Natural product chemistry

Published on 02/03/2015 by admin

Last modified 02/03/2015

Print this page

This article have been viewed 2347 times

Chapter 6 Natural product chemistry

This chapter looks briefly at the chemistry of natural products. This is the study of chemicals produced by the many diverse organisms of nature, including plants, microbes (fungi and bacteria), marine organisms and more exotic sources such as frog skins and insects.

Natural products are defined here as organic compounds in the molecular weight range 100–2000. In a broader sense, the term natural products can also be applied to bulk substances from nature, such as crude plant material, foodstuffs, resins and exudates from plants (e.g. myrrh and frankincense) or extracts of plant material (water or alcoholic extracts). The treatment of this subject will be confined to pure single chemical entities (i.e. chemicals with a well-defined structure and purity).

Historically, natural products have formed the basis of medicines and, even now, many of the compounds that are pharmaceutically and medicinally important are derived from natural sources. The reasons for this are complicated, but probably stem from the ability of nature to produce a fantastic array of structurally complex and diverse molecules. A number of theories have been proposed as to why these compounds are produced and although we can only theorize as to why these products occur in nature, it is highly likely that many of them are produced as part of a chemical defence system to protect the producing organism from attack. Examples of this defence include the synthesis of antimicrobial compounds by plants that are infected by bacteria and fungi (compounds known as phytoalexins), or the synthesis of highly toxic natural products in the skins of Central American frogs to deter predation by other animals. Whatever the reasons for the presence of these compounds in nature, they are an invaluable and under-exploited resource that can be used to find new drug molecules.

Natural products in drug discovery

A survey of any pharmacopoeia will show that natural products have a key role as biologically active agents; in fact, it has been estimated that 20–25% of all medicines are derived from such sources. In this definition, the medicinal agent may be a natural product isolated straight from the producing organism (e.g. the β-lactamase inhibitor clavulanic acid isolated from the bacterium Streptomyces clavuligerus), a natural product that has undergone a minor chemical modification (semisynthetic) (e.g. aspirin, derived from salicylic acid, which occurs as esters and glycosides in Salix spp.), or a compound that was totally synthesized based on a particular natural product possessing biological activity (e.g. pethidine, which was based on morphine from the opium poppy, Papaver somniferum). It is sometimes difficult to see how the fully synthetic compound was modelled on the natural product (Fig. 6.1).

Fig. 6.1

Natural products are historically the core of medicines and they are still a major source of drug leads, which is a term used to describe compounds that may be developed into medicines. A particular example of a natural product that is currently one of the best selling drugs is paclitaxel, marketed as Taxol (Fig. 6.2). This drug was developed by BristolMyers Squibb and marketed for the treatment of ovarian and mammary cancers, and became available for use in the USA in 1993. The compound was initially isolated from the bark of the Pacific yew tree, Taxus brevifolia, and demonstrates the best possible qualities of a natural product, being highly functional and chiral. Additionally, paclitaxel occurs in the bark with a wide range of structurally related compounds (taxanes diterpenes); this is a further important and valuable quality of natural products when they are considered as a source in the search for biologically active drug leads. Paclitaxel has many functional groups and chiral centres (11) and these qualities give rise to its distinct shape and fascinating biological activity. It is important not to be overwhelmed at such a complex molecule, but to look at the functional groups that make up the total structure of the compound. Even a natural product that is as structurally complex as paclitaxel can be broken down into the simple chemical features of functional groups and chiral centres.

Fig. 6.2

Interest in natural drugs has long captured the imagination of the general public who has the impression that natural products are safe and non-toxic, but, as we will see, some of the most potent poisons are derived from nature. This view of natural medicines is in part based on the romantic notion of bioprospecting for new drug leads from areas with an exceptionally high level of biodiversity and beauty, such as the Amazonian rain forest. Although public interest in natural drugs is high, investment in the discovery of natural drugs by industry has been highly fashionable and cyclical, due to the development of alternative ways of finding new drug leads (e.g. combinatorial chemistry). An additional perceived benefit of compounds derived from nature is that they are ‘eco-friendly’ and that they may be produced as a renewable resource by growing the plants or by fermenting the micro-organisms that produce them. This approach has both advantages and disadvantages over the synthetic production of biologically active agents, but synthetic chemistry cannot yet readily mimic the ability of organisms to produce such structurally complex and diverse natural product molecules.

There are a number of approaches that can be used to discover new drug leads from nature, and all of the following have been used by large and small pharmaceutical companies in an attempt to harness the biological potential of natural products.

In the ethnobotanical approach, knowledge of the use of a particular plant by an indigenous people is used to direct a search for a drug lead. In this case, observation of a particular usage of a plant, usually made by a highly trained observer (ethnobotanist), allows the collection of that plant and its subsequent testing for biological activity. Examples of such uses include arrow poisons made from trees by South American Indians as a way of hunting animals for food. One of these agents, curare (from the liana Chondrodendron tomentosum) acts as a muscle relaxant, which kills by paralysing the muscles required to breathe. The active component in curare is tubocurarine. This ethnobotanical observation of curare used as an arrow poison led to the development of a muscle relaxant used in surgery known as atracurium. The use of ethnobotanical information is greatly under-exploited and, as many of the more remote regions of the planet become more readily accessible, without trained personnel to interview and retrieve this information, it is highly likely that much valuable local medicinal knowledge will be lost.

In the chemotaxonomic approach, knowledge that a particular group of plants contains a certain class of natural product may be used to predict that taxonomically related plants may contain structurally similar compounds. This approach is highly useful when the chemistry and biological activity of a compound is well described and compounds with similar chemical structure are needed for further biological testing. A good example of this is the plant family Solanaceae, which is a rich source of alkaloids of the tropane type. The knowledge that deadly nightshade (Atropa belladonna) produces hyoscyamine (a smooth muscle relaxant) would enable one to predict that the thorn apple (Datura stramonium) would contain structurally related compounds, and this is certainly the case, with hyoscine being the major constituent of this solanaceous plant (Fig. 6.3).

Fig. 6.3

Using the random approach, plants are collected regardless of any existing previous knowledge of their chemistry or biological activity. This approach relies on the availability of plants that are abundant in a certain area (and may previously have been extensively studied). Plants that are rare or only exist in a specific habitat (e.g. alpine or parasitic plants) may be neglected and access to chemical diversity lost. This approach is purely serendipitous in that there is a chance that random plant selection will give access to extracts (and therefore compounds) with biological activity (bioactivity).

The information-driven approach utilizes a combination of ethnobotanical, chemotaxonomic and random approaches together with a database that contains all of the relevant information concerning a particular plant species. The database is used to prioritize which plants should be extracted and screened for bioactivity. This approach is favoured by large organizations (particularly pharmaceutical companies) interested in screening thousands (in some cases hundreds of thousands) of samples for bioactivity as it may reduce costs by a process known as dereplication – the process of avoiding the repeated discovery of common or known drugs. This is most important where millions of dollars are spent in the natural product drug lead discovery process.

An example of when a database would be of use in the information-driven approach is in the discovery of anti-tumour agents. Should bioactivity be demonstrated upon screening extracts from the English yew (Taxus baccata), then chemotaxonomic knowledge could be entered into the database indicating that this species is related to Taxus brevifolia and consequently may produce related chemical constituents and should be prioritized accordingly. This is certainly the case as Taxus brevifolia produces the antitumour drug paclitaxel.

The discovery of drugs from nature is complex and is depicted schematically in Fig. 6.4. The biomass (plant, microbe, marine organism) is collected, dried and extracted into a suitable organic solvent to give an extract, which is then screened in a bioassay to assess its biological activity (bioactivity). Screening or assessment of biological activity is generally divided into two formats depending on the number of extracts to be assessed. In low-throughput screening (LTS), small numbers of extracts (a single extract up to hundreds of extracts) are dispensed into a format that is compatible with the bioassay (e.g. a microtitre plate, sample tubes). This approach is used widely in academic laboratories where only a relatively low number of extracts are assessed. In high-throughput screening (HTS), thousands of extracts are dispensed into a format (usually microtitre plates with many wells, e.g. 384 wells per plate) and screened in the bioassay. This approach is favoured by the pharmaceutical industry, which may have hundreds of thousands of samples (both natural and synthetic) for biological evaluation. This large-scale approach means that decisions can be made rapidly about the status of an extract, which has an impact on the cost of the discovery process.

Fig. 6.4

Active extracts are fractionated using bioassay-guided isolation, in which chromatographic techniques are used to separate the extract into its individual components; the biological activity is checked at all stages until a pure active compound is obtained. The natural product isolated will be designated as a lead compound and will be assessed for biological activity in a bank of other assays. This process is known as cross-screening and will give information on how selective the compound is – i.e. is it active in all the assays or does it exhibit specificity for one particular assay? This is an important consideration as one of the criteria for the selection of a compound for further development is specificity. Whilst biological evaluation is on-going, structure elucidation will be necessary to determine the three-dimensional structure of the active molecule. This will enable a search to be done to establish whether the compound is novel, what chemical class it belongs to and whether that type of compound has previously been reported to possess biological activity in the bioassay of interest or other bioassays.

Once novelty and potent biological activity have been established, large amounts of the lead compound are isolated and the decision is made as to whether the compound can be synthesized de novo or whether chemical modification needs to be made to enhance the biological activity. The lead compound will undergo extensive in vivo studies to establish activity, toxicity and efficacy; these studies are sometimes known as preclinical studies. Only once all of these steps have been completed will a drug lead finally enter clinical trials, which is the most extensive evaluation stage of a drug candidate during which many drug leads fail through toxicity or lack of efficacy in humans. Successful completion of these trials usually results in a product licence, which means that the compound is now a drug.

Given the complexity of the process described above, it is not surprising that many natural product drug leads fail to make their way onto the market. Some estimates state that only 1 in 10,000 drug leads may actually make their way to the market. The process is also very lengthy and it may take 12–15 years from the collection of the original biomass to the granting of a licence for a new natural product drug. Additionally, the process is very costly but the rewards are enormous. For example, although high costs could occur for the development of Taxol (US$300 million), these can be readily recovered, with initial sales in excess of US$1 billion per annum. In 2009 the best selling brand name drug was Lipitor (Atorvastatin), which made $13.3 billion; this highlights the potential value of the drug discovery process.

The expense, complexity and time of the natural drug lead process have militated against natural products in the past, but the fact remains that natural products are a tried and tested source and there are many examples of natural drugs. The most important strengths of natural products are their complex chemistry and structural diversity.

The polyketides

Polyketide natural products form an immense group of therapeutically important compounds comprising many antibiotics (macrolides and tetracyclines), fatty acids and aromatic compounds (anthrone purgative glycosides and anthracyclic antitumour agents).

Polyketides are mainly acetate (C₂) derived metabolites and occur throughout all organisms (as fatty acids and glycerides), but it is the microbes, predominantly the filamentous bacteria of the genus Streptomyces, that produce structurally diverse types of polyketides, especially as antibiotic substances. The biosynthesis of these compounds begins (Fig. 6.5) with the condensation of one molecule of malonyl-CoA (CoA is short for coenzyme A) with one molecule of acetyl-CoA to form the simple polyketide acetoacetyl-CoA. In this reaction (Claisen reaction), one molecule of CO₂ and one molecule of HSCoA are generated. The reaction occurs because the carbon between both carbonyl groups of malonyl-CoA (the acidic carbon) is nucleophilic and can attack an electropositive (electron-deficient) centre (e.g. the carbon of a carbonyl group).

Fig. 6.5

The curved arrows in Fig. 6.5 indicate the movement of a pair of electrons to form a bond. Further condensation reactions between another molecule of malonyl-CoA and the growing polyketide lead to chain elongation, in which every other carbon in the chain is a carbonyl group. These chains are known as poly-β-keto esters and are the reactive intermediates that form the polyketides. Using these esters, large chains such as fatty acids can be constructed and, in fact, reduction of the carbonyl groups and hydrolysis of the -SHCoA thioester leads to the fatty acid class of compounds. The expanding polyketide chain may be attached as a thioester to either CoA or to a protein called an acyl-carrier protein. Multiple Claisen reactions with additional molecules of malonyl-CoA can generate long-chain fatty acids such as stearic acid and myristic acid.

The poly-β-keto ester can also cyclize to give aromatic natural products, and the way in which the poly-β-keto ester folds determines the type of natural product generated (Fig. 6.6). If the poly-β-keto ester folds as A1, then loss of a proton, followed by an intramolecular Claisen reaction of intermediate A2 (by attack of the acidic carbon on the carbonyl), would result in the formation of a cyclic polyketide enolate A3 which will rearrange to the keto compound with expulsion of the SCoA anion, resulting in the ketone A4. This ketone would readily undergo keto-enol tautomerism to the more favoured aromatic triphenol A5 (phloroacetophenone).

Fig. 6.6

Should the poly-β-keto ester fold as B1, then an aldol reaction on intermediate B2 will occur by attack of the carbonyl by the acidic carbon, and, with the addition of a proton, an alcohol is formed, resulting in intermediate B3. This alcohol can then dehydrate to the conjugated alkene B4, which can also tautomerize and, via hydrolysis of the thioester-SCoA, the aromatic phenolic acid orsellinic acid (B5) is formed.

The reactive nature of poly-β-keto esters gives rise to many useful pharmaceuticals, and, because they are oxygen-rich starting precursors, the final natural products are generally rich in functional group chemistry. Ketone groups are often retained, but reduction to alcohols and the formation of ethers is common and many polyketides, particularly certain antibiotics and antitumour agents, also occur as glycosides.

Fatty acids and glycerides

This group of polyketides is widely distributed and present as part of the general biochemistry of all organisms, particularly as components of cell membranes. They are usually insoluble in water and soluble in organic solvents such as hexane, diethyl ether and chloroform. These natural products are sometimes referred to as fixed oils (liquid) or fats (solid), although these terms are imprecise as both fixed oils and fats contain mixtures of glycerides and free fatty acids and the state of the compound (i.e. liquid or solid) will depend on the temperature as well as the composition. Glycerides are fatty acid esters of propane-1,2,3-triol (glycerol). They are sometimes referred to as saponifiable natural products, meaning that they can be converted into soaps by a strong base (NaOH). The term saponifiable comes from the Latin word sapo meaning ‘soap’. Saponification of fatty acids and glycerides with sodium hydroxide results in the formation of the sodium salts of the fatty acids (Fig. 6.7).

Fig. 6.7

Glycerides can be very complicated mixtures as, unlike the example given in Fig. 6.7, the substituents on the glycerol alcohol may be different from each other, and it is not uncommon for lipophilic plant extracts to contain many types of glycerides.

Fatty acids are very important as formulation agents and vehicles in pharmacy and as components of cosmetics and soaps. Table 6.1 lists the common names, chemical formulae, sources and uses of the more common fatty acids.

Table 6.1 Common fatty acids

The saturated fats are widespread in nature. The three most common (myristic, palmitic and stearic acids) differ in two methylene groups and contain no double bonds.

The unsaturated fatty acids contain a varying number of double bonds. This, together with the length of the carbon chain, is indicated after the name of the fatty acid. For example, oleic acid (18:1), which is widespread in plants and is a major component of olive oil from the olive tree Olea europaea (Oleaceae), has an 18-carbon chain and one double bond. α-Linolenic acid (18:3) is a constituent of linseed oil from Linum usitatissimum (Linaceae) and is used in liniments and as a highly valued additive in oil-based paints. The related acid, γ-linolenic acid (18:3), is found in evening primrose oil from Oenothera biennis (Onagraceae) and is widely used as a dietary supplement. This acid (like α-linolenic acid) is an essential fatty acid and is a precursor to the prostaglandins, which are involved in many biochemical pathways. Evening primrose oil has gained increasing popularity as an aid to alleviating symptoms associated with multiple sclerosis and premenstrual tension. Ricinoleic acid is the main purgative ingredient of castor oil from the seeds of Ricinus communis (Euphorbiaceae), which was used as a domestic purgative but is now used as a source of the oil for the manufacture of soap and as a cream base.

The polyunsaturated fatty acids contain three or more double bonds and are particularly beneficial in the diet as antioxidants. A number of health-food supplements are available as oils or capsules containing fish liver oils from cod and halibut, which are rich in polyunsaturated fats.

Natural oils that are high in fatty acids and glycerides are also used as components of oral formulations and vehicles for injections of pharmaceuticals. Some of the most common oils used in oral preparations include cocoa, olive, almond and coconut oils.

In man, the saturated fats are precursors for the biosynthesis of cholesterol, high serum levels of which are implicated in heart disease through the formation of atherosclerotic plaques in arteries. By reducing fat intake or by consuming foods that are high in unsaturated fats (particularly polyunsaturated fats), the risk of heart disease is reduced.

The tetracyclines

These polyketide-derived natural products are tetracyclic (i.e. have four linear six-membered rings, from which the group was named) and were discovered as part of a screening programme of extracts produced by filamentous bacteria (Actinomycetes), which are common components of soil. The most widely studied group of actinomycetes are species of the genus Streptomyces, which are very adept at producing many types of polyketide natural products of which the antibiotic tetracycline (Fig. 6.8) and the anthracyclic antitumour agents (see Chapter 8) are excellent examples.

Fig. 6.8

The key features of this class of compound are shown in Fig. 6.8. Although tetracycline has numerous functional groups, including a tertiary amine, hydroxyls, an amide, a phenolic hydroxy and keto groups, it is still possible to see that tetracycline is a member of the polyketide class of natural products by looking at the lower portion of the molecule. C₁₀, C₁₁, C₁₂ and C₁ are oxygenated, indicating that the precursor of this compound was a poly-β-keto ester. C₁₀ and C₁₁ and C₁₂ and C₁ form part of a chelating system that is essential for antibiotic activity and may readily chelate metal ions such as calcium, magnesium, iron or aluminium and become inactive. This is one of the reasons why oral formulations of the tetracycline antibiotics are never given with foodstuffs that are high in these ions (e.g. Ca²⁺ in milk) or with antacids, which are high in cations such as Mg²⁺. This group of antibiotics has been long known and they have a very broad spectrum of activity against Gram-positive and Gram-negative bacteria, spirochetes, mycoplasmae, rickettsiae and chlamydiae. Tetracycline comes from mutants of Streptomyces aureofaciens, and the related analogue oxytetracycline from S. rimosus (Fig. 6.9). These antibiotics are widely used as topical formulations for the treatment of acne, and as oral/injection preparations.

Fig. 6.9

Minocycline and doxycycline are produced semisynthetically from natural tetracyclines. Minocycline has a very broad spectrum of activity and has been recommended for the treatment of respiratory and urinary tract infections and as a prophylaxis for meningitis caused by Neisseria meningitides. Doxycycline (Vibramycin) has use in treating chest infections caused by Mycoplasma and Chlamydia and has also been used prophylactically against malaria in regions where there is a high incidence of drug resistance.

Griseofulvin

Another polyketide antibiotic is griseofulvin (Grisovin) from the fungus (mould) Penicillium griseofulvum (Fig. 6.10). This compound was originally isolated by researchers at the London School of Hygiene and Tropical Medicine.

Fig. 6.10

Griseofulvin is a spiro compound; that is, it has two rings that are fused at one carbon. Initially, the compound was used to treat fungal infections in animals and plants, but it is now recommended for the systemic treatment of fungal dermatophytic infections of the skin, hair, nails and feet caused by fungi belonging to the genera Trichophyton, Epidermophyton and Microsporum. Its main use is in veterinary practice for the treatment of ringworm in animals; it is marketed as Fulcin and Grisovin.

Erythromycin A

Erythromycin A is a complex polyketide from Saccharopolyspora erythrea (Actinomycetes), which is a filamentous bacterium, originally classified in the genus Streptomyces. This compound is a member of the natural product class of macrolide antibiotics; these can contain 12 or more carbons in the main ring system. The term macrolide is derived from the fact that erythromycin is a large ring structure (macro) and is also a cyclic ester referred to as an olide (a lactone). As can be seen from Fig. 6.11, erythromycin A has the best features of natural products, being highly chiral (possessing many stereochemical centres) and having many different functional groups, including a sugar, an amino sugar, lactone, ketone and hydroxyl groups.

Fig. 6.11

The therapeutic antibiotic is marketed as a mixture containing predominantly erythromycin A with small amounts of erythromycins B and C. Erythromycin is used to treat Legionnaire’s disease and for patients with respiratory tract infections who are allergic to penicillin. A number of semi-synthetic analogues (including clarithromycin and azithromycin) have been made and there has been interest in the genetic manipulation of Saccharopolyspora erythrea to produce ‘un-natural’ natural products. These compounds could potentially be analogues of the erythromycin class and might also have antibiotic activity.

The statins

A further group of polyketide-derived natural products is the statins, so named for their ability to lower (bring into stasis) the production of cholesterol, high levels of which are a major contributing factor to the development of heart disease. The rationale behind the use of these compounds is as inhibitors of the enzyme hydroxymethylglutaryl-CoA (HMG-CoA) reductase, which catalyses the conversion of HMG-CoA (Fig. 6.12) to mevalonic acid, one of the key intermediates in the biosynthesis of cholesterol. HMG-CoA reductase became a target for the discovery of the natural product inhibitor mevastatin, which was initially isolated from cultures of the fungi Penicillium citrinum and Penicillium brevicompactum (Fig. 6.12).

Fig. 6.12

Following this discovery, the methyl analogue lovastatin was isolated from Monascus ruber and Aspergillus terreus and is also an inhibitor of HMG-CoA reductase. Simvastatin (Zocor) is the dimethyl analogue of mevastatin and all three compounds are prodrugs, being activated by the hydrolysis (ring opening) of the lactone ring to β-hydroxy acids by liver enzymes. These acids are similar in structure to HMG-CoA and are inhibitors of the reductase enzyme. Pravastatin (Lipostat) is semisynthetically produced by microbial hydroxylation of mevastatin by Streptomyces carbophilus. Unlike the previous examples, the lactone ring has been opened to form the β-hydroxy acid which has then been converted into the sodium salt, increasing its hydrophilic water-soluble nature.

Shikimic-acid-derived natural products

Shikimic acid, sometimes referred to as shikimate, is a simple acid precursor for many natural products and aromatic amino acids, including phenylalanine, tyrosine, tryptophan, the simple aromatic acids that are common in nature (e.g. benzoic and gallic acids) and aromatic aldehydes such as vanillin and benzaldehyde that contribute to the pungent smell of many plants (Fig. 6.13).

Fig. 6.13

A number of natural product groups can be constructed from the amino acid phenylalanine, in particular the phenylpropenes, lignans, coumarins and flavonoids, all of which possess a common substructure based on an aromatic 6-carbon ring (C₆ unit) with a 3-carbon chain (C₃ unit) attached to the aromatic ring (Fig. 6.14). Many reactions can occur to this 9-carbon unit, including oxidation, reduction, methylation, cyclization, glycosylation (addition of a sugar) and dimerization, all of which contribute to the value of natural products as a resource of biologically active compounds and enhance the qualities of structural complexity with the presence of chirality and functionality.

Fig. 6.14

Phenylpropenes

The phenylpropenes are the simplest of the shikimic-acid-derived natural products and consist purely of an aromatic ring with an unsaturated 3-carbon chain attached to the ring. They are biosynthesized by the oxidation of phenylalanine by the enzyme phenylalanine ammonia lyase, which through the loss of ammonia results in the formation of cinnamic acid. Cinnamic acid may then undergo a number of elaboration reactions to generate many of the phenylpropenes. For example, in Fig. 6.15, cinnamic acid is reduced to the corresponding aldehyde, cinnamaldehyde, which is the major component of cinnamon oil derived from the bark of Cinnamomum zeylanicum (Lauraceae) and used as a spice and flavouring. Cinnamon has a rich history, being used by the ancient Chinese as a treatment for fever and diarrhoea and by the Egyptians as a fragrant ingredient in embalming mixtures.

Fig. 6.15

Cinnamon leaf also contains eugenol, the major constituent of oil of cloves derived from Syzygium aromaticum (Myrtaceae). Clove oil was used as a dental anaesthetic and antiseptic, both properties of which are due to eugenol, and the oil is still widely used as a short-term relief for dental pain. These phenylpropenes may have many different functional groups (e.g. OCH₃, O-CH₂-O, OH) and the double bond may be in a different position in the C₃ chain (e.g. eugenol versus anethole) (Fig. 6.16). They are common components of spices, have highly aromatic pungent aromas and many are broadly antimicrobial, with activities against yeasts and bacteria. Some members of this class can also cause inflammation.

Fig. 6.16

Myristicin is a component of nutmeg (Myristica fragrans, Myristicaceae) and is thought to be the hallucinogenic component when this spice is ingested in large quantities. This phenylpropene is very lipophilic due to the presence of methylenedioxy and methyl ether substituent groups and it has been proposed that in vivo the double bond of this compound is aminated (an amino group is added), resulting in the formation of an ‘amphetamine-like’ compound. It should be noted, however, that high doses can be fatal and the ingestion of large amounts of nutmeg should be avoided. Safrole, and particularly trans-anethole, are the major components of anise-flavoured essential oils from star anise (Illicium verum, Illiciaceae), aniseed (Pimpinella anisum, Apiaceae) and fennel (Foeniculum vulgare var. vulgare, Apiaceae). These oils are components of popular Mediterranean beverages such as anisette, ouzo and raki. When water is added to these drinks a cloudy white suspension results, which is attributable to a decrease in the solubility of these phenylpropenes as they are more soluble in ethanol than in water.

The phenylpropenes are generally produced by steam distillation of plant material (e.g. cloves) to produce an essential oil, which is normally a complex mixture of phenylpropenes and other volatile natural products such as hemiterpenes, monoterpenes and sesquiterpenes (see below). The steam distillation procedure involves boiling the plant material with water and trapping the vapour in a distillation apparatus. The condensed liquid is transferred to a separating funnel and, as the oils are immiscible with water and form a less dense layer, they can be readily removed.

Lignans

Lignans are low molecular weight polymers formed by the coupling of two phenylpropene units through their C₃ side-chains (Fig. 6.17) and between the aromatic ring and the C₃ chain. A common precursor of lignans is cinnamyl alcohol, which can readily form free radicals and enzymatically dimerize to form aryltetralin-type lignans of which the compounds podophyllotoxin, 4′-demethylpodophyllotoxin and α- and β-peltatin (from Podophyllum peltatum and Podophyllum hexandrum, Berberidaceae) are examples (Fig. 6.17).

Fig. 6.17

This class of compound is common in the plant kingdom, especially in the heartwood and leaves and as major constituents of resinous exudates from roots and bark. The resin obtained from the roots of P. peltatum has long been used as a treatment for warts by North American Indians, and some preparations still exist which contain podophyllin, which is an ethanolic extraction of the resin rich in podophyllotoxin. This lignan is a dimer of two 9-carbon (C₆-C₃) units and is polycyclic (possessing more than one ring). This natural product has a five-membered lactone ring or cyclic ester and there are many examples of these types of lignans possessing this functional group.

Much work has been done on the podophyllotoxin (aryltetralin) class of lignans, the major active cytotoxic principle, podophyllotoxin, being isolated in the 1940s. This compound inhibits the enzyme tubulin polymerase which is needed for the synthesis of tubulin, a protein that is a vital component of cell division (mitosis). Podophyllotoxin is highly toxic and not used clinically for the treatment of cancers, but this class of compounds was an excellent template on which to base the semisynthetic analogues etoposide and teniposide (see Chapter 8).

Coumarins

The coumarins are shikimate-derived metabolites formed when phenylalanine is deaminated and hydroxylated to trans-hydroxycinnamic acid (Fig. 6.18). The double bond of this acid is readily converted to the cis form by light-catalysed isomerization, resulting in the formation of a compound that has phenol and acidic groups in close proximity. These may then react intramolecularly to form a lactone and the basic coumarin nucleus, typified by the compound coumarin itself, which contributes to the smell of newly mown hay.