341–9.
Kolattukudy PE, Liu T-YJ. Biochem Biophys Res Commun 1970;41:1369–74.
Kolattukudy PE, Buckner JS, Liu T-YJ. Arch Biochem Biophys 1973;156:613–20.
Placing S, Kannangara CG, Mikkelsen JD, Simpson D, von Wettstein-Knowles P. Barley Genet Newslett 1979;
9:75–7.
Hamilton JA. J Lipid Res 1998;39:467–81.
Zhang F, Kamp F, Hamilton JA. Biochemistry 1996;35:16055–60.
Smith R, Tanford C. Proc Natl Acad Sci USA 1973;70:289–93.
Glatz JFC, Van der Vusse GJ. Prog Lipid Res 1996;35:243–82.
Bernlohr DA, Coe NR, LiCata VJ. Cell Devel Biol 1999;10:43–9.
Storch J, Thumser AEA. Biochem Biophys Acta 2000;1486:28–44.
Stremmel W, Pohl J, Ring A, Herrmann T. Lipids 2001;36:981–9.
Rasmussen JT, Borchers T, Knudsen J. Biochem J 1990;265:849–55.
Rasmussen JT, Rosendal J, Knudsen J. Biochem J 1993;292:907–13.
Rasmussen JT, Faergeman NJ, Kristiansen K, Knudsen J. Biochem J 1994;299:165–70.
Hills MJ, Dann R, Lydiate D, Sharpe A. Plant Mol Biol 1994;25:917–20.
Engeseth NJ, Pacovsky RS, Newman T, Ohlrogge JB. Arch Biochem Biophys 1996;331:55–62.
Chye M-L, Huang B-Q, Sze YZ. Plant J 1999;18:205–14.
Ohlrogge JB, Kuhn DN, Stumpf PK. Proc Natl Acad Sci USA 1979;76:1194–5.
Liu D, Post-Beittenmiller D. J Biol Chem 1995;270:16962–9.
Jones P, Davies HM, Voelker TA. Plant Cell 1995;7:359–71.
Voelker T. In: Setlow, JK, editor. Genetic engineering, 18. New York: Plenum Press; 1996 p. 111–33.
Dörmann P, Voelker TA, Ohlrogge JB. Arch Biochem Biophys 1995;316:612–8.
Dörmann P, Voelker TA, Ohlrogge JB. Plant Physiol 2000;123:637–43.
Joyard J, Stumpf PK. Plant Physiol 1981;67:250–6.
Douce R, Joyard J. Adv Bot Res 1979;7:1–116.
Kaneko Y, Keegstra K. Protoplasma 1996;195:59–67.
Gaigg B, Neergaard TBF, Schneiter R, Krogh-Hanson J, Færgeman JJ, Jensen NA, et al. Mol Biol Cell 2001;
12:1147–60.
Jenks MA, Rich PJ, Ashworth EJ. Int J Plant Sci 1994;155:506–18.
Wyatt SE, Carpita NC. Trends Cell Biol 1993;3:413–7.
Kohorn B. Plant Physiol 2000;124:31–8.
Hecht K. Beitr Biol Pflanzen 1912;11:133–45.
Oparka KJ, Prior DAM, Crawford JW. In: Smallwood M, Knox JP, Bowles DJ, editors. Membranes: specialized functions in plants. Oxford: BIOS Scientific; 1996. p. 39–56.
Lang-Pauluzzi I, Gunning BES. Protoplasma 2000;212:174–85.
Staehelin LA, Chapman R. Planta 1987;171:43–57.
Craig S, Staehelin LA. Eur J Cell Biol 1988;46:80–93.
Larsson C, Møller IM, Widell S. In: Larsson C, Møller IM, editors. The plant plasma membrane. Berlin, Heidelberg: Springer-Verlag; 1990.p. 1–13.
Grabsky S, de Feijter W, Schindler M. Plant Cell 1993;5:25–38.
Moreau P, Bessoule S, Mongrand E, Testet E, Vincent P, Cassagne C. Prog Lipid Res 1998;37:371–91.
Brown DA, London E. J Biol Chem 1998;275:17221–4.
Bagnat M, Kernen S, Shevchenko A, Shevchenko A, Simons K. Proc Natl Acad Sci USA 2000;97:3254 -3259.
Ho JK, Moser H, Kishimoto Y, Hamilton JA. J Clin Invest 1995;96:1455–63.
Millar AA, Wrischer M, Kunst L. Plant Cell 1998;10:1889–902.
Samuels AL, Kunst L. [unpublished observations].
Decottignies A, Gotteau A. Nature Genet 1997;15:137–45.
Klein I, Salkadi B, Vàradi A. Biochem Biophys Acta 1999;1461:237–62.
Rea PA, Li Z-S, Lu Y-P, Drozdowicz YM, Martinoia E. Ann Rev Plant Physiol Plant Mol Biol 1998;49:727–60.
80
[127]
[128]
[129]
[130]
[131]
[132]
[133]
[134]
[135]
[136]
[137]
[138]
[139]
[140]
[141]
[142]
[143]
[144]
[145]
[146]
[147]
[148]
[149]
[150]
[151]
[152]
[153]
[154]
[155]
[156]
[157]
[158]
[159]
[160]
[161]
[162]
[163]
[164]
[165]
[166]
[167]
[168]
[169]
L. Kunst, A.L. Samuels / Progress in Lipid Research 42 (2003) 51–80
Sánchez-Fernández R, Emyr-Davies TG, Coleman JOD, Rea PA. J Biol Chem 2001;276:30231–44.
Jasinski M, Stukkens Y, Degand H, Purnelle B, Marchand-Brynaert J, Boutry M. Plant Cell 2001;13:1095–107.
Sterk P, Booij H, Schellekens GA, Van Kammen A, De Vries SC. Plant Cell 1991:907–21.
Jenks MA, Rick PJ, Ashworth EN. Int J Plant Sci 2000;155:506–18.
Pyee J, Yu H, Kolattukudy PE. Arch Biochem Biophys 1994;311:460–8.
Kader J-C. Ann Rev Plant Phys Plant Mol Biol 1996;47:627–54.
Bernhard WR, Thoma S, Botella J, Somerville C. Plant Physiol 1991;95:164–70.
Thoma S, Kaneko Y, Somerville C. Plant J 1993;3:427–36.
Gincel E, Simorre JP, Caille A, Marion D, Ptak, Vovelle F. Eur J Biochem 226: 413–22.
Shin DH, Lee JY, Qwang KY, Kim KK, Suh SW. Structure 1995;3:189–99.
Gomar J, Sodano P, Sy D, Shin DH, Lee JY, Suh SW, et al. Proteins 1998;31:160–71.
Tassin-Moindrot S, Caille A, Douliez J-P, Marion D, Vovelle F. Eur J Biochem 2000;267:1117–24.
Han GW, Lee JY, Song HK, Chang C, Min K, Moon J, et al. J Mol Biol 2001;308:263–78.
Keresztessy Z, Hughes M. Plant J 1998;14:523–33.
Zachowski A, Guerbette F, Grosbois M, Jolliot-Croquin A, Kader J-C. Eur J Biochem 1998;257:443–8.
Arondel V, Vergnolle C, Cantrel C, Kader J-C. Plant Sci 2000;157:1–12.
Clark AM, Bohnert HJ. Plant Cell Physiol 1999;40:69–76.
Treviño MB, O’Connell MA. Plant Physiol 1998;116:1461–8.
Sohal AK, Pallas JA, Jenkins GI. Plant Mol Biol 1999;41:75–87.
Hollenbach B, Schreiber L, Hartung W, Dietz K-J. Planta 1997;203:9–19.
Buhot N, Douliez J-P, Jacquemard A, Marion D, Tran V, Maume BF, Milat M-L, et al. FEBS Lett 2001;
509:27–30.
McCann M, Roberts K. In: Lloyd CW, editor. The cytoskeleton in plant growth and form. London: Academic
Press; 1991. p. 109–29.
Carpita NA, Campbell M, Tierney M. Plant Mol Biol 2001;47:1–340.
Roberts K. Plant Physiol 2001;125:127–30.
Baron-Epel O, Gharyl PK, Schindler M. Planta 1988;175:389–95.
Jeffree CE. New Phytol 1975;75:539–49.
Jeffree CE In: Dickinson CH, Preece TF, editors. Microbiology of aerial plant surface. London: Academic
Press; 1976. p. 119–58.
Jetter R, Riederer M. Planta 1994;195:257–70.
Condit CM, Meager RB. Nature 1986;323:178–81.
Sachetto-Martins G, Fernandes LD, Felix DB, deOliveira DE. Biochem Biophys Acta 2000;1492:1–14.
Gòmez J, Sánchez-Martinez D, Stiefel J, Rigau P, Puigdomènech P, Pagès M. Nature 1988;334:262–4.
Sachetto-Martins G, Fernandes LD, Felix DB, deOliveira DE. Int J Plant Sci 1995;156:460–70.
Neufeld EB, Cooney AM, Pitha J, Dawidowicz EA, Dwyer NK, Pentchev PG, et al. J Biol Chem 1996;
271:21604–13.
Bender ML, Komiyama M. Cyclodextrin chemistry. Berlin: Springer-Verlag; 1978.
Jeffree CE In: Kerstiens G, editor. Plant cuticles. Oxford: BIOS Scientific; 1996. p. 33–82.
Merk S, Blume A, Reiderer M. Planta 1998;204:44–53.
Reynhart EC, Riederer M. J Phys D: Appl Phys 24:478–86.
Kirsch T, Kaffarnik F, Riederer M, Schreiber L. J Exp Bot 1997;48:1035–45.
Neinhuis C, Koch K, Barthlott W. Planta 2001;213:427–34.
Rashotte AM, Jenks MA, Feldmann KA. Phytochemistry 2001;57:115–23.
Barthlott W, Neinhuis C, Cutler D, Ditsch F, Meusel I, Theisen I, Wilhelmi H. Bot J Lin Soc 1998;126:237–60.
Jenks MA, Ashworth EN. Hor Rev 1999;23:1–68.
Reiderer M, Schreiber L. In: Hamilton RJ, editor. Waxes: Chemistry, molecular biology and functions. Dundee: Oily Press, 1995. 131–56.