Lessons from natural molecules

Journal name:
Nature
Volume:
432,
Pages:
829–837
Date published:
DOI:
doi:10.1038/nature03194
Published online

Abstract

Natural products have inspired chemists and physicians for millennia. Their rich structural diversity and complexity has prompted synthetic chemists to produce them in the laboratory, often with therapeutic applications in mind, and many drugs used today are natural products or natural-product derivatives. Recent years have seen considerable advances in our understanding of natural-product biosynthesis. Coupled with improvements in approaches for natural-product isolation, characterization and synthesis, these could be opening the door to a new era in the investigation of natural products in academia and industry.

At a glance

Figures

  1. Medically significant natural products and synthetic molecules.
    Figure 1: Medically significant natural products and synthetic molecules.

    a, Natural products. Vancomycin (1), an antibiotic for bacterial infections; staurosporine (2), a lead compound for the development of selective kinase inhibitors for cancer; rapamycin (3), a compound for immunosuppression; Taxol (4), an anti-cancer agent; b, Synthetic molecules. Viagra (5) for erectile dysfunction; Prozac (6) for depression; Lipitor (7) for hypercholesterolaemia; and Gleevec (8) for chronic myelogenous leukaemia. Natural products have strong conformational biases based on stereogenic centres (1, mauve circles), ether and ring fusions (1, yellow ovals), strategically placed substituents to select a single conformation (3, green circles), macrocyclization (3, blue oval), and conjugation (3, yellow oval). Staurosporine's (2) interlocking rings lead to a completely rigid core structure. c, Three-dimensional structural representations of rapamycin and Viagra.

  2. Natural products that exploit reactive functional groups.
    Figure 2: Natural products that exploit reactive functional groups.

    Compounds 916 all illustrate nature's ability to either mask or fine-tune the reactivity of functional groups. a, The enediyne group (red) in calicheamicin (9) and dynemicin (10) is activated to give a diradical intermediate that damages DNA (as shown for dynemicin). b, The carbolamine group in ecteinascidin (light blue; 11) is converted to an iminium ion that reacts with DNA. c, The dithian-1,3-oxide group (dark blue) in leinamycin (12) is activated to form an episulphonium intermediate that alkylates DNA. d, Fumagillin (13) and epoxomicin (14) contain reactive epoxide groups (green) that trap proteases. e, The masked or explicit β-lactones (mauve) in lactacystin (15) and salinosporamide (16), respectively, target the proteasome.

  3. Natural products that exploit shape and polarity complementarity to biological targets.
    Figure 3: Natural products that exploit shape and polarity complementarity to biological targets.

    Biological targets include: G quartets for telomestatin (17); tubulin for discodermolide (18) and hemiasterlin (19); and ion channels for saxitoxin (20) and zetekitoxin AB (21).

  4. Template diversification.
    Figure 4: Template diversification.

    a, The benzodiazepine core is a common template for synthetic diversification because the groups indicated here as R1, R2, R3 and R4 can be varied widely. b, Dysidiolide (23) has been used as a template for a natural-products-based diversity library. The native structure was simplified and a single diversity element was used to create the library (24). c, Compounds 2528 illustrate natural template diversification. Atropine (25) and cocaine (26) are plant alkaloids with mydriatic and local anaesthetic properties, respectively. Epibatidine (27) is a non-opioid analgesic isolated from the skin of an Ecuadoran poison frog, and anatoxin A (28) is the Very Fast Death Factor produced by cyanobacteria.

  5. Biosynthesis of natural products.
    Figure 5: Biosynthesis of natural products.

    The rapamycin synthase assembly line consists of four multimodular proteins (RapA, RapB, RapC and RapP). Fourteen polyketide synthase modules are distributed in RapA–C and the fifteenth, a nonribosomal peptide synthase module (NRPS), comprises the RapP protein. RapA–C comprise the three-subunit assembly-line machinery for the polyketide-chain initiation and elongation. Each of the 15 modules has a carrier-protein domain (peptidyl carrier protein, PCP in RapP). This is post-translationally modified with a phosphopantetheinyl arm containing a terminal cysteine on which the elongating acyl chains are assembled. The most downstream acyl intermediate is shown on the PCP domain of RapP as it undergoes an intramolecular cyclization, thought to be catalysed by the second condensation domain (C) of RapP. The first C domain makes the acyl–N linkage to the pipecolyl moiety of the acyl chain, while the adenylation domain (A) selects, activates and incorporates the pipecolyl moiety. All the atoms of pre-rapamycin come from the four building blocks malonyl CoA, methylmalonyl CoA, pipecolate and dihydroxycyclohexenoate, as shown. After cyclo-release from the assembly line, pre-rapamycin undergoes a series of oxidative and O-methylation-tailoring steps to yield the final product: rapamycin.

  6. The role of oxidation in the construction of natural products.
    Figure 6: The role of oxidation in the construction of natural products.

    a, The oxidative tailoring of vancomycin by three haem-containing proteins introduces aryl ether (C–O) bonds and aromatic (C–C) crosslinks (shown in red) that rigidify the vancomycin skeleton. b, The spectacular series of oxidations that convert taxadiene to Taxol. Eight oxygen atoms are introduced into the scaffold by cytochrome P450 mono-oxygenases, and these are further modified into carbonyl, ether or ester links. The intermediates shown have been identified, but not all the responsible enzymes have been characterized; some of the transformations require more than one enzyme57, 82. c, A key step in the biosynthesis of morphine and other opium alkaloids involves the oxidative coupling of two phenol radicals to form the key bond shown in red.

  7. Recent natural products obtained from nontraditional sources.
    Figure 7: Recent natural products obtained from nontraditional sources.

    a, The nodulisporic acids (29, 30) and b, the guanacastepenes (3136) are from endophytic fungi, the large group of fungi that live inside higher plants; c, halomon (37) is from a red alga; d, pederin (38), which was long believed to be an insect metabolite, is produced by bacteria; e, apratoxin (39) and jamaicamide (40) are from marine cyanobacteria; f, epothilone (41) and tubulysin (42) are from myxobacteria.

Author information

Affiliations

  1. Department of Biological Chemistry and Molecular Pharmacology, Harvard Medical School, Boston, Massachusetts 02115, USA

Competing financial interests

J. Clardy and C.Walsh advise several natural product-based companies including Eisai Research Institute, Novobiotics, Kosan Biosciences and Vicuron Pharmaceuticals.

Corresponding author

Correspondence to:

Author details

Additional data