Sterols in Algae and Health

Phytosterols (plant sterols) are triterpenes that are important structural components of plant membranes, and free phytosterols serve to stabilize phospholipid bilayers in plant cell membranes just as cholesterol does in animal cell membranes. Most phytosterols contain 28 or 29 carbons and one or two carbon-carbon double bonds, typically one in the sterol nucleus and sometimes a second in the alkyl side chain. Phytostanols are a fully-saturated subgroup of phytosterols (contain no double bonds). Phytostanols occur in trace levels in many plant species and they occur in high levels in tissues of only in a few cereal species. Phytosterols can be converted to phytostanols by chemical hydrogenation. More than 200 different types of phytosterols have been reported in plant species. In addition to the free form, phytosterols occur as four types of "conjugates," in which the 3beta-OH group is esterified to a fatty acid or a hydroxycinnamic acid, or glycosylated with a hexose (usually glucose) or a 6-fatty-acyl hexose. The most popular methods for phytosterol analysis involve hydrolysis of the esters (and sometimes the glycosides) and capillary GLC of the total phytosterols, either in the free form or as TMS or acetylated derivatives. Several alternative methods have been reported for analysis of free phytosterols and intact phytosteryl conjugates. Phytosterols and phytostanols have received much attention in the last five years because of their cholesterol-lowering properties. Early phytosterol-enriched products contained free phytosterols and relatively large dosages were required to significantly lower serum cholesterol. In the last several years two spreads, one containing phytostanyl fatty-acid esters and the other phytosteryl fatty-acid esters, have been commercialized and were shown to significantly lower serum cholesterol at dosages of 1-3 g per day. The popularity of these products has caused the medical and biochemical community to focus much attention on phytosterols and consequently research activity on phytosterols has increased dramatically.

1 Introduction Cholesterol and its relatives possessing the 1,2-cyclopentanoperhydrophenanthrene ring system (Figure 1) form the sterolome, which comprises a chemical library of more than 1000 natural products found in all forms of eukaryotes and some prokaryotes that serve a myriad of biological functions. 1,2 The structural and stereochemical commonality of these compounds derive in large part from the action of oxidosqualene synthases (formerly cyclases) that generate the parent sterol frame. In their 1985 Nobel lecture, Brown and Goldstein stated that “cholesterol is the most highly decorated small molecule in biology”, a comment supported by the many Nobel prizes awarded to individuals who devoted a large part of their lives to research one or more aspects of the chemistry and chemical biology of sterols, their metabolites, or other isoprenoids. 2,3 Figure 1 Common tetracyclic steroid frame containing the 1,2-cyclopentanoperhydrophenanthrene ring skeleton. The first known sterol, cholesterol, was discovered by French chemists as a crystalline component of human gallstones over 230 years ago. In 1789, Francois Poulletier de La Salle observed an alcohol-soluble portion of bile stones, which 10 years later was reported by De Fourcroy to be identical to a waxy material in the fat of putrefied corpses referred to as adipocire.(4) Shortly thereafter, Michel Chevreul established that the crystalline component of bile stones, which gave a melting point of 137 °C (accepted mp for cholesterol is 149 °C), was distinct from adipocire or that of another waxy material from whale, spermaceti (their mp ranged 44–68 °C), and named it “cholesterine” from two Greek words: chole, meaning bile and stereos, meaning solid. 4,5 In English-speaking countries, the name cholesterin was replaced with cholesterol after recognition that the substance was as a secondary alcohol. The correct formula (C27H45O–, which here shows that it has a hydroxyl group although it contains 46 hydrogen atoms) of cholesterol was proposed in 1888 by F. Reinitze, yet it took another 30 years to establish the exact steric representation of the molecule, efforts that led to the Nobel Prizes in chemistry for Wieland (1927) and Windaus (1928). The first connection between cholesterol and human health appeared in 1843 as Vogel showed that cholesterol was present in arterial plaques.(5) Cholesterol was subsequently determined to be widely distributed in the animal kingdom and its isomeric forms, termed “cholesterol bodies” or phytosterol, were shown to frequent the vegetable kingdom.(6) These early attempts at natural product surveys of sterols were based on measurements of melting point, color reaction, optical rotation, and crystalline form. Sterol research changed in the mid- to late 20th century and centered on biomimetic chemistry, tracer work, enzymology, and structure determination using high-field NMR and X-ray diffraction methods, culminating with a broad outline for cholesterol synthesis and the partial or complete purification of all the microsomal-bound enzymes that act on sterols between lanosterol and cholesterol. 7−13 Ergosterol, then cholesterol, was discovered to play multiple cellular roles associated with membrane (bulk role) and signal (sparking) functions, which could be differentiated by structure and amount of compound; 14−17 notably sterols accumulate in cultured plant and animal cells at approximately 3000 fg/cell and in yeast at 20 fg/cell.(18) During this time, there was a renewed focus on regulation of cholesterol synthesis and human physiology.(19) These studies, which involved oxysterols, led to the development of an inhibitor of hydroxymethyl glutaryl CoA reductase (HMGR), atorvastatin (Lipitor), which in 2006 was the best-selling drug in the world grossing more than $12 billion dollars in sales. 20−22 Further work on sterol biosynthesis demonstrated that lanosterol is the product of squalene-oxide cyclizaton in organisms of a nonphotosynthetic lineage and cycloartenol is the product of squalene-oxide cyclization in organisms of a photosynthetic lineage. 23−25 These findings show that sterol biosynthesis can proceed by phyla-specific pathways. Central to the advances of the past two decades is the development of molecular genetic approaches that have witnessed the cloning, primary amino acid sequences, and functional characterization of a large number of enzymes that act on sterol and revealed unexpected inborn errors of cholesterol metabolism. The full exploitation of these genes lies in medical diagnostics, treatment, and the ability to ultimately engineer phytosterol pathways to generate plants with tailored sterol profiles for commercial production. Another spectacular discovery involving 13C-isotopically labeled compounds supplied to microorganisms and plants was the demonstration that the classic acetate–mevalonate pathway to animal cholesterol can be replaced in the biosynthesis of algal sterols and other isoprenoids by a mevalonate-independent pathway. 26,27 The work on the biosynthesis of cholesterol and other sterols is ongoing, and many talented investigators all over the world are contributing to the latest surge of biochemical investigation. Their efforts constitute a collective undertaking of significant importance to several remarkable advances toward the completion of the enzymatic inventory of sterol synthesis, and these new findings, together with a brief examination of prior art in the field of sterols, are discussed in this review. 2 Structure and Distribution 2.1 Sterol Frame and Functional Domains Sterols are amphipathic compounds that originate in isoprenoid biosynthesis with the main frame composed of a nucleus and side chain (Figure 2). Accordingly, the sterol molecule possesses four indispensible domains. In domain A, the polarity and tilt of the C3 OH-group contribute functionally to hydrogen-bond interactions. In domain B, the C4 and 14-methyl groups can affect the A ring conformation and back face planarity, respectively. Alternatively, the number and position of double bonds in the nucleus can affect the shape of the sterol and tilt of the 17(20)-bond. In domain C, the natural configuration at C20, R, determines the conformation of the side chain to orient into a “right-handed” side chain. In domain D, the conformation and length of the side chain, in addition to the stereochemistry of the C24-alkyl group in phytosterols, are critical to intermolecular contacts. These molecular features are crucial for the Δ5-sterol molecule to function in membranes as a flat, elongated compound of approximate volume 794 A3; notably, the cholesterol shape can be mimicked and its structure replaced by sterol-like pentacylic compounds (hopanoids) in bacteria having a smaller volume of 745 A3.(28) Cholesterol has eight stereocenters that give rise to 28 stereoisomers (256), yet only one of them, the natural enantiomer with the 3R, 20R configurations, is utilized as a membrane insert.(17) On the other hand, the size and direction of the 24-alkyl group have phylogenetic significance; 24β-methyl sterols predominate in less-advanced organisms, including many fungi and protozoa, and 24α-ethyl sterols populate the sterol mixture of advanced organisms, typified by vascular plants. 29−31 Figure 2 Perspective drawings of the cholesterol molecule showing four domains of functional importance (left) and the flat elongated structure presumed to form in the membrane (right). 2.2 Nomenclature, Stereochemistry, and the Isoprene Rule The earliest chemical definition for a sterol was provided by Fieser and Fieser. These substances will possess the characteristic perhydro-1,2-cyclopentanophenanthrene ring skeleton (Figure 1).(32) The revised 1989 nomenclature system recommended by the International Union of Pure and Applied Chemistry (IUPAC) and the International Union of Biochemistry (IUB) for the naming of sterols and related isoprenoids permits unambiguous assignments of configuration and introduces a convenient and logical uniformity to the nomenclature of all tetracyclic triterpenoids (Figure 3B).(33) However, the recommendations can cause confusion with regard to the phytosterol side chain structure and nomenclature, particularly with respect to the application of the R and S designation, and in NMR assignments. The revised system has further complications and becomes cumbersome as the complexity of side-chain modifications by biosynthetic alkylations increase. 34,35 We have adopted the older convention (Figure 3A) and rely on the biosynthetic side chain rule, which, for instance, states that the E-methyl group (C26) associated with the Δ24(25)-double bond in the sterol side chain can be shown to originate in C2-mevalonic acid and the Z-methyl group (C27) to arise from C6 (C3′) of mevalonic acid with the numbering of these methyl groups retained in the case of the cholesterol side chain. 2,36 This subtle differentiation between the chemically equivalent but biosynthetically distinct terminal methyl groups can be lost in the numbering procedure adopted by the IUPAC–IUB recommendations. Another convention relevant to this rule involves chiral substituents at C24 to be designated α and β, but this α/β notation is unrelated to the similar notation for substiuents on the ring structure. The corresponding family of phytosterols are then related to structures based in ergostane (24β-methyl), campestane (24α-methyl), and stigmastane (24α-ethyl) or poriferstane (24β-ethyl). The systematic names of the sterols are treated as derivatives of 5α-cholestan-3β-ol or that of a related stanol. As relevant, the corresponding trivial name for a sterol may be used in place of the systematic name for its familiarity with the reader such as cholesterol rather than cholest-5-en-3β-ol. 1,2,5,12 Figure 3 Two systems recognized for numbering of carbon atoms of the sterol nucleus and side chain. The conventional system, which incorporates the α/β-chiral descriptors in nucleus and side chain is based on the Fieser and Fieser book Steroids and biosynthetic considerations as discussed in refs (1, 2, and 5) (panel A), versus the 1989 IUPAC recommendations (panel B). Because sterols are derived from the C30 squalene, they are a class of triterpenoids. These tetracycles are generated from the linear combinations of the C5-isoprenoid building block, isopentyl diphosphate. In order to conform to the empirical isoprene rule, triterpenes (C30) are the products of joining 6 × 5 C5-units.(37) Thousands of triterpene products have been reported from plants, and they arise from more than 80 different carbon skeletons; the triterpene cyclase–baruol synthase recently cloned can generate as many as 23 products.(38) Sterols therefore form a unique family of triterpenes that can be defined on biosynthetic reasonableness. For this alternate definition, the focus is on the reaction mechanism and ignores the precise isoprenoid character of the cyclization product, assuming only that the intermediate adopts steroidal character during cyclization to produce a tetracycle compound of specific structure and stereochemistry. Thus, a true sterol is formed by the electrophilic cyclization reactions that pass through a transition state similar to the trans–syn–trans–anti–trans–anti configuration affording a protosteroid C20 cation.(2) The cyclization product(s), lanosterol or cycloartenol and in rare cases parkeol, should contain absolute configurations at C3 and C20 of the R-orientation. In light of recent work from Corey’s laboratory, the intermediate C20 cation is now known to assume the 17β-side chain (pseudoaxial side chain generated by a 17α-hydrogen atom) required of the synthase to establish the natural C20 configuration of the sterol product. 39,40 The structure and stereochemistry for several structural isomers in the tetracyclic triterpene series, including lanosterol, cycloartenol, euphol, tirucallol, and cucurbitacin, have been established by X-ray crystallography. 41−44 The lanostane and cycloartane sterol skeletons possess relative stereochemistry of 5α, 8β, 10β, 13β, 14α, 17α, and 20α configurations, consistent with their stereospecific formation involving the protosteroid cation, which agree with the stereochemistry in the cholestane frame resulting from their metabolism (Figure 4). Figure 4 The cyclization of squalene-2,3-oxide (i) folded in either the chair–chair–chair–boat–unfolded (ii) or chair–boat–chair–boat–unfolded (ii) conformations to yield a cation (ii), which can stabilize to produce the dammarane, tirucallane, euphane, and cucurbitane skeletons (via path a) or cycloartane and lanostane skeletons (via path b). Generally, the cyclization products contain a 3β-OH group and a Δ8-bond or in the case of cucurbitacin a Δ5-bond. 2.3 Variation in Sterol Construction Principal differences in cholesterol and related phytosterols are in the side chain (C20–C26/C27), which has different degrees of substitution and unsaturation. Chemical surveys of the sterol composition of prokaryotes, eukaryotes, and sedimentary organic matter show that there are at least 250 sterols and related steranes; in corn, 60 different sterols have been characterized. 45−47 Higher plants are a mixture of 24-alkylated sterols, 24-ethyl and 24-methyl sterols generally accounting, respectively, for more than 70% and less than 30% of the total sterol. Djerassi and his associates, using a computer-assisted program, calculate that natural sterols may have as many as 1778 different structures; many of them may be found in marine organisms, which are known to synthesize highly bioalkylated sterol side chains.(48) The number of cholesterol variants can increase by an order of magnitude by including cholesterol/24-alkyl sterol derivatives to esters, glycosides, and sulfates and those metabolites that retain the perhydro-1,2-cyclopentanophenanthrene skeleton. In animals, cholesterol is converted to sex hormones, bile acids, vitamin D, and different classes of “oxysterols”, in insects to ecdysteroids, and in nematodes to dauer steroids, whereas in plants, cholesterol is converted to C27-spirostanols to form glycoalkaloids and saponins or C23 cardenolides. 1,2,49 Alternatively, the C28- and C29-phytosterols metabolize to the plant defense withanolides (from 24(28)-methylene cholesterol), to plant growth hormone brassinosteroids (from campesterol), or to fungal sex hormones antheridiol and oogoniol (from fucosterol) in the Oomycetes. 50−52 3.0 Biogenetic Considerations: Historical Pedagogy 3.1 Ionic Reactions According to the currently accepted hypothesis, the formation of steroids proceeds by a cationic cyclization process. This theory has it roots in the biogenetic isoprene rule of Ruzicka and his associates in Zurich who considered biosynthesis of triterpenoids was initiated by an electrophilic attack on a double bond of a linear polyprenoid substrate forming a cyclic (or polycyclic) intermediate cation, which, in turn, can then undergo various transformations and rearrangements.(37) Carbocationic intermediates can also arise in the presterol segment of isoprenoid biosynthesis, and they include allylic pyrophosphates that generate allylic carbocations on elimination of pyrophosphate. Ionization of the allylic pyrophosphate leads to formation of a charge-stabilized allylic cation. The cations can react with alkenes to form new carbon–carbon bonds using Δ2-IPP (dimethyl allyl pyrophosphate) as a source of allylic carbocations, such that after trans elimination of a proton, a new prenyl diphosphate is generated that is five carbons larger (Figure 5). 5,53 For these biochemically rare electrophilic additions to a double bond to proceed, the divalent metal ion binds to the pyrophosphate moiety of the allylic cosubstrate so as to make it a better leaving group in the ionization step. Two structurally unrelated classes of isopentyl diphosphate isomerase (IDI) are known. Type I IPP (IDI-1) utilizes a divalent metal in a protonation–deprotonation reduction. In contrast, the type II enzyme (IDI-2) requires reduced flavin, raising the possibility that the reaction catalyzed by IDI-2 involves the net addition or abstraction of a hydrogen atom.(54) Other coupling reactions contribute to squalene biosynthesis, the most notable being that involved in the two-step reaction yielding squalene from farnesyl diphosphate (FPP). 55,56 First, one molecule undergoes loss of PPi and addition of the allylic carbocation to the alkene end of the other molecule of FPP, accompanied by loss of a proton, to form presqualene pyrophosphate. In the second step, presqualene pyrophosphate (PSPP) loses PPi, and the presumptive cyclopropylcarbinyl carbocation undergoes ring opening and reduction by NADPH to squalene (Figure 6). Figure 5 The prenyl transferase reaction and formation of farnesyl diphosphate (FPP). Prenyl (C5) units composed of isopentyl diphosphates (IPP) are assembled in a head to tail fusion to yield C10 (geranyl diphosphate, GPP) and in repeat fashion to C15 (FPP) isoprenoids. Alternate isoprene unit assemblies yield a variety of different structures notable in terpene metabolism. Enz = enzyme and M++ can be Mn2+ or Mg2+ ions. Figure 6 Proposed mechanism for the coupling of farnesyl diphosphate (FPP) to form squalene. Adapted with permission from ref (53). Copyright 2009 American Chemical Society. Rahier and co-workers have drawn attention to ionic processes that generate sterol cations concomitant with allylic rearrangements, double bond C-methylations, or reduction reactions in the postsqualene segment of sitosterol biosynthesis (Table 1).(11) The most celebrated and best characterized of these electrophilic processing enzymes has been the sterol C24-methyltransferase (24-SMT), attracting justly deserved attention, not only for its centrality in phytosterol diversity but for possible inactivation by catalyst-specific drugs in ergosterol biosynthesis to treat human infections. A broad unifying mechanistic framework for the origin of all phytosterol side chains that included the intermediacy of a cationic intermediate at C25 in the conversion of Δ24(5)- to Δ24(28)-sterols was provided by Castle et al. in their demonstration of the incorporation of the S-methyl of methionine into sitosterol in Pisum sativum.(57) Table 1 Electrophilic Reactions Catalyzed by Sterol Biosynthesis Enzymes 3.2 Formation of Steroidal Backbone The usual pathway to the C30 primary sterols proceeds along the well-established isoprenoid trail that leads from the “active isoprene unit”, isopentenyl diphosphate, to the C30 triterpenoid squalene. The biosynthesis of C30 sterols from squalene and thence to cholesterol can be outlined in three major stages as envisioned by Bloch:(58) acetate → isoprenoid intermediate → cyclization product → cholesterol. Stage 1 involves (i) conversion of acetyl CoA into acetoacetyl CoA catalyzed by acetoacetyl CoA thiolase (AACT), (ii) conversion of acetoacetyl CoA into 3-hydroxyl-3-methylglutaryl CoA by hydroxyl-3-methylglutaryl CoA synthase (HMGS), (iii) conversion of 3-hydroxy-3-methylglutaryl CoA into mevalonic acid (MVA) by 3-hydroxy-3-methylglutaryl CoA reductase (HMGR), (iv) conversion of MVA into phosphomevalonate by phosphomevalonate kinase (MK), (v) conversion of phosphomevalonate into diphosphomevalonate by phosphomevalonate kinase (PMK), and (vi) conversion of diphosphomevalonate into isopentyl diphosphate (Δ3-IPP) by mevalonate diphosphate decarboxylase (MVD). Thus, in the first stage MVA is transformed into IPP by two phosphorylation steps at C5 of MVA and a decarboxylation/elimination step; IPP, the basic C5 building block, is then added to prenyl diphosphate cosubstrates to form longer chains. Δ3-IPP itself is insufficiently reactive to undergo ionization to initiate the condensation of higher isoprenoids. Therefore, it is first isomerized to the allylic ester Δ2-IPP through an antarafacial rearrangement followed by head to tail condensation of Δ2- and Δ3-IPP to form geranyl diphosphate by geranyl diphosphate synthase (DPS). In the second stage, the condensation reaction is repeated by the addition of Δ3-IPP producing the C15 allylic product farnesyl diphosphate. Two molecules of farnesyl diphosphate condense tail to tail to the C30 acyclic polyene squalene by the action of squalene synthase (SQS). The C30 symmetric olefin undergoes oxidation to form S-oxidosqualene via an NADPH-dependent mono-oxygenase reaction catalyzed by squalene epoxidase (SQE), and this substrate can be cyclized by an oxidosqualene–sterol synthase to yield the steroidal backbone structure represented in lanosterol. In stage 3, lanosterol is converted to cholesterol (section ). In the early phase of cholesterol research, it was not immediately apparent that the C27 structure of cholesterol was related to lanosterol, since it failed to be divisible by C5 units. To establish isoprenoid character, several groups incubated [1-14C]acetate and [2-14C]acetate with liver slices affording a decisive pattern in the distribution of acetate carbon atoms in the labeled cholesterol. As shown in Figure 7, where “M” denotes an acetate methyl and “C” an acetate carboxyl carbon, there are three places in the molecule where a repeating C5 pattern typical of isoprene can be recognized. These tracer studies also provided the foundation for the acetate–mevalonate pathway in sterol biosynthesis. 1,2,5 The stereochemistry of the enzymatic reactions involved with biosynthetic steps that lead to squalene, lanosterol, and cholesterol have been completely elucidated by the work of Popjak and Cornforth using six species of MVA stereospecifically labeled with either 2H or 3H at C2, C4, or C5. 5,59 Seo and co-workers complemented this work by feeding [5-13C2H2]MVA together with [2-13C2H3]acetate or [1,2-13C2]acetate to yeast and cultured plant cells.(60) Using 13C NMR spectroscopy, they confirmed the fate of all the relevant hydrogen atoms comprising MVA incorporated into 13C-labeled ergosterol, cycloartenol, and sitosterol. Figure 7 Distribution of acetate carbon atoms found in cholesterol; a repeating pattern of five carbon atoms (isoprene unit), surrounded by dotted lines, is recognizable in three places in the molecule. Adapted with permission from ref (5). Copyright 1986 Plenum Press, New York. Since the late 1960s, it has been known that two major cyclization pathways exist for the conversion of oxidosqualene to steroidal tetracycles; lanosterol is formed in organisms of a nonphotsynthetic lineage, and cycloartenol is formed in organisms of a photosynthetic lineage by independent synthase enzymes. 2,9 This biosynthetic bifurcation in sterol biosynthesis is one of the most interesting phylogenetic markers available because it has no apparent influence on the structure of the functional steroid at the end of the pathway and the enzymes themselves are membrane-bound though not in plastids (e.g., cycloartenol). By carrying out experiments with doubly labeled [2-14C(4R)4-3H1]MVA fed to animal and plant tissues, Goad and Goodwin demonstrated that cycloartenol having a 3H/14C atomic ratio of 6:6 is not produced by rearrangement of lanosterol having a 3H/14C atomic ratio of 5:6, consistent with retention of label at C19 in the biosynthetically formed cycloartenol.(61) Altman and co-workers studied cycloartenol biosynthesis in the alga Ochromonas malhamensis. Incubation of oxidosqualene bearing a chiral methyl group (H,2H,3H) at C6 with the plant synthase revealed that the stereochemistry of the cyclopropane ring closure proceeds with retention of configuration.(62) These variant synthase-mediated substrate transformations were formerly postulated to involve a concerted ring annulation that led to a C20 protosterol intermediate in which the cationic side chain at C17 is α-oriented (C17β-hydrogen), affording the unnatural C20S arrangement. To form the natural 20R configuration, it was postulated that intervention of an electron-donating “X-group” on the synthase takes place to transiently neutralize the C20 charge. 63,64 Utilization of the X–-group permits rotation about the 17(20)-bond of 120° prior to hydride migration. A final elimination reaction takes place at C9α to produce lanosterol. 65,66 In contrast to formation of a “flat” Δ8(9)-sterol, formation of “bent” cycloartenol requires the product bend through almost 90° to accommodate the 9β,19-cyclopropane-8β-H bridgehead in which the 19-CH3 and 8H-atom are cis to one another. 5,14,24,67,68 For “bent” cycloartenol formation, a distinct enzyme–substrate interaction from the one involved at C20 is considered to intercept the positive charge generated at C9α (Figure 8). 9,67 Ring construction in this pathway requires a 1,3-anti elimination of a proton from the C19 angular methyl group and the added nucleophile. If this is not done, the final step, migration of a hydrogen from C19-methyl, will be cis to the C9 hydrogen transfer instead of trans in order to conform with the biogenetic isoprene rule. Subsequent withdrawal of the X– then permits closure of the cyclopropane ring in a trans manner with concomitant removal of the C19 proton, which is tantamount to a double inversion mechanism. These postulates are revisited in section . Figure 8 Interpretation of the mechanism of squalene-2,3-oxide cyclization to lanosterol and cycloartenol according to refs (61, 63, and 64). The cyclization mechanisms are hypothesized to require “X– group” (any electron-donating group on the enzyme) participation to generate the C20R configuration, which requires the side chain of the X– group bound protosteroid to rotate from left to right about C20 and formation of “bent” cycloartenol from a “flat” 9β,19-cyclosteroid intermediate. See text. Adapted with permission from refs (61 and 64). Copyright 1977 American Chemical Society and 1968 Biochemical Journal. 4 Recent Advances in Sterol Biosynthesis 4.1 The Genome–Sterol Metabolome Congruence The different molecular libraries that constitute isoprenoid–sterol metabolomes across Kingdoms are organized through a series of discrete assemblies of enzymatic reactions, which are characterized compartmentally. The acetate–MVA pathway to squalene oxide is considered to be the main route to the production of steroidal backbones. Recent international efforts have resulted in the complete sequencing of the model plant Arabidopsis, nonpathogenic fungus Saccharomyces, and human genomes. 69−71 In the entirely sequenced genomes of these organisms, homologues for all genes of the acetate–MVA pathway to sterols are present. The functional genomics approach together with the establishment of defective biosynthesis steps in humans and generation of yeast and plant mutants in sterol biosynthesis enabled the elucidation of structural genes for the individual enzymes in sitosterol, ergosterol, and cholesterol formation. 72−74 Notably, sterol biosynthesis in plants and fungi differ markedly from that in animals since these organisms contain more sterol genes than do animals. In nonanimal systems different sterolic genes can encode for similar reaction steps, for example, sterol C24-methyltransferase, sterol C14-demethylase, or sterol methyl oxidase, whereas mammals generally have only a single gene for each enzymatic step. The best described gene–gene product pairing is in the yeast pathway for which every relevant gene in the conversion of lanosterol to ergosterol has been identified. The principal enzymes of lanosterol conversion to cholesterol are coded for by nine genes (section ). The fungal 24-alkyl sterol pathway has acquired at least four additional genes from the human pathway that give rise to a sterol C24-methyltransferase, sterol C22-desaturase, sterol C24–24(28)-reductase and sterol Δ24(28)- to Δ24(25)-isomerase as reported in the ascomycetes, Saccharomyces, and Gibberella.(18) Plants, variably from algae to tracheophytes, have evolved all 11 of the fungal/animal sterol genes and can code for two more of them, the sterol Δ25(27)-reductase and 9,19-cyclopropyl to Δ8-isomerase. To date, the number of sterol genes involved in conversion of lanosterol or cycloartenol to Δ5-sterol is 15. Isoforms can exist to alter the number of enzymatic reactions in a specific pathway. 4.2 Biosynthesis of Squalene: MVA versus MVA-Independent Pathways At least two alternative biosynthetic pathways to Δ2-(dimethyl allyl pyrophosphate) and Δ3-isopentenyl diphosphate (formerly pyrophosphate) have been shown, by biosynthetic labeling studies, that can supply prenyl units to squalene, which in turn is converted to sterols (Figure 9). One involves acetate directly incorporated into the MVA pathway after activation into acetyl CoA. The other, known as the methyl-d-erythritol 4-phosphate (MEP) pathway or mevalonate-independent pathway, utilizes triose phosphate units as the precursor and is located in the plastid of plants. 75−77 The former pathway uses seven enzymes, some of which have crucial isoforms that contribute to drug sensitivities, compared with the latter pathway that consists of eight reactions catalyzed by nine enzymes. Many of these enzymes have been characterized structurally. In the newly discovered mevalonate-independent pathway reported independently by Rohmer who detected it in eubacteria and Arigoni who detected it in higher plants, 75,76 Δ2-IPP (dimethyl allyl diphosphate) is formed directly from pyruvate rather than via acetyl CoA–MVA and Δ3-IPP. Terpenoids of the C10, C15, C20, and C40 skeletons are generally synthesized within the plastid and sterols synthesized in the cytosol of higher plants. Goad and co-workers reported that mitochondria of some parasitic protozoa have the capability to convert leucine to MVA, which then converts to IPP and ergosterol via the cytosol.(78) Alternatively, isopentenyl diphosphate can undergo catabolism through an “MVA shunt” pathway (also known as the Popjak shunt”) that forms a non-sterol-forming pathway to redirect flux to small molecules or fatty acids in animal, fungal, or plant tissues. 79−81 The MVA shunt, a multidirectional pathway, can allow carbon from amino acids to be incorporated into IPP; isovaleric acid has in its metabolism intermediates common to those in the shunt pathway. Figure 9 Overview of compartmentalized isoprenoid–sterol biosynthesis pathways. Feeding studies of [2-13C]leucine or [1-13C]glucose distinguish the leucine–HMGCoA, acetate–mevalonic acid (MVA) or the 2-C-methyl-d-erythritol-4-phosphate (MEP) pathways that contribute carbon to isopentenyl diphosphate (IPP). The 13C-labeling pattern of ergosterol (synthesized in all major eukaryote Kingdoms) reveals which of the pathways operate in a given organism and whether cross-talk between the pathways exist. In the forward direction, IPP converts to ergosterol or IPP can retroconvert via the MVA shunt to HMGCoA and to acetate. See text. It has been possible to establish the contribution of these variant isoprenoid pathways to sterol biosynthesis in plants and microorganisms by the retrobiosynthetic approach. When, for example, [1-13C]glucose, [2-13C]acetate, l-[2-13C]leucine, [2-13C]MVA, or [2,3,4,5-13C4]1-deoxy-d-xylulose is used, the MVA and MEP pathways yield distinctive labeling patterns in IPP and sterols. Thus, in higher plants, sitosterol can be made by either the MVA or MEP routes; the MVA pathway is always senior to the MEP pathway. In the diatom Rhizosolenia setigera that can synthesize 24-ethyl sterol, the MEP pathway is favored.(27) In the photosynthetic alga Scenedesmus obliquous and nonphotosynthetic yeast-like Prototheca wickerhamii (cycloartenol route) that synthesize 24-methyl sterols, the MEP pathway operates to the exclusion of the acetate–MVA as shown by the distinctive 13C-labeling pattern of ergosterol, which possesses 10 rather 15 of its carbons enriched with 13C after incubation with [1-13C]glucose (Figure 9). 82,83 4.3 Cyclization of Squalene Oxide to Lanosterol or Cycloartenol By far the best studied of the enzymes involved in sterol biosynthesis are the lanosterol and cycloartenol synthases. For cyclization, the reaction requires (3S)-2,3-oxidosqualene first to adopt a preorganized chair–boat–chair conformation. Protonation of the epoxide ring then triggers a cascade of stereospecific ring-forming reactions to protosterol 3 channeled to specific outcomes for the different synthases (Figure 10). Recent findings of the Corey group, studying lanosterol formation, demonstrated that the initial substrate cation in the chair–boat–chair–boat ring conformer 3 generates a 17β-side chain (17α-hydrogen) that connects to the sterol side chain at C20; the resulting cis C17H and C20H geometry orients the side chain into a “left-handed” structure.(83) The protosterol cation converts to lanosterol 11 by elimination of the 9α-H atom (originally the 9βH of the protosteroid cation), which requires nuclear rearrangements that involve a series of 1,2-shifts of hydride and methyl groups in an antiparallel manner coupled to quenching of the positive charge at C20. The C9β-hydrogen on the protosteroid cation is then lost to the medium as a proton when the 8α-methyl migrates to C14 to form intermediate 5, which undergoes proton abstraction at C9α to form the 8,9-double bond in 7.(66) Lanosterol cyclization may be considered complete when sterol 7 assumes the physiological conformation of the energetically favored “right-handed” side chain (C17α-H and C20R-H trans to one another in the usual view of the molecule) after product release (Figure 9).(43) Figure 10 The cyclization of squalene-2,3-oxide (1) catalyzed by the lanosterol synthase (LAS) or cycloartenol synthase (CAS) to give true sterols. Ring annulation proceeds by synthase (cyclase)-specific cationic mechanisms involving a common protosteroid cation (17α-H/17β-side chain orientation, 3). The cyclization cascade terminates with formation of either (i) intermediate 4 (LAS) followed by its deprotonation at C9 or C7 to form lanosterol 11 or lanosta-7,24-dienol 12, respectively or (ii) intermediate 6 (CAS), which depending on the position of deprotonation at C7, C19, or C11 yields lanosta-7,24-dienol 13, cycloartenol 14, or parkeol 15, respectively; trans-cyclization of the chair–boat type of system in 6a (natural intermediate compared with 6b as discussed in the text) affords the 9β-H instead of a 9α-H, which is necessary to complete biosynthesis of cholesterol. The asymmetric conformation of bound intermediates 4 and 6 with a 20α-H-atom (20R-configuration) eclipsed to 17α-H-atom produce an enzyme-bound “left-handed” rotamer at C20. After release of true sterols 7–10 from the active site, the sterol side chain can rotate about C20, such that C21 and C22 lies in a 1,3-diaxial relationship with C18 to form the energetically favorable “right-handed” conformation of 11 to 15 utilized in membranes. The postulated mechanism of cyclization of squalene oxide into cycloartenol (and parkeol) is essentially the same as that for lanosterol, except for the final 9β,19 cyclopropane ring closure with the 9β-H atom migrating to C8 instead of C9α proton elimination. Nes found that the solid state and solution studies unambiguously show that cycloartenol is “flat” of A/B/C-rings in the chair/half-chair/twist-chair conformer (Figure 11).(44a) An alternative reaction sequence to generate cycloartenol from intermediate 3 (C9β) is to proceed to 6b (C8β) of constrained boat structure (chair–boat–chair) and to 14, rather than to 6a of chair–chair–chair conformation, which has resemblance to 4 to 12 (Figure 10); noticeably, the C9-cations in 6a and 6b considered in the final cyclopropane formation are on opposite sides of the reaction intermediate. Figure 11 Structure of “flat” cycloartenol established by NMR (left) and X-ray (right) measurements. Reprinted with permission from ref (44a). Copyright 1998 American Chemical Society. Cloning and mutagenesis studies of cycloartenol and lanosterol squalene oxide (= oxidosqualene) synthases [OSS] suggest that the native enzymes in plants and fungi are catalytically and structurally similar to the human oxidosqualene–lanosterol and bacterial squalene–hopene cyclases. 65,84−86 The cloned and purified human oxidosqualene cyclase is active as a monomer, whereas the corresponding purified enzyme from bovine liver exhibited active forms of 70 and 140 kDa, suggesting dimer organization. 87,88 Insight into lanosterol scaffold formation from the recently determined structure of human oxidosqualene synthase reveals the presence of a cavity, shaped so as to accept the substrate prefolded product-like conformation.(86) Structural analysis, mutagenesis, and inhibition experiments have shown that the protruding part of the protein contains a lipophilic channel leading to the active-site cavity that can influence substrate recognition (squalene versus oxidosqualene) and access to the active site.(89) Directed mutagenesis of cycloartenol and lanosterol synthases have shown that point mutations in the active site can alter the cyclase product specificity. Mutations of aliphatic residues in the cycloartenol synthase of Arabidopsis thaliana and Dictyostelium discoideum resulted in a partitioning shift from cycloartenol formation in the wild-type to lanosterol, parkeol, and 9β-lanosta-7,24-dienol.(68) In a study on cycloartenol synthase redesign, Matsuda and co-workers achieved the nearly complete shift from cycloartenol to lanosterol (99%) formation through engineering a double mutant of His477Asn and Ile481Val.(90) These results show that only small changes in the active site topography of these enzymes are required to fashion a fungal/animal synthase from a plant synthase. Lanosterol, parkeol, and lanostane derivatives with the 9βH-sterochemistry have recently been found to occur naturally in plants, 12,91 raising questions of their biosynthetic origin, that is, from preformed cycloartenol. Detailed investigations of the steric and electrostatic effects of conserved aromatic residues through site-directed mutagenesis experiments have revealed cation−π interactions that may influence the product outcome. Examination of yeast ERG7 homology structure and human OSC structure suggest that Tyr707 and Tyr99 residues in the Saccharomyces cerevisiae ScERG7 might play an important role in stabilizing the C8 cation during the formation of the second cyclohexyl ring and the final lanosterol C9 cation.(65) On the other hand, Thomas and colleagues suggest the relevant active site amino acid in the human synthase occurs at His232, which is a strictly conserved residue among the oxidosqualene cyclases,(89) whereas Christianson noted that Tyr503 could possibly accept this proton because it is closer and better-oriented with regard to the C9 proton that is ultimately eliminated to form lanosterol.(92) Although cycloartenol is the first cyclized product in plants, lanosterol has been considered as an intermediate in phytosterol biosynthesis. Several reports involving plant sterol biosynthesis show similar rates of [2-3H] (or 14C)-lanosterol and [2-3H] (or 14C)-cycloartenol incorporation into phytosterols suggesting that cycloartenol might transform to lanosterol via a 9β,19-cyclopropane to Δ8-isomerase catalyzed reaction as a preliminary step to phytosterol formation. 93−95 This reaction would allow for the cyclosterol intermediate (cycloartenol or some other cyclosterol) to rearrange into the more stable structure that can yield products having the C9α-H in Δ5-sterols. However, primary sequences of lanosterol synthase were discovered recently in three different laboratories from dicotyledonous plant species, including Arabidopsis thaliana, Panax ginseng, and Lotus japonica, using a yeast expression system that suggest that lanosterol can be synthesized directly and independently from cycloartenol. 96−98 Conservative patterns suggest that the Ile481 residue is partly responsible for the catalytic differences between the lanosterol and cycloartenol synthases.(68) These findings together with reports showing that radiolabeled squalene oxide converts to lanosterol in a cell-free latex preparation of Euphorbia lathyris and intact Arabidopsis thaliana converts [613C2H2]MVA to lanosterol and phytosterol with retention of three 2H-atoms at C19 in the labeled products show the cycloartenol–lanosterol bifurcation in sterol biosynthesis is not absolute. 99,100 Moreover, the cycloartenol and lanosterol synthases have been cloned and functionally expressed from the same bacterium Stigmatella aurantiaca, and the lanosterol synthase has been cloned from other primitive organisms, including Methylococcus capsulatus and Gemmata obscruiglobus.(101−104) 4.4 The Core Pathway: Lanosterol Conversion to Cholesterol Three different approaches have been used to study the nature and sequence of steps involved with sterol biosynthesis: organic, enzymatic, and genetic approaches. The enzymatic approach to understand and control formation of the sterol structure was hampered by the low abundance of sterol enzymes in cell-free preparations, as well as difficulties associated with purifying microsomal proteins to homogeneity. By the early 1980s, it was evident primarily from the pioneering efforts of Gaylor and his co-workers at cataloging the properties of enzymes that act on sterols that an accurate sequence of 10 enzymatic reactions performed by nine distinct enzymes can be written for the lanosterol conversion to cholesterol as outlined in Table 2.(10a) These enzymes were shown to be integral membrane-bound proteins of the endoplasmic reticulum. Table 2 Biosynthesis Steps Involved in the Conversion of Lanosterol to Cholesterola a Steps 3 to 4 and 5 to 6 involve two discrete enzymatic reactions, 4-SMO and 4-SDC; see text. Cholesterol synthesis requires one molecule of C30-4,4,14-trimethyl sterol converted to C27-4,4,14-trisdesmethyl sterol followed by the formation of a saturated isooctyl side chain and Δ5-bond in the final product. The stoichiometry of cholesterol synthesis from lanosterol is therefore lanosterol + 15NADPH + 4H+ + 10O2 → cholesterol + 2CO2 + HCOOH + 15NADP+. This process can be divided into two stages: stage 1 (reactions 1–6) represents the nuclear demethylation reactions that fashion the lanosterol frame into the cholesterol structure. The product of reaction 6 lacks methyl groups at C4 and C14 and contains a 3β-OH group that is distinct from the one generated by the cyclization of squalene-2,3-oxide to lanosterol. During these conversions of the C4-sterol to a C4-desmethyl sterol, a stable 3-keto sterol intermediate is formed. For stage 1, the process consumes 12 NADPH, 2 NAD+, 9 O2, and 1 H+. Reactions 3 and 4 are enzymically coupled and repeated with the resulting C4 methyl product catalyzed in reactions 5 and 6; in both sets of reactions two distinct enzymes, sterol C4-methyl oxidase (4-SMO) and sterol C4-decarboxylase (4-SDC, also referred to as 3β-hydroxysteroid dehydrogenase/C4-decarboxylase), are utilized to carry out the overall conversion of the sterol C4-demethylation reaction. Key to this series of reactions is that the equatorial (in the plane of the sterol nucleus) methyl group of the 4,4-dimethyl and 4-monomethyl substrates is recognized for catalysis. To maintain stereochemical consistency in 4,4-dimethyl substrate recognition by 4-SDC, during decarboxylation the methyl group that occupies the 4β-(axial) position of the C4-dimethyl sterol epimerizes to the more stable 4α-(equatorial) position in the 4-methyl 3-ketosteroid product and, in doing so, re-establishes the C4α configuration. Thus, the natural occurrence of C4β-methyl sterols seems unlikely. Stage 2 (reactions 7–10) involves the rearrangement of the Δ8-bond to the Δ5-position and saturation of the side chain double bond. Reactions in stage 2 consume 3 H+, 3 NADPH, and 1 O2, and the enzymes act sequentially to convert the ring structure from a Δ8 to Δ7 to Δ5,7 to Δ5. Two enzymatic studies completed on the cloned human sterol C8–C7 isomerase and microsomal rat C24(25)-reductase after Gaylor postulated a lanosterol–cholesterol pathway clarified the later sequence of reactions to cholesterol. 105,106 These enzymatic studies reveal that the Δ24(25)-double bond of zymosterol is reduced prior to Δ5-desaturation. The nine catalysts that convert lanosterol to cholesterol may be considered the core enzymes of Δ5-sterol biosynthesis. As more enzymes have been isolated from an increasing number of sources, it has become clear that they fall into only a small number of reaction types and that the chemical names give varied indication of this. In an attempt to rationalize this situation, Table 2 includes trivial names of the parent catalysts and a related system of abbreviations proposed by the author. Although cholesterol biosynthesis is often linked to the formation and metabolism of C5 (isoprenoid = isopentenoid = terpenoid)-units, the “sterol biosynthesis pathway” can be defined as that set of enzymes that act on the sterol structure. The most direct route to cholesterol will be dependent upon the relative specificities of the enzymes for a particular sterol substrate thereby giving rise to the kinetically favored pathway. When multiple routes are postulated, it has been possible to draw a sterol biosynthesis matrix for an organism, which can predict the main or trace sterols which might be present, particularly after exposure to a sterol biosynthesis inhibitor or genetic defect. In the case of cholesterol biosynthesis from lanosterol, two intersecting routes have been postulated. The choice of pathway is determined by the stage at which the double bond at C24 in the sterol side chain is reduced. If C24 double bond reduction is retained until the last reaction, cholesterol synthesis proceeds via cholesta-5,24-dienol (desmosterol) (Bloch pathway). On the other hand, early Δ24-reduction involving lanosterol can proceed to cholesta-5,7-dienol (7-dehydrocholesterol) and cholesterol (Kandutsch–Russell pathway). A common interpretation regarding which pathway is utilized involves the positioning of the Δ24-reductase in cholesterol biosynthesis such that skin and intestines, which have higher sterol C24-reductase activities than liver or brain, proceed via the C24–C25-terminal intermediates. 10b−10d Regardless of tissue specificity, the kinetically favored pathway for cholesterol biosynthesis appears to involve the Kandutsch–Russell pathway. The relevant committed step that distinguishes sterol from isoprenoid–triterpenoid biosynthesis occurs at the cyclization of oxidosqualene. Major control points in sterol biosynthesis may arise in the primary pathway before squalene formation at hydroxymethyl-glutaryl-CoA reductase (HMGR) (coarse control)(3) or after squalene formation at the sterol C24-methyltransferase (24-SMT) step (fine control)(107) specific to organisms other than animals. Cofactor control by differential allocation of oxygen, NADPH, and AdoMet can further influence reaction rates and product distributions in cholesterol or phytosterol biosynthesis. 4.5 C24-Alkylation–Reduction Bifurcation in Phytosterol Synthesis Early attempts to deduce sterol biosynthetic pathways in systems other than animals were based largely on indirect approaches that included structural and stereochemical correlations of co-occurring metabolites and in vivo tracer studies. Using microsome preparations of corn, Rahier, Benveniste, and their co-workers systematically worked through the major enzymatic reactions from cycloartenol to sitosterol and established substrate preference for many of these enzymes.(11) The accumulated evidence supports the kinetically favored pathway to Δ5-sterol synthesis from cycloartenol illustrated in Figure 12. 2,11,72 Cholesterol can accumulate in plants by reduction of the Δ24-bond depending on the expression of the sterol C24-methyltransferase enzyme, as demonstrated in genetically modified Arabidopsis plants.(108) The 24-SMT is recognized first in the phytosterol reaction sequence in plants and in some fungi. The typical products of transmethylation of the substrate Δ24-bond are the C24-methyl Δ24(28)- and Δ25(27)-olefins, or in rare cases they become the 24-methyl Δ23(24) (corn)- or Δ24(25) (protozoa)-olefins (Figure 13). 109−111 Terrestrial organisms, as well as many pathogenic fungi and protozoan parasites synthesize 24-alkyl sterols, whereas animals, especially humans, lack the 24-SMT gene. Biomimetic studies of electrophilic alkylations of a remote double bond support the detection of multiple outcomes of enzyme-generated 24-alkyl sterol olefins(112) Separate biosynthetic routes, which have the following differences—major sterols are 24α-ethyl sterols or 24β-ethyl sterols or cholesterol and cyclopropyl or 24-methyl(ene) sterols replace 24-ethyl sterols as the major compounds in the sterol mixture, can be generally categorized and result in the production of diverse sterol compositions. Figure 12 A generalized route from cycloartenol 1 to Δ5-24α/β-ethyl sterols, stigmasterol 18 and poriferasterol 23, respectively. Figure 13 The sterol alkylation reaction pathways operating in different organisms; routes 1 in fungi, protozoa, and plants to form C9 and C10 side chains and routes 2 and 3 to form C11 extended side chains in marine organisms. In organisms that live in the marine world, sterol methylation patterns vary greatly and reflect the complexity of mixtures of sterols arising through the food chain. The 24-SMT of these organisms catalyze branched and highly alkylated side chains of distinct stereo- and regiochemistry, including tris- or quadruple alkylation of the C11 side chain, not typically observed in the sterol methyltransferase catalyzed products of terrestrial organisms that produce C8 and C9 side chains (Figure 13). Whereas plants can synthesize as many as 60 sterols, marine organisms have been shown to contain as many as 74 sterols in a single organism.(113) Dinoflagellate algae that form the ocean ecosystem are different from other algae and plants in their ability to synthesize 22,23-cyclopropyl sterols, effectively from an ergosterol-containing side chain. Biosynthesis of many other nonconventional phytosterols are postulated to result from nucleophiles for these reactions that originate in Δ22-, Δ23-, Δ24-, and Δ25-olefin-containing side chains. The stereochemistry of the transmethylation reactions and subsequent side chain modifications by reduction have been subjected to detailed investigation.(114) The translational significance for unearthing unconventional sterols, such as the 24-isopropyl steranes, in fossil remains is that their chemical fingerprint provides relevant biomarker information about sterol evolution and its relationship to speciation. 115,116 5 Sterol Enzyme Action 5.1 C24 Methylation In reviewing the action of enzymes that catalyze sterol formation, it has been found convenient to divide topics according to the properties of the enzyme that include its specificity, mechanism, inhibition, and where evidence is available results of mutagenesis experiments. We start with the enzymatic C24-methylation reaction, where progress has been made in identifying the genes and catalytic properties of the corresponding proteins. Four different sterol C24-methyltransferase enzymes (24-SMT) have been detected and classified according to the substrate favored by the enzyme for catalysis. Plants synthesize two 24-SMTs, SMT1, which prefers cycloartenol ((EC 2.1.1.142), and SMT2, which prefers 24(28)-methylene lophenol (EC 2.1.1.143); some fungi and protozoa synthesize 24-SMT1 of substrate preference for zymosterol (EC 2.1.1.410), and other fungi also synthesize 24-SMT with a substrate preference for lanosterol. 111,117−120 In all cases studied, the catalysts shared similar properties of subunit organization as homotetramers of M r 38 000–43 000, as determined by SDS-PAGE and gel permeation, and each contained a single active site for sterol and AdoMet. An interesting feature of several of the cloned 24-SMTs is their ability to produce multiple and distinct product sets at either the C1- or C2-stage of methylation. 111,120,121 So far as is known, 24-SMTs vary in sequence identity (49–77%), the whole family containing only about 63 completely conserved residues (less than 20% of an average of ERG6). Twenty five of these residues lie in the substrate binding segments, and another 38 lie outside of these areas.(122) The cloned enzymes share similar slow turnover numbers of k cat = 0.01 s–1 and pH optima of 7.5. Steady-state kinetic studies show fungal SMT1 (S. cerevisiae) AdoMet and zymosterol bind randomly to the enzyme. Alternatively, the plant SMT1 (G. max) exhibits an ordered binding mechanism, AdoMet binds first, followed by the methyl acceptor cycloartenol. 118,123 The structure of 24-SMT has not been established. In the absence of atomic coordinates, the general characteristics of the 24-SMT active site have been probed by means of photoaffinity labeling using AdoMet, suicide substrates using 26,27-dehydrozymosterol and site-directed mutagenesis experiments. These studies, focused on the S. cerevisiae 24-SMT, have identified several conserved amino acids that cluster in sets of approximately 11 residues in four substrate binding segments of the protein; homology modeling shows that region II contains the Rossmann fold, which consists of central β-sheets flanked by α-helices, and the experimental data shows contact amino acids that can interact with AdoMet and regions I, III, and IV. The sterol binding segments were defined through mutagenesis experiments to give rise to gain-in-function activities that upon mutation can lead to the second C1-tranfer reaction characteristic of plant 24-SMTs. 121,124,125 The 24β-chirality in algal ergosterol and the 24α-chirality of vascular plant campesterol and sitosterol is considered to arise from direct C24-alkylation or through a Δ24(28)-sterol intermediate, respectively. To account for the product diversity, ionic and X– group C24-alkylation mechanisms have been proposed as shown in Figure 14. 125−127 The different stereochemical outcomes observed in ergosterol and sitosterol were proposed to involve a C24-methylation reaction that led to a C24S-methyl group (fungi) or C24R-methyl/ethyl group (plant). In either case, C24-methylation was thought to generate the C25R-H chirality, which is analogous structurally to the stereochemistry at C25 in animal cholesterol. As envisioned in the mechanism for squalene oxide cyclization to sterols, the C24-alkylation mechanism should include a nucleophilic center on the 24-SMT (X–, unidentified amino acid) to form a covalent bond with the C25 or C24 carbocation resulting from the destruction of the Δ24(25)- or Δ24(28)-bonds. The covalent species are shown in mechanism b, and the formation of such intermediates would obviate the necessity to form discrete carbocations. The X– group mechanism considered in 24-alkyl sterol biosynthesis is based entirely on stereochemical considerations that relate to changes in the conformation of covalent-bound sterol-24-SMT of different organisms directing the chirality in enzyme-generated products. Figure 14 Mechanisms for methyl transfer to Δ24(25)-substrate in the action of sterol C24-methyltransferase; dot shows to 13C-labeled carbon. In mechanisms A and C, the carbocationic intermediates are discrete species, and an important difference between the two mechanisms involves the SN2 (Si, β)-attack of methyl cation on the substrate double bond coupled to the regiospecific deprotonation at C28 to yield the 25S stereochemistry typical of 24-alkyl sterols of the ergostane and stigmastane family. In mechanism B, the formation of carbocation intermediates is avoided by postulating the participation of an enzymatic nucelophile-X to form covalent bonds with incipient cations to yield the C25R stereochemistry typical of animal cholesterol. Adapted with kind permission of Springer Science and Business Media from ref (67), Copyright 1977 Springer-Verlag, and from ref (123), Copyright 2003 The American Society of Biochemistry and Molecular Biology. In the ionic mechanisms, illustrated in paths a and c (see Figure 14) the biomethylation reaction involves discrete carbocationic intermediates that can be the result of synchronous changes in bonding that occur in a concerted fashion to form a single product or involve a stepwise process where topology is maintained between the initiation and termination steps affording multiple products. Ammonium and sulfonium analogs of the hypothetical 24-methyl sterol C25/C24 cation intermediates have been found to inhibit the C24-methylation, consistent with a cationic reaction mechanism in sterol alkylations.(128) Equilibrium binding of the transition state analog 25-azalanosterol (charged at physiological pH) to yeast 24-SMT indicated that its K d was similar to that of the sterol acceptor and AdoMet, ca. 4 μM, whereas the K i for 25-azalanosterol against zymosterol (K mapp = 35 μM), determined by steady-state kinetic analysis, is ca. 20 nM. These results suggest that the 24-SMT undergoes a conformational change upon inhibitor binding for the substrate analog to interrupt catalysis. 130,131 Arigoni and colleagues reported the preparation of [1H2H3H]AdoMet (only one enantiomer of the chiral methyl group was present) and have used it to probe the stereochemical course of a methyl transfer to carbon in fungal production of ergosterol. The results obtained indicate that the chirality of the acetate derived from the methyl group at C24 in ergosterol was the same as that of the methyl group of methionine but with a reduction in chiral purity.(132) This addition–elimination reaction proceeds with inversion at the transferring, chiral methyl center yielding an SN2 mechanism for the methyl transfer. The question as to which side of the double bond of the Δ24-sterol acceptor the incoming methyl group is added was resolved by studying the metabolic fate of the pro-Z methyl of various Δ24-13C27-sterols incorporated into ergosterol and comparing the isotopically labeled natural product, found to have the 25S-configuation, to chemically synthesized 25-S-[13C]ergosterol. 133,134 The results show that the first C1-transfer reaction mediated by the plant and fungal 24-SMTs proceeds by alkylation from the Si-face of the substrate Δ24(25) (yielding β-methyl stereochemistry) followed by migration of a hydrogen from C24 to C25 across the Re-face. Yagi et al. investigated the second C1-transfer reaction by feeding a set of [26-13C], [27-13C], and [26,27-13C2]desmosterols (cholesta-5,24-dienol) to cultures of Ajuga reptans and found that the C24-alkylation takes place in a specific manner wherein the C26- and C27-methyl groups of the substrate become C26 (vinyl methyl) and C27 (exomethylene carbon), respectively of the 24β-ethyl Δ25(27)-sterols synthesized by the plant.(135) Regiospecificity directed at Δ25(27)-olefin formation was confirmed using microsomal preparations of Prototheca wickerhamii by the demonstration that 13C-label at C27 (pro-Z methyl) in the lanosterol substrate is retained in the enzyme-generated 24β-methyl Δ25(27)-product.(136) Using cloned Arabidopsis 24-SMT1, we subsequently demonstrated that conversion of [27-13C]fecosterol (C1) to [27-13C]24(28)Z-ethylidenefecosterol (C2) takes place with retention of configuration at C25.(135) Studies of synthesized [28E-2H]- and [28Z-2H]-24-methylene sterol acceptors transmethylated to give 24-ethyl(idene) sterol products established that the second methylation proceeded in such a manner that addition of the methyl group and proton loss occur on opposite faces of the original Δ24(28)-double bond. 123,137 On the basis of these observations and recognizing that 24-SMTs can be bifunctional and capable of operatong successive transfer reactions to yield multiple products, we proposed the C24-methylation reaction illustrated in Figure 15, in which the C28 elimination occurring on the isofucosterol cation during the second C1-transfer reaction occurs by an anti-mechanism.(123) Figure 15 C1-Transfer mechanisms in the conversion of Δ24(28)-substrate into 24-ethyl(idene) products. In each case, the isotopically labeled substrate contains a 13C-27 atom (dot) and a Z- or E-2H-atom selectively introduced at C28. The “cis-process” has been eliminated by experiment. Adapted from ref (123). Copyright 2003 The American Society of Biochemistry and Molecular Biology. Kinetic studies involving 2H-substrates have revealed another important feature of the reaction catalyzed by the soybean 24-SMT enzyme. Thus, incubation of AdoMet versus [methyl- 2H3]AdoMet as substrate against saturating concentrations of cycloartenol afforded a kinetic isotope effect of V CH3 /V CD3 close to unity.(123) Similarly, incubation of [24-2H]cycloartenol failed to produce any apparent isotopes effects related to the 1,2-hydride shift of H24 to C25, consistent with a concerted mechanism in the first C1-transfer reaction. However, substrate binding experiments of [28E-2H]- and [28Z-2H]-24-methylene sterol acceptors to determine isotope effects associated with formation of multiple product resulting from the second C1-transfer reaction demonstrated that [28E-2H]-olefin afforded a KIE ((k H/k D) observed) of 0.92 for the deprotonation step whereas for the [28Z-2H]-olefin a KIE of 1.23 was obtained. The inverse value established by the isotoptically sensitive branching experiments for the second C1-transfer reaction is strong evidence that the proton transfer from the substrate to the active site base has to come to equilibrium prior to the rate-determining transition state and supports the proposal of a discrete carbocation intermediate. We reasoned further that it would be possible to differentiate between the X-group and ionization mechanisms by selectively substituting hydrogen with fluorine or other atoms in the vinylic substrate. By changing the nucleophilicity of the sterol Δ24-bond by substitution of methyl or fluorine for hydrogen significantly decreases the rate at which soybean 24-SMT catalysis proceeds without affect on the binding properties of the substrate appreciably. The observed order of V max/K m values, H (34) 43% identity) and possess a unique stretch, FLFASQDASS, which contains a highly conserved alanine residue considered to be a contact amino acid in the substrate binding recognition site. 238,239 The mechanism of the 22-SD has been studied using MVA labeled with tritium at C2 and C5 and with synthetic substrates incubated with cell-free preparations of 22-SD. 240,243,244 These studies revealed that stereoselectivity of removal of hydrogen from C22 and C23 varies between fungi and plants and protozoa (Figure 23); in fungi two hydrogen atoms are removed from the α-face, whereas in plants and protozoa the two hydrogen atoms are removed from the β-face. In addition, a hydroxylation–dehydration mechanism for the cis-removal of hydrogen atoms, considered based on metabolite profiling,(249) was discounted by Giner and Djerassi who incubated 23R- and 23S-hydroergostenols with the yeast 22-SD and found no evidence of metabolism.(236) The different sterospecificities and substrate preferences in the 22-SD catalyzed reaction among fungi and plants and protozoa appear influenced by specific amino acids whose positions and orientation are responsible for interacting with distinct side chain variants that determine the level of substrate discrimination that occurs with each enzyme. Figure 23 Two mechanisms for the cis-dehydrogenation reaction catalyzed by sterol C22-desaturase (22-SD), also known as CYP51A1 for the fungal enzyme and CYP107A for the plant enzyme, that leads to the introduction of the C22(23)-E-double bond in the sterol side chain; H R and H S refer to hydrogen atoms originally part of the MVA molecule. 6 Concluding Remarks We have seen that the outline for similarity and differences in sterol biosynthesis across kingdoms is now available; these originate in the cloning and characterization of enzymes that form lanosterol and cycloartenol, elucidation of the pathways to cholesterol and 24-alkylation leading to ergosterol and sitosterol and the emergence of cloned enzymes that can be overexpressed affording an exacting analysis of native and mutant protein properties. Yet, the mechanistic details involved in the reaction course of many enzymes that catalyze sterols still need to be worked out. In all the years of pharmaceutical research, sterol biosynthesis has been a major source for new drug discoveries. These studies have contributed to our basic knowledge concerning regression of cholesterol biosynthesis in general. With the increase in fungal resistance and the opportunity to develop novel means for controlling sterol biosynthesis pathways that are absent from humans, the development of new and more potent antifungal and antiparasitic drugs has been one of the impetuses of recent biomedical research. Thus, the search for phyla-specific pathways that contain unconventional reaction sequences to 24-alkyl sterols are currently being pursued in a number of laboratories. In addition, with the recent determination of the X-ray structure of sterol enzymes and the finding that azole drugs used to treat diseases of ergosterol biosynthesizing organisms can bind differentially to human and trypanosome 14α-demethylase enzymes opens the door for a new era of rational drug design.(184) Understanding and treating genetic modifications in sterol biosynthesis that affect human health and researching genetically modified sterol pathways for improvement of specific traits or the addition of new traits to economically important plants is another major worldwide objective. Finally, a word on sterol evolution. Bloch gave us the first insight into the role of oxygen and sterol features affecting sterol competency and biosynthesis, while Ourisson and Rohmer provided a blue print for a rational progression in time for the biosynthesis of terpenes to hopanoids to true sterols, which starts in prebiotic systems. 245,246 However, the recent demonstration of horizontal gene transfer complicates our view for connections to bacterial origins in sterol biosynthesis.(247) There is a correlation between the presence of the nervous system (animals) and the absence of certain biosynthetic steps, notably the absence of sterol C24-alkylation.(29) It is clear that convergent evolution in ergosterol biosynthesis exists since the ergosterol synthesized in Prototheca wickerhamii occurs by the MVA-independent pathway and the cycloartenol route utilizing the Δ25(25)-alkylation path to 24β-methyl group synthesis, whereas the ergosterol formed in fungi occurs by the classic acetate–mevalonate pathway and the lanosterol route utilizing the Δ24(28)-alkylation path to 24β-methyl group synthesis. Equally intriguing to consider are the prospects that sterol biosynthesis and kinetic control of it evolved by a patchwork assembly with the individual members having undergone a duplication event followed by divergent evolution to give rise to the unique specificities. In this regard, we have recently shown that the SMT2 with a distinct product set and substrate specificity likely originates from SMT1 in phytosterol biosynthesis and that a correlation exists between the activation barrier (measured as E a) to sterol C24-methylation and the product outcome; a higher energy route (and therefore more primitive) exists to form Δ25(27)-24β-methyl sterol compared with the formation of Δ24(28)-methylene sterols. 248,249 Although many of the membrane-bound enzymes of sterol biosynthesis continue to be recalcitrant to crystallization and they occur in low abundance in tissues, with tools of protein chemistry, mechanistic enzymology, and structural biology, it is now possible for a renewed sustained attack on the of the individual steps in the pathway to address questions that remain: What about the protein structure that governs product diversity or recognizes one substrate better than another? Are there any allosteric modulators that can affect activity? What changes occurred in enzymes of the oxidosqualene–lanosterol and cycloartenol pathways and of 24-alkyl sterol biosynthesis that fashioned distinct stereospecifity of reactions that are of phylogenetic and functional significance. There is still much to learn about sterol biosynthesis and the enzymes involved in the pathway. A determination of the structure and function of all the sterol catalysts is now within our grasp. The forthcoming decades should be an exciting time for basic bioorganic chemistry and biochemistry regarding sterol biosynthesis and production.