Phytosterolomics

The study of metabolic networks and feedback response to developmental and environmental conditions involves the profiling of natural products. In this context, phytosterolomic research can be defined as the complete identification and quantification of all sterols present in a specific biological sample. Phytosterol fingerprinting is another aspect of phytosterolomics where the unusual sterol composition of a pathogen can serve as a signature lipid for disease such as for immunocompromised patients harboring pneumocystis (Kaneshiro et al., 2002; Zhou et al., 2002). Phytosterol profiling involves techniques such as thin layer chromatography (TLC) and high-performance liquid chromatography (LC) mass spectroscopy (LC/MS) and gas chromatography mass spectroscopy (GC/MS) where individual components are separated and tentatively identified and quantified by one set of techniques with final characterization of structure determined by ¹H or ¹³C NMR spectroscopy (Kalinowska et al., 1990; Nes et al., 1998a, 1999, 2003). HPLC equipped with a diode array detector to monitor UV is useful in the identification of the type and amount of conjugation in the sterol molecule (Nes et al., 1985; Norton and Nes, 1991; Xu et al., 1988). Separation factors in GLC and HPLC based on the retention of cholesterol relative to the retention time of sterol standards have been reported to enable the prediction of the identity of an unknown sterol (Xu et al., 1988). High-throughput screening of phytosterol mixtures by conventional chromatography and spectroscopy has not yet progressed to the point where a single method can establish the identities of a complicated set of structurally similar phytosterols. Although the sterol patterns of many plant extracts deduced by GLC are rather simple, usually consisting of three or four major sterols, the sterol composition of specific plant parts or the type of sterol in primitive versus advanced plants can be strikingly different (Nes, 1990).

The sterol side-chain can be modified by the addition of one or two supernumerary carbon atoms at C-24 with either α- or β-chirality. Establishing the stereochemistry of the 24-alkyl group of individual sterols is a formidable task and can be achieved by identifying specific side-chain carbons by high-field NMR (Fig. 9.7, Panels A–C). The assignments of the signals from the ¹H or ¹³C NMR spectra of ¹H- or ¹³C-labeled sterols are diagnostic for the origin of the methyl group at C-24 and the biosynthetic stereospecificity of phytosterol sidechains (Guo et al., 1996, 1995; Popják et al., 1977). The relative proportion of α- to β-isomer of a 24α/β-methyl cholesterol mixture can also be quantified by reversedphase HPLC (Guo et al., 1995) (Fig. 9.7, Panel D) (Parker and Nes, 1992). The ratio of these diastereoisomers can change as plants evolve from less to more advanced (Nes et al., 1977).

FIGURE 9.8 NOE networks that indicate whether cycloartenol can orient into a flat or bent shape. Adapted from Nes et al. (1998b).

A more exacting analysis of sterol structure and stereochemistry involves the use of several spectroscopic techniques, including high-field ¹H and/or ¹³C-NMR, X-ray crystallography, as well as molecular modeling (Nes et al., 1998a). An example where the use of several techniques has played a major role in our understanding of conformational analysis of plant sterols is the characterization of cycloartenol (with a 9β,19-cyclopropane ring) and lanosterol (with a Δ⁸⁽⁹⁾- bond). There is a hypothesis that the structural isomers are shaped differently, bent versus flat (Bloch, 1983; Goodwin, 1981; Rahier et al., 1984). The putative bent 9β,19-cyclosterol was considered to be an unique structural trait of intermediate plant sterols, and bent compounds were acceptable only to either specific plant enzymes or membrane systems. The rationale for 9β,19-cyclosterols to be bent is based on the syn–cis configuration at theA/B and B/C ring junctions, resulting in an unfavorable interaction between the 9,10-bridgehead

FIGURE 9.9 Partial X-ray crystallographic structures of cycloartenol, lanosterol, and sitosterol. Adapted from Nes et al. (1991b).

and the 8β-hydrogen atomat C-8. Furthermore, on the basis of manipulation of ball and stick models, it was proposed that ring B in 9β,19-cyclosterols becomes a boat and theA/B/Crings orient in the chair-boat-chair conformation. However, we discovered for the first time excellent agreement between a set of crystallographically observed cycloartenoltype structures and their solution conformations deduced from 2D-NMR spectroscopic analysis, analysis of the NOE networks (Fig. 9.8) and MM/MD calculations to be pseudoplanar (A/B/C rings are chair/half-chair/twist-chair conformer), and the sterol side-chain to orient to the ‘‘right’’ (Nes et al., 1984, 1991a). These results with sitosterol, lanosterol, and related compounds reveal that the compounds possess similar three-dimensional shapes (Nes et al., 1991b) and differ in the tilt of the C-3 group and C-17(20) side-chain (Fig. 9.9), structural features that can be responsible for the different activities of sterols (Nes et al., 1991b, 1993).

Using the modern methods of sterol analysis, it has been possible to show that phytosterols appear in all cells at all stages of development, but the type and amount of sterol is greatly influenced by ontogeny and speciation, and consequently seems to be a carefully regulated process (Nes, 1990). In spite of findings that show variability of the sterol side C-24 alkyl structure occurs widely, suggesting some sort of association between sterol structure and plant biology, relatively little is known about individual sterols or sterol sets and their role in plant physiology. As many as 60 sterols have been reported in a vascular plant, Zea mays (Guo et al., 1995), which is about one-third to one-half the number of phytosterols found in living systems; 39 sterols have also been identified in an ascomycetous fungus, Gibberella fujikuroi (Nes et al., 1989a).

The ratio of 4,4-dimethyl and 4-desmethyl sterol intermediate to 4-desmethyl sterol end product, easily determined by TLC, can be as much as 1:3 in pollen, 1:3 in the seed, and 1:9 in vegetative parts (Heupel et al., 1986; Marshall et al., 2001; Nes and Schmidt, 1988; Nes et al., 1991c). These differences in intermediate to end product ratios suggest that the phytosterol pathway is undergoing significant changes during development. Consistent with this observation, as plants mature from seed to flower the amounts of total sterol in individual parts and various cell types differ dramatically. For example, in sunflower the sterol levels are reported to be (Nes, 1990, 1991b,c) in: a seed, 83 µg; the 4-day dark grown sprout, 15 µg; the primary leaf, 15 µg; mature leaf, 100 mg; immature green flower bud, 35 µg; disk flower, 11 µg; cultured cells, 3,000 fg/cell; and the ray mesophyll cells released by macerase digestion, 500 fg/cell. Pollen grains were found to be the richest source of sterol, 585,000 fg/grain.

Sterols accumulate in vegetative parts mostly as 24-ethylsterols (e.g., sitosterol and stigmasterol [22-dehydrositosterol]) whereas pollen can accumulate significant amounts of 24-methylene sterols and 24-desalkylsterols (4,4-dinorcycloartenol; 24-dehydropollinastanol) (Guo et al., 1995; Nes and Schmidt, 1988). However, the sterols of some pollen and seeds can be mostly sitosterol and related phytosterols. Stem trichomes separated from sunflower had no detectable sterol within a GLC limit of detection set at <0.1% total sterol. Alternatively, the leaf wax contains cholesterol, as reported for sorghum leaves (Nes, 1990). Cycloartenol, usually an intermediate that accumulates to trace levels in the sterol mixture, can accumulate to as much as 25% in aging tomato leaves (unpublished data) and similarly can be a major sterol in soybean seeds (Marshall et al., 2001). The inability to detect cycloartenol in some instances may be due to the manner in which sterol was prepared for analysis. For instance, in many early studies sterol purification involved digitonin precipitation that fails to precipitate 4,4-dimethyl sterols along with the major phytosterols; therefore only the Δ⁵-sterol content in the tissues was determined.

FIGURE 9.10 Correlation of the sterol content with the increase in size of seeds. From unpublished data.

In contrast to other lipids, a correlation exists between the amount of sterol and seed size. Thus, as the seed increases in size the amount of sterol increases from as little as a few micrograms per seed in Arabidopsis to approximately 3,000 µg/seed in Strychnos nux vomica (Fig. 9.10). In similar fashion, the amount of sterol in vegetative plant parts (e.g., leaves) increases as the blade size increases during growth and as the plant height approaches maturity (Fig. 9.11). It would appear that the free sterol content of a system is roughly related to the cell number; therefore the amount of total sterol of a system can be an approximate measure of the total number of cells. Thus, Arabidopsis seeds will contain a smaller number of cells than Strychnos seeds.

The amount of total phytosterol in seedlings has been correlated to the level of SMT activity during plant maturation (Fig. 9.12). Transcript expression levels of SMT from Arabidopsis and soybean also vary during plant maturation

FIGURE 9.11 Correlation of the sterol content with the increase in shoot height of sorghum. Inset is a correlation of the increase in sterol content with maturation of the leaf blade, measured as changes in leaf length. Adapted from Heupel et al. (1986).

(Carland et al., 2002; Diener et al., 2000; Shi et al., 1996). For example, young roots, leaves, and stems had higher levels of steady state transcript levels as compared with mature leaves, suggesting a high rate of sterol biosynthesis in growing vegetative tissues. Growing apical meristems appear to be a major site for sterol biosynthesis (Devarenne et al., 2002). In plants, phytosterol synthesis is measurably a slow event, ca. 1.0 pmol/h/100 shoots, consistent with their primary function as architectural components of membranes, and cycloartenol is turned over rapidly at 24 pmol/h/mg protein consistent with its role as an intermediate (Guo et al., 1995). Thus, to meet the continued needs of campesterol and sitosterol formation, carbon from the isoprenoid pathway must be continually made available to cycloartenol synthesis.

The phytosterol composition can change in regard to the structure of the sterol side chains with plant development. A ‘‘switching’’ mechanism in the biosynthesis of phytosterols has been shown in the cucurbits to produce different 24-alkyl sterol stereoisomers with development (Kalinowska et al., 1990). The seeds of squash and pumpkin synthesize 24β-ethyl sterols whereas the seedlings of

FIGURE 9.12 Correlation of the rate of phytosterol synthesis with plant growth and SMT activity. Data adapted from Guo et al. (1995).

these plants synthesize 24α-ethyl sterols (Nes, 1987a). Similarly, the ratio of 24α-ethyl sterol to 24β-ethyl sterol can switch in Kalanchoe; roots (total sterol about 285 µg/fwt) contain exclusively the α-isomer and leaves (total sterol 232 µg/fwt) contain the β-isomer. Flowers (total sterol 160 µg/fwt) have an equal mixture of 24α/β-ethyl sterols. These profiling results strongly suggest that the level and perhaps type of SMT expressed in individual cells and cell types during ontogeny is regulated differentially and is an important determinant of phytosterol diversity.

The principal approach adopted seeks to derive developmental and evolutionary understanding of sterol synthesis by combining the sterol and genetic composition of plants with sterol functions and the morphological evidence of plants at different time points. These studies indicate that sterol biosynthesis can proceed in plants by a cycloartenol pathway and in animals and fungi by a lanosterol pathway (Goodwin, 1981; Nes et al., 1990). In addition, the size and direction of the 24-alkyl group of the sterol side-chain can be an indicator of an early or late stage of development and whether the plants are less or more advanced. Primitive organisms synthesize phytosterols with a 24β-methyl group and vascular plants synthesize sterols with a 24α-ethyl group (Kalinowska et al., 1990). Since C-methylation is an energy expensive process, it seems unlikely that the cell would commit energy to produce the side-chains of ergosterol or sitosterol unless the methylation event was functionally significant. Although few pure enzymes of sterol catalysis have been examined in any detail, it does seem safe to suggest that a relatively small number of SMT enzymes determine the basic structural character of the phytosterol side chains produced in a given species.

The type and amount of phytosterol can change during the life history of plants and fungi yet there is a sterol homeostasis maintained at the cell level throughout development. The question arises as to why certain sterols accumulate and what biology maintains their balance in the cell. In other words, is one sterol as good as another? Can a single sterol play multiple roles? Is there something special about a particular sterol cocktail? One way to get at this problem would be to determine what sterols actually do, to find a way to quantify the value of individual features, and then to determine the extent to which deviations in sterol homeostasis affect function. The occurrence of intermediates is also problematic since different pathways can lead to the same end product. In the first case, the biosynthetic sequence determines the structure of the end product. In the second case, a choice presumably is made in a way that has no impact on the structure of the end product. Nes discussed the possible multiple roles of sterols in plants and fungi (Nes, 1980) and the difference between functional and phylogenetic control of biosynthesis (Parker and Nes, 1992). Evidence in support of phylogenetic control is the cycloartenol-lanosterol bifurcation whereby either precursor can give rise to ergosterol. In contrast, functional control is related to the structure and function of enzymes that act on sterols that, once impaired, interrupt flux to end products resulting in aberrant morphology or cell death. In support of the latter hypothesis is recent work from animal and plant studies that show mutations in the terminal segments of the cholesterol and sitosterol pathway lead to malformations (Clouse, 2002; Herman, 2003).

The hypothesis that sterols play multiple roles in plants and fungi has its origins with the work of Clark and Bloch (1959) who studied the insect nutritional requirements of sterol. Insects cannot make their own sterol due to a block in the pathway before squalene oxide cyclization, therefore they are dependent on an exogenous or dietary source of sterol to satisfy their structural and physiological needs for them (Clark and Bloch, 1959). Clark and Bloch discovered a sterol function other than the bulk membrane role for sterols on the basis of structural and quantitative requirements. Thus, feeding different amounts of a mixture of cholesterol and cholestanol to the insect resulted in the finding that cholestanol can spare (partly replace) cholesterol. The function of cholestanol was considered to be that of a regulatory molecule, not necessarily as a precursor to a hormone, for example, ecdysteroid. Fungi have been found to utilize sterols in multiple roles as well and various actions similar to sparing cholesterol in insects have been described with the yeast ergosterol (Nes, 1987b). The requirement for sparing amounts of a 24-ethyl sterol was demonstrated using cultured celery cells (Haughan et al., 1987). Support for the involvement of the different phytosterols in different physiological roles has been obtained with the fungus G. fujikuroi where there is delayed expression of certain sterolic enzymes that generate different sterol compositions (Nes and Heupel, 1986) and mutation studies in plant sterol synthesis that show changes in sterol composition affects morphology (Lindsey et al., 2003; Schaller, 2004).