In Comprehensive Natural Products II Chemistry and Biology

Full text

Provided for non-commercial research and educational use only.
Not for reproduction, distribution or commercial use.
This article was originally published in the Comprehensive Natural Products
II Chemistry and Biology published by Elsevier, and the attached copy is
provided by Elsevier for the author’s benefit and for the benefit of the
author’s institution, for noncommercial research and educational use
including without limitation use in instruction at your institution, sending it
to specific colleagues who you know, and providing a copy to your
institution s administrator.
All other uses, reproduction and distribution, including without limitation
commercial reprints, selling or licensing copies or access, or posting on
open internet sites, your personal or institution’s website or repository, are
prohibited. For exceptions, permission may be sought for such use through
Elsevier’s permissions site at:
http://www.elsevier.com/locate/permissionusematerial
Giovanni Appendino, Gabriele Fontana, and Federica Pollastro. In
Comprehensive Natural Products II Chemistry and Biology;
Mander, L., Lui, H.-W., Eds.; Elsevier: Oxford, 2010; volume 3, pp.205–236

Author's personal copy
3.08 Natural Products Drug Discovery
Giovanni Appendino, Universita` del Piemonte Orientale, Novara, Italy
Gabriele Fontana, Indena S.p.A., Milano, Italy
Federica Pollastro, Universita` del Piemonte Orientale, Novara, Italy
ª2010 Elsevier Ltd. All rights reserved.
3.08.1 Introduction 205
3.08.2 The Current Pharmaceutical Scenario 206
3.08.3 Why Natural Products Are Intrinsically Useful for Drug Discovery 207
3.08.3.1 Molecular Bases for the Biomedical Relevance of Natural Products 208
3.08.3.2 Evolutionary Bases for the Biomedical Relevance of Natural Products 210
3.08.3.3 Structural Bases for the Biomedical Relevance of Natural Products 212
3.08.4 Possible Reasons for the Current Downsizing of Natural Products Drug Discovery 214
3.08.4.1 Intellectual Property Issues 217
3.08.4.2 Access to Natural Chemical Diversity 217
3.08.4.3 The Biodiversity Crisis 218
3.08.4.4 Supply Issues 219
3.08.4.5 Methodological Issues 219
3.08.4.5.1 Entourage effects 220
3.08.4.5.2 False positives/negatives and reproducibility 220
3.08.4.6 Dereplication 221
3.08.4.7 Advent of Combinatorial Chemistry and Progress in Synthetic Chemistry 221
3.08.4.8 Poor Relevance to Noncytocidal Targets 222
3.08.5 Strategies in Natural Products Drug Discovery 223
3.08.5.1 Ethnopharmacology 223
3.08.5.2 Ecology 224
3.08.5.3 Unconventional Natural Products Sources 225
3.08.5.4 Edible Plants 226
3.08.5.5 Derivatization, Diverted Total Synthesis, Diversity-Oriented Synthesis, and
Semisynthesis 227
3.08.5.6 Extract Engineering 229
3.08.5.7 Engineered Biosynthesis (Mutasynthesis, Combinatorial, and Transgenic
Biosynthesis) 229
3.08.6 Conclusions 231
References 232
3.08.1 Introduction
Over the past two decades natural products drug discovery has been increasingly de-emphasized by pharmaceutical
companies. Although heralded on the verge of a comeback several times,
1,2
attention for natural products has so far
failed to materialize in big pharmaceutical companies (pharma), remaining relegated, with a few exceptions, to small
biotech companies. Many review articles have analyzed this issue, pointing out the past success of natural products in
drug discovery and their big and still largely untapped potential to provide new drugs for a host of unmet medical
needs.
3–19
In this chapter, we will attempt to analyze from a pharma perspective why natural products have fallen out
of favor in drug discovery despite their intrinsic utility for biomedical research. After an introduction on the current
state of drug discovery, the reasons for the inbuilt utility of natural products for biomedical research will be
highlighted, attempting next to explain why, despite so many advantages, it is so difficult for mainstream drug
discovery to interface with natural products research. Finally, several strategies to improve natural products drug
discovery and make it more efficient and attractive from a pharma side will be discussed.
205

Author's personal copy
3.08.2 The Current Pharmaceutical Scenario
A byproduct of the late-19th-century chemical business, pharmaceutical research thrived for more than a century
by finding chemical combinations to treat diseases. But, after contributing substantially both to human health and
drug-industry profits, it has failed to produce significant innovations in recent years.
A. Johson, Wall St. J. 12 December 2007
Despite the introduction of a wealth of ingenious and innovative strategies to help and/or direct drug discovery
(combinatorial chemistry, diversity-oriented synthesis, fragment-based drug discovery, chemical biology, in
silico screening),
20
the number of new chemical entities (NCEs) reaching the pharmaceutical market has
suffered a downward trend over most of the past decade. To explain this dismal performance, a host of
arguments have been proposed, from the shortage of low-hanging drug fruits to the suitability of the much-
hyped modern techniques to pick the higher-hanging fruits identified by genomics in the drug orchard.
20
Thus,
aminergic G-protein coupled receptors (CPCRs) are relatively easy, highly druggable targets since their
endogenous ligands are small molecules (serotonin, dopamine, adrenaline, histamine), while interfering with
protein–protein interactions requires larger and more lipophilic agents and this inflation of physical properties
promotes binding to unwanted targets and raises problems regarding absorption, distribution, metabolism, and
excretion (ADME).
20
Depending on the therapeutic area, the capital needed to bring a drug to the market has now skyrocketed to
$0.8–1.7 billion, with cost splitting (breakdown) being approximately 10% for discovery, 15% for preclinical,
15% for manufacturing and process, 55% for clinical trials, and 5% for postmarketing.
21
Since the attrition rate
of clinical development is currently estimated at an appalling 93–96%,
22
a market potential of at least $300
million is considered the lowest limit for a big pharmaceutical company to be interested in a product.
23
This
combined with the lengthening of the R&D time, currently at around 12 years, is responsible for the
‘blockbuster-itis’ that is plaguing the pharmaceutical industry.
Faced with a looming and massive patent cliff in the first half of the next decade and with an arthritic
drug pipeline, drug companies have been increasingly relying for innovation on biological big molecules
(monoclonal antibodies, vaccines, and nucleic acids) and techniques (stem cells). Developments in this area
have been remarkably fast. Thus, the first paper describing RNA interference (RNAi) gene silencing in
mammal cells was published in 2001,
24
but six therapeutic programs based on this concept had already
moved to clinical trials in 2007,
24
and also cutting-edge genetic tools like small inhibitory RNAs (sRNA)
made the leap to drug candidates in record time.
24
Compared to small molecules, biologics enjoy a longer
exclusivity, since clinical trials of bioequivalence are generally needed for generic versions (biosimilars),
25
but have higher costs of development and manufacturing and are therefore more expensive. A recent
Swedish survey estimated that the annual cost of four biological antirheumatic drugs (etanercept, infliximab,
adalimulab, and anakinara) is 50–70-fold higher than the average annual drug cost per person
(E10 800–14 400 per year versus E170 per year),
26
and the financial burden of the biological anticancer
agents is even higher. When generalized, these costs will become unsustainable for any national health
insurance, since the total health costs in Western countries has already reached very high levels (16% of the
gross domestic product (GDP) in the United States in 2005).
27
Despite these limitations, the pharmaceutical
industry has embraced a ‘biopharma’ approach to drug discovery, diversifying its traditional focus on small
molecules with a growing commitment on biologics. Thus, the term ‘new molecular entities’ is rapidly
replacing the term NCEs in new drugs inventories. Unsurprisingly, drug companies have also started to shed
their chemical workforce at an unprecedented rate. The laying off of Robert Sliskovic, the discoverer of
atorvastatin (Lipitor, 1), made headline news and was considered emblematic of the pharmaceutical
industry’s declining fortunes.
28
Atorvastatin is the best-selling pharmaceutical product ever, having gener-
ated over $80 billion sales to Pfizer since its introduction into the clinic, and the laying off of Sliskovic was
commented in a front page article by The Wall Street Journal.
28
The list of major drug companies that
announced programs of cost reduction and biotech refocusing in 2007 includes, apart from Pfizer,
Bristol–Myers Squibb (BMS), Novartis, and Astra-Zeneca, with Merck and Lilly already adopting this
model a few years earlier.
29
206 Natural Products Drug Discovery

Author's personal copy
With a mean time of approximately 12 years from drug discovery to launch, the current problems of
pharmaceutical research have their roots in choices dating to the 1990s and any strategy pursued today will
have only marginal influence on the near-future late-stage development pipelines. The current downsizing of
natural products drug discovery should therefore be seen as a part of a general trend of pharma research to focus
on biological drugs at the expenses of small molecules. Confronted with an acute productivity crisis, main-
stream pharmaceutical research will undoubtedly explore alternative strategies of discovery, with ‘old’ assets
like natural products possibly coming of age again. Newer screening paradigms, shortened discovery times, and
integration into a multifaceted drug discovery scenario will, however, be necessary to foster the long-awaited
renaissance of natural products drug discovery.
3.08.3 Why Natural Products Are Intrinsically Useful for Drug Discovery
When you have no idea where to begin in a drug discovery program, Nature is a good starting point.
L. H. Caporale, Chem. Eng. News 13 October 2003, p 89
A large number of biological processes involve the interaction of a small molecule (ligand) with a macromolecular
target (receptor). Preeminent examples are the interaction of neurotransmitters with their protein targets, of
intercalating agents with specific DNA or RNA polynucleotidic sequences, and of macromolecular carriers with
the ligands they transport across biological membranes. In general, the interaction of a small molecule with a
macromolecular target represents an opportunity to perturb a biological system, to study its function, and to assess
its druggability, that is, its amenability to pharmaceutical exploitation.
30
In molecular terms, perturbing is equal to
knowing and natural products hold a special position as molecular perturbators, since their role to reveal
interesting biology, to provide the tools necessary for its study, and to generate molecular clues (hits and leads)
for its ultimate therapeutic exploitation is at the very base of modern medicine. Tubulin, one of the most
important anticancer targets, is a remarkable example of this process.
31
Thus, tubulin was discovered because of
the availability of colchicine (2), a specific ligand obtained from the medicinal plant Colchicum autumnale L. The
biological profile of tubulin was next furthered using a variety of natural products, including podophyllotoxin (3),
and the manipulation of tubulin was eventually translated chemically with the development of Vinca alkaloids,
taxanes, and epothilones into effective anticancer drugs. Natural products were also instrumental for the
identification of hsp90 as an anticancer target and to study its function.
32
Several natural products drugs aimed
at the inhibition of this chaperone are currently under clinical development, making it possible that, thanks to
natural products, hsp90 biology will also soon be translated into hsp90 drugs
33
(for more details, see Chapter 3.06).
Natural Products Drug Discovery 207

Author's personal copy
Natural products have a personality: they come with attractive names, awe-inspiring structures and dazzling
biological properties, and there are clear molecular, evolutionary, and structural bases for their relevance in
drug discovery. Because of their three-dimensional structural complexity and inbuilt affinity for biological
surfaces, natural products are in fact privileged structures for drug discovery from both a chemical and a
biological standpoint and qualify nature not only as the ultimate synthetic chemist but also as the ultimate
pharmacist.
3.08.3.1 Molecular Bases for the Biomedical Relevance of Natural Products
Macromolecular biological targets (proteins, nucleic acids) ‘see’ small molecules as three-dimensional surfaces
bearing specific binding elements (charges, polarities, and hydrogen-bonding donor and acceptor elements).
Surreptitious complementarity between biogenetic enzymes and animal protein domains is the general mole-
cular basis for the preeminence of natural products in biomedical research and drug discovery.
34
Most natural
products are the results of enzymatic reactions and their shape must therefore be complementary to that of the
biological protein surface that fitted their ultimate precursor(s). A biosynthetic origin therefore involves the
imprinted capacity to recognize protein surfaces. All organisms, whether bacteria, plants, or animals, are the
results of permutations and combinations of the same four basic nucleotides and 22 amino acids. The number of
unmodified, or ‘naked’ proteins coded by the human genome is between 80 000 and 100000 but the total number
of different proteins produced by human cells is magnified by posttranslational modifications like glycosylation,
acetylation, methylation, hydroxylation, and phosphorylation and is probably around 1 million.
35
Assuming a
10% of druggability,
36
the number of drug targets is therefore enormous, around 100 000, but only 324 of them
have so far been identified for all the approved therapeutic drugs.
37
Despite the large number and structural
variety of these protein targets, the number of peptide domains, which is the way secondary structures like -
helices and -sheets can arrange themselves in space independently from the rest of the structure, is fairly
limited. It has been estimated that the widely diverse function of proteins derives from the combination of only
600–8000 domains. The number of distinct protein families is therefore fairly limited
38
since a similar structural
domain can be used by many proteins in more or less modified forms generated by divergent evolution, as
shown, for instance, by leukotriene A4 hydrolase and thermolysin. These two enzymes catalyze different
reactions (vinylogous opening of an epoxide ring and peptide hydrolysis) but share a very similar fold and
profile of inhibitors.
38
Therefore, natural products are characterized by an intrinsic, biogenetically imprinted
shape complementary to biological surfaces and this topological property can be translated into a reversible
interaction with a druggable target totally unrelated to their original biosynthetic enzymes. In other words, since
protein domains are conserved, biological surfaces homologues to those present in biogenetic enzymes might
exists in animals, even though they fundamentally lack the enzyme machinery to produce secondary metabolites.
For instance, the flavonoid biogenetic enzymes chalcone isomerase, chalcone synthase, and anthocyanidin
synthase share a flavonoid recognition site with mammalian protein kinases and it is therefore unsurprising
that the flavonoid genistein (4) was one of the first inhibitors of tyrosine kinases to be discovered.
39
The
biosynthetic enzymes/drug targets correlation makes it possible that small-molecule ligands might exist for all
the druggable human targets and neuroactive plant alkaloids represent a striking example of this assumption.
Cocaine (5) inhibits the reuptake of mammalian biogenic amine neurotransmitters (dopamine, serotonin,
noradrenaline, adrenaline), a system that has no counterpart in plants, which, in turn, also contain a host of
compounds that target the cholinergic, adrenergic, dopaminergic, and gamma-aminobutyric acid (GABA)ergic
systems of animals, all structures that do not fundamentally exist, even in cognate form, in plants.
34
208 Natural Products Drug Discovery

Author's personal copy
Some natural products are not the direct result of an enzyme reaction but rather of a spontaneous reaction
cascade triggered by an enzyme reaction. In this so-called bomb-like strategy, a stable precursor is enzyma-
tically transformed into an unstable species (chemical bomb) that undergoes an intramolecular rearrangement
to a more stable but still reactive compound, capable of interacting with a variety of targets rather than with a
single target. Examples from plant secondary metabolites are isothiocyanates from cruciferous plants and
thiosulfinates from garlic, formed from glucosinolates and sulfoxidic amino acids by the action of specific
enzymes (myrosinase and alliinase, respectively) compartmentalized in different cells or cellular stores with
respect to their substrates.
40
Hydrolysis of the thioglucosidic bond of glucosinolates generates the unstable
sulfate of an N-hydroxythioimidate, which then undergoes Lossen rearrangement to a reactive isothiocya-
nate,
40
while allinase triggers a -elimination reaction that splits alliin into pyruvic acid and a sulfinic acid,
which then spontaneously dimerizes to allicin, a reactive thiosulfinate ester (Scheme 1).
41
Compounds like
isothiocyanates and thiosulfinates target nucleophilic sulfhydryl sites of a host of proteins via a polar trapping
mechanism, while the microbial enediyne antibiotics cleave DNA with a radical mechanism, mediated by the
formation of a 1,4-diyne (p-benzine) via a Bergman reaction.
42
Reactive secondary metabolites have generally been perceived as nonselective and indiscriminate in their
activity and therefore of little relevance for drug discovery. Lists of chemically or biologically reactive groups
that should be avoided in drug discovery (aldehydes, exomethylene enones, epoxides, furans, etc.) can be found
in any book on drug discovery. Nevertheless, reactive functional groups do occur in natural products and are,
surprisingly, often associated with a specific activity. Thus, the exomethylene--lactone parthenolide (6), an
NF-B inhibitor and the hallmark constituent of feverfew (Tanacetum parthenium L.) has been considered for
clinical development for a variety of conditions, including headache, but has also raised considerable interest as
a selective cytotoxic agent for stem cells,
43
while the even more reactive nitrogen mustard CpdA (Compound
A, 7) is a selective glucocorticoid receptor modifier, capable of dissecting the transactivation and the transre-
pression properties of these hormones.
44
CpdA is as effective as dexamethasone in counteracting acute
inflammation in vivo but lacks the hyperglycemic side effects of glucocorticoids and, despite its reactive nature,
is considered of ‘great potential for therapeutic use’.
45
CpdA is a stable (sic) analogue of the unstable active
principle of the Namibian shrub Salsola tuberculatiformis Botsch., a plant used in the bushman folklore as a
contraceptive, and implicated in a syndrome of prolonged gestation and fetal postmaturity in sheep.
44
The
active principle of this plant is a very unstable compound that quickly decomposes to synephrine (8) in acidic
water and that has been suggested to be a precursor of the phenolic aziridine (9). CpdA was prepared from
synephrine by treatment with acetyl chloride and its biological profile was comparable to that of the elusive
active principle of the plant.
44
Although phenolic aziridines are extremely unstable in vitro, the active principle
of S. tuberculatiformis and its derived aziridine are apparently stabilized in vivo by binding to plasma proteins.
44,46
Scheme 1 Enzymatic ‘detonation’ of glucosinolate and allicin chemical bombs.
Natural Products Drug Discovery 209

Author's personal copy
There is also growing evidence that reactive Michael acceptors can interact with biomolecules in a
reversible fashion, not remaining bound to their ‘proximate’ targets but bouncing within a multitude of thiol-
containing proteins and eventually ‘landing’ on key regulatory proteins having thiol group especially sensitive
to electrophiles. This mechanism is the basis of the activity of CDDO (10), a semisynthetic homotriterpenoid
nitrile considered one of the most promising anticancer agents in current clinical development.
47,48
Compounds
like CDDO do not fit into the conventional models of drug action based on a single high-affinity receptor–
ligand interaction, since their activity involves transient molecular interactions with multiple targets that share
the critical and reactive cysteine residues.
49
If Michael acceptors that react reversibly with thiol groups have
different affinity for different target proteins, then low concentrations of these compounds would preferentially
interact with the most sensitive targets. A host of networks, in this case those responsible to the redox state of a
cell, are therefore affected by their activity. This mechanism might well apply to a series of multifunctional
chemopreventive agents like sulforafane (11), curcumin (12), and the organosulfur compounds from garlic,
49
all
compounds capable of targeting multiple proteins through Michael adduct mechanisms.
50
Compounds like
CDDO undoubtedly stretch the concept of ligand–receptor interaction well beyond the classic lock-and-key
metaphor and it should be remarked that several tyrosine kinase inhibitors also interact in a combinatorial way
with a host of their enzyme substrates, modifying the whole ‘kinomic’ profile of a cell rather than a single
effector target and thus modulating the activity of several signaling networks.
51
3.08.3.2 Evolutionary Bases for the Biomedical Relevance of Natural Products
The term ‘secondary’ in secondary metabolites does not imply a ‘less important’ status but simply addresses a
different functional level, just like the ‘secondary’ structure of a protein is equally important as the ‘primary’ or
the ‘tertiary’ structure. Indeed, natural products have been detected in fossils
52,53
and presumably appeared
very early in the evolution of life, possibly already in the RNA world, since many antibiotics bind specific RNA
sequences.
52,53
Evolutionarily, the synthesis of natural products is a conserved and highly successful trait. Since
it is not uncommon that scores of specific enzymes are involved in the biosynthesis of a single natural product,
the production of secondary metabolites involves an enormous investment in terms of protein-coding DNA.
Given this investment, natural products must have beneficial functions for their producers, all based on the
interaction with macromolecular targets and their functional perturbation in terms of promotion or inhibition.
The activity of natural products has been shaped to overcome environmental stress and provide defense against
natural enemies and an adverse environment and was optimized by evolution during millions of years of
environmental high throughput screening (HTS). In nature, the struggle for survival is waged at the molecular
level and natural products, by reflecting eons of wisdom and refinement, have an intrinsic evolutionary and
ecological meaning. They represent working examples of biological ‘intelligent design’ and it is therefore
regrettable that the futile cat-and-mouse contest between science and creationism has not been waged at a
210 Natural Products Drug Discovery

Author's personal copy
molecular level, spurring research into the elusive meaning of secondary metabolites. The ecological role of
secondary metabolites is extremely difficult to investigate. They are like words of a language that we can read
but do not understand, just like Etruscan. An important exception is capsaicin (13), probably the only secondary
metabolite whose natural function has been elucidated in detail.
54
After survival and reproduction, dispersal of
seeds is third on the list of priorities for a plant. To this purpose, peppers produce fleshy and colored fruits to
attract consumers and colonize new areas. By using nonpungent peppers consumed by both mammals and birds,
it was found that fruit ingestion by mammals inhibits seed germination, while, conversely, consumption by birds
does not damage seed germinability, rather promotes it. Birds swallow the fruits and promote the dispersion of
seeds while mammals chew the fruits with their teeth and physically damage them. Hence, mammals behave as
seed predators and birds as seed dispersers, acting as living ‘vessels’ to carry chillies to new turfs. Capsaicin
targets the vanilloid receptor TRPV1, a molecular thermometer whose activation is perceived as pungent pain
by mammals but not by birds, which have a mutated form of TRPV1, insensitive to capsaicin.
55
Owing to the
selective sensitivity of the mammalian version of TRPV1 to capsaicin, this compound functions therefore as a
specific inhibitor of seed predation.
There are strict relationships between sensitivity to natural products and genomics, with the possibility of
adaptive coevolution between the production of secondary metabolites and their target. For instance, it has
been demonstrated that camptothecin (CPT, 14)-producing plants like Camptotheca acuminata,Ophiorrhiza
pumila, and Ophiorrhiza liukiuensis have Topo1’s with point mutations that confer resistance to CPT.
56
Remarkably, one of these substitutions (Asn722Ser, human Topo1 numbering) is identical to that found in
CPT-resistant human cancer cells while a phylogenetic analysis of Topo1’s in CPT-producing and
CPT-nonproducing plants suggests that mutations in Topo1 occurred before the CPT-producing enzymatic
machinery appeared.
56
In other words, CPT was ‘planned’ taking advantage of a previous biochemical
enzymatic unicity. Since the evolutionary history of most natural products is unknown, we lack information
on the molecular milieu and the temporal range in which their production evolved. In this context, a theory
alternative to surreptitious complementary has been developed to explain why animals have receptors for
compounds produced by plants and, conversely, plants produce compounds for receptors they lack. According
to the vestigial receptor hypothesis,
34
receptor–ligand pairs evolved in primitive organisms that predated the
divergence of animals and plants. Since then, organisms that then evolved into plants lost the need of receptors
and maintained that of ligands, while those that evolved into animals retained receptors but lost the need for
their ligands. In this context, it is instructive to ponder that many compounds involved in cell signaling in plant
cells have receptors also in animals, such as abscisic acid (15),
57
salicylates,
58
and genistein (4).
59
While being
necessarily vague and simplistic, evolutionary observations and hypotheses on the function of natural products
can nevertheless have interesting implications for drug discovery, as shown by the adaptation strategies
developed by plants and microorganisms to overcome self-poisoning from their own metabolic products.
Natural Products Drug Discovery 211

Author's personal copy
As already discussed with CPT, organisms accumulating inhibitors of basic biological processes like mitotic
poisons (taxanes, Vinca alkaloids, colchicine) and antibiotics, must also have evolved strategies to avoid self-
poisoning. These involve, apart from compartmentalization, the expression of export pumps, of antibiotic-
modifying enzymes, and of target-protecting mechanisms that can mimic the molecular bases of drug resistance
in the clinical setting.
60
Thus, antibiotics are produced to create, by disrupting basic biological processes, a local
protective environment inhospitable to invading organisms. Clearly, antibiotic-producing microorganisms
must have evolved self-resistance mechanisms prior to the production of antibiotics and, indeed, horizontal
gene transfer from nonpathogenic bacteria, the major source of antibiotic resistance, has been ultimately traced
to the antibiotic-producing organisms themselves.
60,61
In other words, acquired antibiotic resistance among
dangerous bacterial pathogens, an increasing medical problem, relies on mechanisms primed at the stage of
antibiotic production. Just like with CPT-producing plants, also in antibiotic-producing Actinomycetes, protec-
tion must have evolved before the production of active metabolites. A possible mechanism has been dubbed
‘feed-forward’ and involves inactive precursors that act as signal to prepare the organism for the later buildup of
toxic levels of antibiotics.
61
More sophisticated mechanisms of self-resistance to ‘endogenous’ secondary
metabolites have also been discovered and they might afford interesting mechanistic clues to the development
of clinical resistance to them. The enediyne anticancer antibiotic calicheamicin (16) is an interesting case. This
agent is a real ‘chemical nuclease’ that fragments DNA via a cycloaromatization-induced radical mechanism.
42
Because of its extraordinary toxicity, in the femtomolar range for some organisms, calicheamicin is used in the
clinics only as a monoclonal antibody conjugate (Mylotarg) and the mechanism of self-resistance to this toxin
has long remained a mystery. Enediyne-binding proteins (chromoproteins) stabilizing these highly unstable
compounds and possibly aiding in self-protection have been detected in some ene–dyne-producing micro-
organisms but not in Micromonospora calicheamicensis, the source of calicheamicins. Recently, a gene (calC)
conferring in vivo resistance to calicheamicin has been characterized in the genome of M. calicheamicensis.
62
The encoded CalC protein turned out to be a stoichiometric self-sacrificing agent against calicheamicin-
induced double strand DNA scission, capable of quenching activated calicheamicin through a direct and
specific hydrogen abstraction that mimics the action of the antibiotic on DNA. The production of CalC,
along with a tightly sequestered biosynthesis and exportation system, makes it possible to escape self-toxicity
from this agent and similar self-sacrificing strategies might underlie the development of resistance to enediyne
also in nonproducing organisms, including cancer cells.
The examples we have described highlight how closely related genomics, metabolomics, and ecology are
and how information from this fertile interface can have far-reaching clinical implications. Indeed, the
evolutionary study of natural products can afford interesting clues for drug discovery in terms of both
mechanism of activity and resistance to it.
3.08.3.3 Structural Bases for the Biomedical Relevance of Natural Products
The chemical complexity and diversity of natural products supercedes anything pharmaceutical companies or
chemists can design, making it possible to explore areas of the biological space that are difficult, or downright
212 Natural Products Drug Discovery

Author's personal copy
impossible, to access with compounds obtained by random synthesis campaigns. Thus, a statistical investigation
of the structural complementarity of natural and synthetic compounds showed that 40% of the chemical
scaffolds of natural products are absent in synthetic compounds.
63
Organic synthesis has evolved rapidly and
nowadays, given enough students and funds, most natural products are within the reach of synthetic preparative
chemistry. However, simple statistic considerations make it very unlikely that the constitutional and config-
urational structural subtleties of natural products might evolve ex novo from a random synthetic program. The
chemical space is in fact huge but fundamentally void in biological terms. Most small biomolecules are made
only by four elements (carbon, hydrogen, nitrogen, and oxygen) and have molecular weight (MW) <500 Da,
while biological peptides are based on only 20 building blocks and have an average of 300 amino acids. Even
with these structural constraints, the number of possible structural choices (permutations in the statistic lingo) is
appalling. Thus, the number of small molecules having MW <500 Da and based on the four most common
biological elements (C, H, N, and O) is 10
60
, while the number of 300 amino acid-long linear sequences based
on the 20 proteogenic amino acids is 10
360
.
63
Life has therefore been extremely selective in molecular terms,
since less than 18 000 natural products are known and, as we have seen, the human genome contains sequences
of less than 100 000 proteins. Most synthetic chemical compounds are therefore biological chuff and, without
opportune clues, it is extremely difficult to enter the highly selective biological space of natural products with
compounds derived from a random synthetic campaign. An apparent exception is the anticancer drug
5-fluorouracil (5FU, 17), which was synthesized over 40 years before being actually described as a natural
product.
64
However, its synthesis was inspired by the RNA base uracil and was not the result of a program of
random synthesis.
65
Considerations of this type, better than the much-hyped similarity between the drug space
and the natural products space, support the view that natural products are special tickets in the drug discovery
lottery. Since natural products have been the major source of chemical diversity for drug discovery, the drug
space was filled mostly by research programs based on endogenous small molecules (neurotransmitters,
hormones) or natural products and therefore this convergence might, in principle, be viewed as the mere
result of past strategies of drug discoveries.
Synthetic libraries are straightforward to assemble but the relatively limited number of synthetic reactions
and building blocks amenable to combinatorial strategies means that the resulting compounds often lack the
structural complexity and diversity required to efficiently explore the biological space. For instance, rigid
molecules occur rarely in combinatorial libraries since they are more difficult to assemble and the higher
number of rotatable bonds means that conformational constraint, often an affinity booster, is missing from these
libraries.
63
Indeed, combinatorial libraries are mainly based on flat structures. These, while very useful to
explore ATP mimics and kinase ligands, are much less suitable to discover agents capable of interfering with
complex processes like protein–protein interactions. Finally, synthetic libraries are often biased by the
requirements of previous focused drug discovery programs and/or by the compliance to certain predefined
criteria like the Lipinsky rule of 5 (RO5)
63
and the absence of reactive functional groups like epoxides, furans,
and -unsubstituted enones. Many bioactive natural products violate RO5 and feature ‘undesirable’ reactive
groups. Thus, the alkylation of an epoxide moiety is essential for the bioactivity of important drug leads, such as
the antiangiogenic methionine aminopeptidase2 (MetAP-2) inhibitor fumagillin (18),
66
the histone deacetylase
(HDAC) inhibitor trapoxin (19),
67
and the exquisitely selective proteasome inhibitor epoxomicin (20),
68
while
the anti-inflammatory agent triptolide (21), a putative transient receptor potential canonical (TRPC) ligand,
69
features a lineup of three adjacent epoxide groups. Other reactive, more exotic, and still biologically critical
functional groups can occur in natural products drug leads, as exemplified by the hydroxamic HDAC inhibitor
trichostatin A (22),
70
the -lactone proteasome inhibitor lactacystin (23),
71
and the diazo derivative kinamycin
(24).
72
Given a suitable molecular framework, reactive functional groups can indeed be implanted in drug leads
Natural Products Drug Discovery 213

Author's personal copy
but this operation is forbidden under current ‘rational’ drug discovery rules, which, by doing so, precludes the
exploration of a significant portion of the biological space. Finally, structural complexity aids specificity and
potency in biological interactions, limiting target promiscuity,
73
while diversity is important to broaden the
chemical space ‘interrogated’ during the biological evaluation.
A host of molecular, evolutionary, and structural reasons therefore underlie the idea that natural products, as
evolutionarily selected ligands to structurally conserved but genetically mobile protein domains, represent the
most validated starting point to explore the druggable section of the chemical space. This view, while amply
shared within the drug discovery environment, is nevertheless difficult to explicitly translate into numbers
since structural diversity is hard to assess objectively with measurable parameters.
74
Even using simple
structural elements, natural products are clearly differentiable from combinatorial chemistry products in
terms of MW (414 versus 393), number of stereogenic centers (6.2 versus 0.4), and cycles (4.1 versus 3.2) and
they incorporate fewer nitrogen, halogen, and sulfur atoms but are more rich in terms of oxygen and are
sterically more complex, with more rings and bridgehead carbon atoms.
74
Based on these observations, a
topological analysis of combinatorial libraries and natural products showed that combinatorial compounds
densely populate a small area of the chemical space while natural products are more largely distributed in terms
of occupancy of chemical space and are more diverse.
74
3.08.4 Possible Reasons for the Current Downsizing of Natural Products Drug
Discovery
The pharmaceutical industry has a conception of the format through which future discoveries will be made, and
natural products are not on their radar. The mavens (sic) of the pharmaceutical industry seem to think that a
discovery made outside that format can’t be worth much. Some of these guys would have turned down the
214 Natural Products Drug Discovery

Author's personal copy
Gettysburg Address because it was handwritten by an aging single author rather than turned out by some pricey
word-manufacturing institute that hit upon it by chance.
S. Danishefsky, Chem. Eng. News 13 October 2003, p 107
The exquisite biological specificity of natural products has laid the foundations of modern medicine. Natural
products save millions of lives every year and generate annual commercial sales of billions of dollars to their
developers and discoverers. Undoubtedly, natural products represent the most successful and validated strategy
of drug discovery and it has been calculated that 60% of the current drugs are natural products, derivatives of
natural products, or synthetic analogues of natural products.
75
Furthermore, high-speed approaches of fractio-
nation and structure elucidation based on hyphenated techniques
76
have considerably expedited the
construction of natural products libraries, alleviating the burden of dereplication.
77
Interestingly, some of
these techniques, such as automated high-performance liquid chromatography (HPLC) separation with fast
gradient elution, were originally developed for the production of combinatorial libraries.
78,79
Overall, the
processing time of crude extracts has been substantially reduced through pretreatment, automated separation,
and computer-assisted structure elucidation. In particular, iterative automated fractionation makes it possible to
detect minor compounds once below the threshold of chemical and biological revelation and therefore
inaccessible.
78,79
The structural complexity and the diversity of secondary metabolites are also ideal to fill
the gap between the growing number of drug targets disclosed by genomics and the dismally low number of
specific ligands available for them.
Nature is undoubtedly the largest and most diverse combinatorial library available but unlocking it is far from
trivial since it requires multidisciplinary expertise, is more time consuming and costly than most current drug
discovery approaches, and poses problems unfamiliar to corporate culture. It is therefore hardly surprising that,
over the past two decades, drug discovery has gradually, and probably myopically, prematurely dismissed research
into natural products as an old-fashioned delivering tool, investing instead in nonvalidated surrogates of biodi-
versity like combinatorial chemistry. The profitability of natural diversity to provide templates for drug discovery
has been questioned and bioprospecting has lost out to high-throughput drug discovery, a process that relies on
combinatorial chemistry and computational drug discovery. Thus, in 2008, Merck cut its natural products program
entirely, despite a long and successful history in this area (lovastatin, caspofungin acetate) and, within the major
pharmaceutical companies, only Novartis and Wyeth retain natural products research divisions with activities that
go beyond the semisynthesis of antibiotics and steroids. Remarkably, less than a decade ago, it was still possible to
claim that ‘‘most major pharmaceutical companies maintain important efforts in natural products’ research.’’
80
The
demise of natural products in drug discovery is even more surprising when one considers that, within the 15 small-
molecule drugs approved by Food and Drug Administration (FDA) in 2007, six were natural products or
derivatives of natural products.
81
This ratio is somewhat skewed by the inclusion, within the count of NCEs, of
three ‘old’ drugs (azithromycin, topotecan, and temsirolimus) for which new indications were approved but is
nevertheless remarkable. Clearly, there must be solid reasons for the pharma industry to abandon the beaten track
of natural products and adventure into remote areas of the chemical space to discover new drugs.
On the other hand, natural products isolation is also becoming a rare area of interest in the US and European
universities. Natural products have long been pursued in academia only because of their unique chemical
structure but, over the past decades, the focus has shifted on the discovery of their properties and the academic
demise of natural products isolation is therefore unjustified. The academic woes of Georg Pettit, a hero of
natural products drug discovery, exemplify this trend.
82
Just like taxonomy, natural products are also becoming
extinct in academia, serving mostly as models for total synthesis. The shift in focus from isolation to synthesis is
ironical at an age when more and more emphasis is placed on applied research since synthesis is often pursued
essentially as a training ground for Ph.D. students and lacks practical application.
The transition from natural products to synthetic drugs was preceded by that from pharmacognosy to
medicinal chemistry. Indeed, by the 1940s, heroic plant drugs like opium, cinchona, ipecacuhana, and mayapple
had already been replaced by their isolated active constituents (morphine, quinine, emetine, and podophyllo-
toxin). Since the activity of most medicinal plants, even as popular as valerian and chamomile, could not be
traced to a single constituent amenable to pharmaceutical development, their gradual downsizing to the health
food realm was inexorable, while, in the wake of the wartime efforts to produce penicillin, large fermentation
programs of drug discovery were launched, filling up the dwindling pipeline of plant natural products drugs.
Natural Products Drug Discovery 215

Author's personal copy
Despite several decades of success, in the past two decades microbial natural products research has also been
de-emphasized in pharmaceutical companies and considered as overexploited and poised to fail to significantly
enrich the yield of pharmaceuticals. Natural products are no longer considered a focal point for hit discovery,
making up for only a tiny percentage of the compound archives of large pharmaceutical companies, typically
comprising up to 5 million compounds. Alternative opportunities to discover novel low MW leads have taken
over and drug discovery is nowadays associated with dazzling new concepts and technologies, such as
functional genomics, combinatorial synthesis, structure-based ligand design, and ultrahigh-throughput screen-
ing (UHTS, >100 000 compounds per week).
20
Conversely, and despite noteworthy methodological evolution,
the strategy to develop new natural products drugs has instead remained the same, involving primary screening
of crude extracts, bioassay-guided fractionation, dereplication of active compounds, and isolation and structure
elucidation of new bioactive constituents. Furthermore, despite the current success of natural products drugs,
there is no recent record of plant-derived drug discovered by big pharmaceutical companies and modern
successful plant drugs like paclitaxel (25a),
83,84
CPT (14),
84
and artemisinin (26)
85
originated from publicly
funded research. After the success of rapamycin, chalicheamicin, tacrolimus, and epothilones, the pipeline of
fermentation products also seems to have become arthritic and doubts linger if platensimycin (27), a lipid
biosynthesis inhibitor and the most notable antimicrobial agent discovered in the past few years,
86
will ever be
developed by Merck.
Conversely, several unsuccessful case histories of natural product-based drug discovery projects have been
reported. Thus, in the early 1990s, Merck paid $1.14 million to InBio, a Costa Rican conservation group, to
screen rainforest species (plants, insects, and microorganisms) for novel chemicals of interest for drug
discovery.
87
Nothing useful apparently came out of this project, which was terminated in 1999. In those
years, Shaman Pharmaceuticals, a company founded in 1989 to develop modern drugs from traditional
medicines, went as far as late-stage clinical trials for an antiviral plant extract but then went bankrupt,
88
while the 1970s–1990s had witnessed the Herculean efforts of National Cancer Institute (NCI) to discover new
natural products anticancer drugs. The program was terminated after the screening of 114 000 extracts
originating from 35 000 plant samples representing 12 000–13 000 species had apparently failed to produce a
single natural product-based anticancer drug.
89
For the sake of comparison, we can mention that the discovery
of the kinase inhibitor sorafenib (28) by Bayer involved the screening of a combinatorial library of 200 000
compounds, followed by the parallel synthesis of 1000 further analogues and was accomplished in only 4
years.
90
However, since sorafenib is the first, and so far the only, drug emerging from the screening of
combinatorial libraries, it is not clear how general for modern drug discovery its genesis is and if its success
can be considered as a real validation of combinatorial chemistry as a delivering tool.
91
Furthermore, the
clinical potential of many leads discovered during the NCI campaign was realized only after its termination.
Thus, CPT derivatives and paclitaxel were introduced into the clinic almost three decades after their
discovery, while a host of compounds from the original NCI program are being developed only now, such as
combretastatin A4 (29)
92
and maytansine (30).
93
216 Natural Products Drug Discovery

Author's personal copy
Of great relevance to the current demise of natural products is also the growing trend toward a more rational
approach to drug discovery, where ligands are designed ex novo or assembled by fragments and where emphasis
is apparently placed on design rather than on discovery. Natural products research is also intrinsically multi-
disciplinary and requires the combined efforts of natural products chemists, pharmacologists, and medicinal
chemists. These expertises are difficult to coordinate and focus and require a highly trained team, difficult to
assemble in a pharmaceutical scenario characterized by a high personnel mobility.
For these reasons, there is no doubt that, compared to synthetic libraries, natural products and extract
libraries pose a host of problems, which, albeit all singularly soluble, nevertheless conjure up to provide an
overall unattractive scenario for big pharma. The future of natural products in drug discovery and their
promotion to full potential will therefore critically depend on how well and quickly these issues will be solved,
making the process less time and resource intensive.
3.08.4.1 Intellectual Property Issues
Natural products cannot be patented as a structure but semisynthetic analogues can, as well as their uses,
isolation/manufacture processes, and specific drug formulations, although intellectual protection is obviously
weaker.
94
Furthermore, the technological asymmetry between biodiversity- and technology-rich countries has
raised considerable and still unresolved proprietary issues on natural products drugs.
95
According to the UN
Convention on Biological Diversity (CBD), countries have sovereign rights over the biological resources within
their boundaries and should establish conditions for the preservation and sustainable use of their biodiversity.
96
Source countries should be involved in projects related to their biodiversity and should share any commercial
benefit resulting from its use. These claims are objectively difficult to translate in terms of pharmaceutical
intellectual property (IP), whose reinforcement is, nevertheless, essential to contribute to the local development
of resources and to make prospecting an engine for biodiversity conservation.
97
As a result of this ambiguity,
many countries have placed barriers on the exporting of biological materials, even for noncommercial
researches. CBD is not retrospective and therefore the examples of earlier natural products discovery that
failed to produce commercial rewards to the source country lack legal meaning. Although the political debate
on biopiracy is colored with examples from developing countries (rapamycin from Easter Island, teprotide, on
which the angiotensin converting enzyme (ACE) inhibitors were molded, and tubocurarine from Brazil),
98
there are far many examples from developed countries, such as cephalosporin, rifampicin, daunomycin, and
mycophenolic acid from Italy, cyclosporin A from Norway, or paclitaxel from United States, just to mention a
few examples.
99
3.08.4.2 Access to Natural Chemical Diversity
Gaining access to biodiversity from natural habitats is legally complicated, especially for broadscale corporate
campaigns involving the collection of hundreds, or even thousands, of species. Developing societies that possess
Natural Products Drug Discovery 217

Author's personal copy
important biodiversity and developed societies endowed with advanced technologies should interact on the
basis of a principle of equity but, as we have seen, it is difficult to specify how.
97,98
In principle, countries with
rich biological resources should be able to charge companies for bioprospecting for either drugs or genetic
information that could lead to new drugs but legally binding formulas to control this ‘trade’ are difficult to
conceive and to implement. In 2002, countries signatory of the Rio Biodiversity Convention agreed on a set of
rules (Bonn Guidelines on Access and Benefit Sharing) that specify how each country should frame licenses to
allow companies to access natural resources but this arrangement was fiercely opposed by many environmental
and economic organizations. Thus, Jeremy Rifkin, who heads the Foundation on Economic Trends claimed
that ‘‘nobody has the right to enter into exclusive deals over the products of millions of years of evolution.’’
100
Political sensitivity regarding access to biodiversity from developing countries is undoubtedly one of the
reasons underlying the phasing out of natural products from big pharma.
Given the current downsizing of basic natural products research in drug companies, it is likely that many, if
not most, of the natural products drugs of the future will originate from publicly funded research, via
government organizations and academic institutions, or from venture capital small biotech companies. Taxol
(paclitaxel, 25a) was the first anticancer drug to reach a billion-dollar yearly sale and the NCI–BMS deal on the
pharmaceutical development of this compound was undoubtedly highly profitable from a corporate view-
point.
101
It was also vociferously criticized, in particular for the hijacking of the taxol name from BMS. On the
other hand, the deal made rapid access to this drug possible and its price sank after it acquired a generic state.
Aventis (then still Rho
ˆne–Poulenc Rhorer) developed a semisynthetic analogue of the natural product
(docetaxel, 25b) that enjoyed a longer protection status and topped the list of best-selling anticancer drugs
for several years.
101
Compared to natural products, synthetic compounds undoubtedly enjoy the advantage of
clarity in terms of IP but the development of taxane anticancer drugs provides an important example of how
the intrinsically difficult IP protection issues associated with natural products can be solved in a pharma-
profitable way. Furthermore, NIH has sponsored several bioprospecting projects of International Cooperative
Biodiversity Groups (ICBG) that combine academic and industrial groups and that might serve as models for
programs that combine drug discovery, conservation of environmental and genetic resources, and the establish-
ing of sustainable economic activities.
102
3.08.4.3 The Biodiversity Crisis
The prospect of turning natural resources to practical use has fuelled human activity for thousands of years and
biodiversity is one of the most valuable but least appreciated natural resources. Some of the most important
wonder drugs came from organisms not usually associated with healing, such as poisonous plants and animals,
and every species has the potential to teach us something new. The calculated yearly loss of 20 000 living
species therefore means that thousands of undiscovered and unique chemicals with potential utility will be lost
forever, along with the genetic information necessary for their assembly.
103
The current inventory of biodi-
versity is very incomplete and will undoubtedly be enlarged as exotic regions and habitats are studied, with
marine ecosystems being the largest unexplored habitat of life. We are monumentally uncertain as to how many
living species there are on earth and even the tenet that most of the higher terrestrial plants have been
discovered has been hotly debated. It has been estimated that there are at least 250 000 species of higher plants,
30 million insect species, 1.5 million species of fungi, and a similar number of algal and prokaryote species.
104
The uncertainty on these number is big since it has been reckoned that over 50 000 plants are still waiting to be
discovered.
104
Whatever the case, only a fraction of the plants known have been investigated chemically or for a
specific bioactivity and many of them were investigated decades ago with relatively crude techniques.
Considerable disagreement also exists as to where plant biodiversity is most concentrated, although a certain
consensus exists that the richest regions in terms of flowering plants are the Amazon basin, Southeast Asia, and
the Mediterranean region.
105
Owing to difficulties in cultivation, only a tiny percentage of bacteria and fungi
are known (12 and 5%, respectively) and most insects and nematodes living are still undiscovered.
105
Translating biological diversity into chemical diversity has long been the aim of phytochemistry and
represents the first step toward a rational utilization of bioresources. Modern developments in separation
techniques and spectroscopy have expedited the ‘molecular cataloguing’ aspect of phytochemistry, just like
technological advances and genomics have simplified the work of taxonomists. Nevertheless, and paradoxically,
218 Natural Products Drug Discovery

Author's personal copy
both taxonomy and phytochemistry are facing a lack of academic prestige and resources that is crippling the
cataloguing of biological and chemical diversity just at the time when it has become most urgent. Taxonomy and
phytochemistry are enabling sciences. They do not generate new ideas or test hypotheses but make it possible to
open new areas of research and translational sciences, with clearly achievable and relevant goals, which are
favored by the current funding system. Thus, a project on the health benefits of a certain diet will surely have
more chances of being funded than one on the phytochemistry of the plants on which the program is based.
3.08.4.4 Supply Issues
Biodiversity can be lost by natural catastrophes (fires, eruptions, diseases) or by human activity and is basically
less reliable than oil-derived feedstocks to secure continuous access to a product. Indeed, resupplying of an
active extract is a major drawback in natural products drug discovery and the lack of a backup sample can cause
substantial delay, especially for nonfermentable biomasses like plants and marine organisms while many
nonmicrobial natural products cannot be produced in bulk by isolation or can be produced only after
considerable efforts, as exemplified by the development of paclitaxel (25a) as an anticancer agent.
101
Furthermore, many nonmicrobial natural products cannot be produced in bulk by isolation. In general, the
reliability of marine feedstock as a bulk source of natural products has often been questioned, being much less
than that of plants and microorganisms, and the belief that marine-derived natural products may furnish better
opportunities to synthetic chemists than to medicinal chemists is still rife. The incredible ordeal represented by
the 60 g synthesis of discodermolide (31), a marine biological analogue of paclitaxel (25a), gives credit to this
idea. Discodermolide was prepared for preclinical studies by a total synthesis that required the combined efforts
of more than 43 Novartis chemists to produce 60 g of product in an overall yield of 0.65%.
106
While a
commercial chemical synthesis of discodermolide is still elusive, there is strong evidence that many marine
secondary metabolites are actually of microbial origin and that the marine source simply represent a macro-
scopic host for the microbial colony actually producing the compound of interest.
107
Thus, it has been firmly
established that tetrodotoxin, possibly the most famous marine natural product, is actually a microbial
compound and that the puffer fish only accumulates it, using it as a hormone.
108
On the positive side, several
techniques have been developed to gain access to natural products or natural products-like compounds, in
nonnatural ways, with plant tissue cultures and combinatorial genetics being the most investigated ‘rain forest’
surrogates in terms of availability of natural products.
107
3.08.4.5 Methodological Issues
The de novo construction of libraries of pure natural products is prohibitively costly compared to synthetic
libraries but a very big library of pure natural products (>10 000 compounds, >85% purity) has been assembled
by the German biotech company AnalytiCon Discovery and made commercially available, unifying, in terms of
the internal logistic of pharmaceutical companies, the screening of natural products and that of synthetic
compounds.
109
However, libraries of crude extracts rather than pure compounds are typically screened in
natural products-based drug discovery campaigns. Screening extracts in both biochemical and cell-based assays
is operatively similar to screening libraries of synthetic compounds but the readouts are plagued by factors that
occur more rarely in synthetic libraries and there is therefore great interest in the production of ‘assay-friendly’
libraries of extracts.
Natural Products Drug Discovery 219

Author's personal copy
3.08.4.5.1 Entourage effects
The isolation of morphine from opium in 1805
110
was the first demonstration that the activity of a medicinal
plant could be attributed to a single chemical constituent, initiating natural products chemistry and the search
for similar ‘quintessential’ principles in other medicinal plants. This approach was successful only for highly
active or poisonous medicinal plants (heroic drugs) while the activity of the majority of medicinal plants could
not be traced to a single constituent (magic bullet).
There is now growing awareness that the activity of most medicinal plants is the result of the synergistic action of
several constituents (magic shotgun).
111
These concepts were deftly exploited to develop Sativex, a combination of
two strains of Cannabis characterized by a high contents of tetrahydrocannabinol (THC; 32) and cannabidiol (CBD;
33), used to relieve the symptoms of multiple sclerosis and which is also under clinical development for the
treatment of cancer pain.
112
CBD, long considered pharmacological ballast, shows anti-inflammatory activity and
modulates the psychotropic effects of THC via its CB1 reverse agonism and by interfering with the hepatic
11-hydroxylation of THC, which increases the brain penetration of this psychotropic compound.
113
The ‘entourage
effect’ has been a deterrent for the mainstream and reductionist pharmaceutical exploitation of medicinal plants. In
other words, extracts of natural origin are complex systems and we do not know how much we can simplify
(fractionate) them and still have them functioning. Chronic degenerative diseases like cancer and Alzheimer’s
disease are multifactorial and mixtures of compounds, or compounds with a pleiotropic mechanism of activity, are in
principle more useful to treat these diseases than a single compound. Indeed, cancer and HIV are treated with
cocktails of drugs and not with a single agent, while synergistic combination drugs like Augmentin, an association of
a-lactam antibiotic and a lactamase inhibitor, have been developed. Nevertheless, synergies are better deduced
than planned and entourage effects are unmanageable in mainstream, magic bullet style, drug discovery campaigns.
3.08.4.5.2 False positives/negatives and reproducibility
False positives can originate from various causes, such as nonspecific hydrophobic binding, poor solubility, the
tendency to form aggregates, or the presence of denaturing agents (tannins), pigments, fluorescent compounds,
nonselective and widespread ligands like linoleic acid, or functional groups that react in a nonspecific way with
protein targets (aldehydes, epoxides, and Michael acceptors).
114
All these issues are more severe in extracts than
in synthetic libraries, where hydrophobicity, solubility, and presence of reactive functional groups and color
can be minimized at the planning stage. Conversely, extracts are generally characterized by a total lack of
information on their molecular composition and, in this sense, they are black boxes. False negatives might
originate from a too low concentration of an active compound in an extract, its chemical instability, the
interferences with the assay readout, and/or the presence of compounds with opposite activity. Again, these
issues are nonexisting or rare in synthetic libraries. Extracts are intrinsically ‘dirtier’ than synthetic libraries but
can be cleaned by prefractionation, an operation that minimizes most of the false positive issues and increases
the concentration of constituents, therefore improving the detectability of trace constituents. Several methods
to remove tannins, protein-precipitating agents, and reactive chemicals from plant extracts have been devel-
oped.
115,116
False negatives might also originate from the presence of compounds with opposite bioactivity and
some potent natural products could probably never have been discovered using modern HTS campaigns. Thus,
fiber cannabis contains THC, a cannabinoid agonist, but also CBD, a cannabinoid reverse agonist that is much
more abundant than THC.
113
Another case is Lycopodium extract, which, despite containing the very powerful
nicotinic agent huperzin A (34), also contains anticholinergic compounds with, overall, little, if any, cholinergic
activity.
117
Clearly, the interrogation of a novel target with a high-throughput campaign based on natural
products extracts might well fail to produce any useful results, since few bioassays are robust enough to
220 Natural Products Drug Discovery

Author's personal copy
withstand the screening of complex mixtures and previous prefractionation is therefore necessary. This
operation of molecular simplification limits the possibility of false positive and negative but it is undoubtedly
labor intensive, time consuming, and costly. Finally, reproducibility of activity and/or composition is often an
issue, being observed in approximately 40% of plant extracts as a result of differences in geography and time of
plant collection, or of the presence of microbial elicitors of the production of secondary metabolites.
118
3.08.4.6 Dereplication
Natural product-based hit discovery campaigns suffer from a complete lack of information on the composition
of the compounds to screen and assays are per se incapable of distinguishing between known and novel
compounds. Dereplication, the identification of known compounds responsible for the activity of an extract
before bioassay-guided fractionation,
119
is therefore important before screening, at least in campaigns aimed at
the identification of structurally novel ligands. It is therefore possible, at least in principle, that obvious ligands
are ‘rediscovered’ in any nondereplicated phytochemical screening. For instance, GABA is widespread within
plants and its presence interferes with assays of GABAergic activity, masking the presence of both GABA
inhibitors (false negative readout) and GABA mimetics (false positive readout).
120
To minimize this problem,
the NCI has developed a dereplication strategy based on HPLC fractionation with diode array detection,
collection of fractions into 96-well microtiter plates, and preparation of daughter plates for either biological
testing or mass spectrometry–electrospray ionization (MS–ESI) detection.
121
3.08.4.7 Advent of Combinatorial Chemistry and Progress in Synthetic Chemistry
The rapid identification of protein, DNA, and RNA pharmaceutical targets has driven the need for
easily prepared, chemically diverse, and target-specific small-molecule ligands.
122
HTS and combinatorial
chemistry have emerged to meet this need. HTS, whose flow rate far exceeded the capacity of standard
proprietary libraries, predates combinatorial chemistry and spurred its development. The design and synthesis
of combinatorial libraries have focused mainly on functional group variation within members of the library,
with, at least at the beginning, little, if any, stereochemical or skeletal diversification.
123
Considerable advances
have been achieved in the past years in terms of purity and structural diversity of combinatorial libraries, which,
however, remain dismally inferior to natural products in terms of diversity. Since it is nowadays accepted that
biological relevance and chemical diversity are more important than the library size, several groups have been
involved in the development of natural products-like libraries based on the combinatorial elaboration of
scaffolds inspired by natural products.
123
Current pharmaceutical research needs increasingly larger number
of compounds spanning as many molecular architectures as possible and phytochemical techniques minimizing
manipulation and purification steps must be developed. Clearly, no magic techniques of high-throughput
isolation exist and, despite all the impressive progress in isolation and structure elucidation techniques, natural
products libraries will never be competitive in terms of availability and rapidity of assembly with synthetic
libraries. At the same time, progress in synthetic chemistry and the spiraling of drug prices have made it possible
to produce by total synthesis drugs that rival the complexity and polyfunctionalization of natural products. The
anti-HIV drug enfuvirtide (Fuzeon) is a remarkable example. This 26 amino acid peptide is not produced by
Roche recombinantly in engineered cells but by total synthesis, with an investment that led to a worldwide
overall cost lowering of all peptide synthesis reagents, starting materials, and equipment.
124
Complex natural
products like huperzine A (34) and galanthamine (35) are nowadays competitively produced by synthesis
rather than by isolation,
125
and the enormous progress of the past years in synthetic methodologies and
efficiency have undoubtedly made synthesis a rival of isolation for both the discovery of new drug hits and
the production of bioactive natural products.
Natural Products Drug Discovery 221

Author's personal copy
3.08.4.8 Poor Relevance to Noncytocidal Targets
Since natural products are essentially chemical weapons, natural product-derived drugs are preeminent in the field
of oncology and anti-infective therapy,
126
while chances to identify natural products leads in screening for other
activities (cardiovascular, neurological, and metabolic) is undoubtedly weaker, since the source organism and
human proteins did not coevolve. These low hit rates should, however, be compared to those of purely synthetic
libraries and there is no shortage of examples of recent discoveries of new natural products leads and new natural
product-related targets in hot areas of research like diabetes, metabolic diseases, and Alzheimer disease. A recent
example is the identification of the dimeric flavone isoginkgetin (36) as a mechanistically new promoter of
adiponectin secretion, an important antidiabetic target.
127
Adiponectin is a hormone secreted by adipocytes that
increases insulin sensitivity and whose plasma level are low in diabetic and obese people. Screening of a library of
drug-like synthetic compounds and natural products identified isogingketin, a constituent of gingko leaves, as a
powerful inducer of adiponectin secretion, acting in a fundamentally distinct way compared to thiazolidinediones,
and involving not peroxisome proliferator-activated receptor-(PPAR-) but rather AMP-activated protein
kinase (AMPK).
127
Regarding the natural product-inspired discovery of new targets, a recent example is the
identification of TRPC6 as the antidepressant target of the phloroglucinol hyperforin (37).
128
This constituent of
St. John’s Worth inhibits the neuronal reuptake of serotonin, dopamine, and norepinephrine, behaving as a
functional biological analogue of synthetic antidepressants. However, hyperforin acts with a basically different
mechanism, inducing sodium and calcium entry mediated by specific binding to TRPC6, a nonselective ion
channel. Since neurotransmitter reuptake requires an efficient sodium gradient, its impairment translates into a
decreased amine reuptake. The therapeutic areas of infectious diseases and oncology have undoubtedly benefited
most from natural products but natural products have been successfully developed to treat human diseases in
almost all therapeutic areas and it should be remarked that statins, the commercially most successful drugs ever,
were molded on the microbial product lovastatin (38) (for more details on Natural Products of Therapeutic,
see Chapter 2.19). In 2006 alone, the sales of statins were over 20 billion dollars.
129
222 Natural Products Drug Discovery

Author's personal copy
3.08.5 Strategies in Natural Products Drug Discovery
How would penicillin have fared had the initial discovery occurred in 2007, in the absence of a clearly defined
molecular target against which were screened a mind-numbing collection of low-pedigree samples, often of
questionably purity?
S. Danishefsky, Chem. Eng. News 13 October 2003, p 103
3.08.5.1 Ethnopharmacology
Traditional medicinal practices predate modern medicine by thousands of years. All indigenous populations
have derived a pharmacopoeia unique to their environment and an enormous amount of information on the
medicinal properties of plants, fungi, and animals exists in ethnic cultures.
130
By analyzing extensive
databases of bioactivity such as the NCI list of ‘active plants’, it was calculated that plants with a traditional
use in medicine were 2–5 times more likely to generate ‘active (cytotoxic) extracts’ compared to plants
without an ethnopharmacological record.
131
Of special interest are poisonous organisms (plants, animals,
mushrooms, and microorganisms), since their ‘bad’ properties can be potentially translated into successful
therapeutic drugs.
132
Physostigmine (39), atropine (40), and tubocurarine (41) and botulinum toxin are
important examples from the past and cyclopamine
133
(42) and conotoxins
134
from current research on
poisonous organisms.
The use of medicinal plants in traditional medicine represents in principle a sort of preexisting clinical
testing and a shortcut to biologically active compounds but the translation of enthobotanical knowledge into
commercialized products is far from simple.
135
For one thing, many traditional medicines are not based on the
Hippocratic principles of disease. Thus, traditional Chinese medicine (TCM) takes a holistic approach to
treatment, emphasizing the balance and harmony of the human body. Central to its practice are concepts like
yin and yang, primal and opposite forces, and the spiritual energy known as qui, whose block causes illness.
These concepts cannot be translated into molecular terms and it is therefore hardly surprising that TCM has so
far contributed so little to mainstream drug discovery.
136
Furthermore, while issues like claim validation and
standardization can be addressed by current pharmaceutical expertise, others like sustainability of the source
and ownership of the intellectual knowledge are unusual, or downright alien, to mainstream pharmaceutical
corporate culture, as is the use of mixtures of compounds like extracts, or even of mixtures of extracts. These
problems are no doubt exacerbated by the current pharmaceutical legislation, which is well suited to cope with
monomolecular drugs or mixtures of active pharmaceutical ingredients (APIs) but is at a loss with complex
active matrixes like extracts. For this reason, special channels have been devised in the US and European Union
Natural Products Drug Discovery 223

Author's personal copy
pharma legislation to accommodate drug derived from ‘evidence-based’ ethnobotanical medicinal discovery.
137
Extracts, a fundamentally rudimental form of drug even in purified and standardized form for current
pharmaceutical standards, represent an important area of drug discovery and the recent FDA approval of
Veregen (polyphenon A), a standardized polyphenolic extract from green tea, for the management of genital
papilloma warts represent an important example on a basically new type of natural products drug, which was
approved without any evidence of mechanism of activity and on the basis of highly positive clinical results
only.
138
Traditional knowledge is disappearing faster than biodiversity and many ‘islands’ of traditional
knowledge remain to be investigated and will undoubtedly get lost forever with the current pace of globaliza-
tion. The study of folk pharmacopoeias and ethnomedicine is the basis of the discovery of several important
drugs and biological leads, as exemplified by digoxin, tubocurarine, ephedrine, atropine, and quinine.
139
Not
only plant-derived compounds but microbial products also owe their origin to ethnopharmacology, as cogently
shown by cephalosporins, whose discovery was related to the study of the so-called ‘Cagliari paradox’, namely
the very low incidence of cholera in this Sardinian town despite the lack of a public sewage system and the habit
of the inhabitants to take a bath in the polluted waters of the Su Siccu beach, later found by Brotzu to be sterile
because of the presence of the antibiotic-producing mold Cephalosporium acremomium.
140
After their isolation by
Brotzu in Cagliari, the development of cephalosporins as antibacterial agents was eventually carried out in
England and their introduction into the clinic brought rich dividends to the National Research Development
Corporation, a body set up in 1949 to exploit discoveries made by British universities and government
laboratories.
141
The clinical translation of the original discovery by Brotzu required considerable efforts from
both academy and industry but in the highly politicized context of bioprospecting, can also be perceived as a
blatant case of exploitation of both tangible (genetic resources) and intangible (knowledge) indigenous
resources.
While ethnopharmacology is undoubtedly an asset for natural products plant discovery, this approach has
some obvious limitations, even under a Hippocratic medicinal context, since many diseases are ill defined in
terms of symptoms. Thus, most cancers show little if any symptoms until the late stages of the disease and they
are not specific. It is therefore difficult to translate ethnopharmacological information into clinical clues for
cancer, despite a monumental attempt by Hartwell.
142
Even for diseases well defined in terms of symptoms,
such as fever and malaria, traditional use might have missed important plants. A striking case is artemisinin (26).
This antimalarial drug was discovered in a Chinese medicinal plant (Artemisia annua L.) that was substantially
overlooked in terms of antimalarial use in the TCM.
136
Indeed, the Jesuit penetration in China in the
seventeenth century was spurred by the healing of the Chinese emperor by Cinchona, the miracle antimalarial
plant traded by Jesuits. Pure artemisinin is not orally available, although it was reported that a certain
absorption takes place from crude extracts containing flavonoids,
143
and A. annua, even with all the limitations
implicit in the translation of folklore indications into modern medicine, was not sufficiently emphasized as an
antimalarial agent in TCM.
136
3.08.5.2 Ecology
The preservation of biodiversity goes beyond the simple cataloguing of living species but also involves the
study of their physiology and the preservation of their relationships. Biodiversity is therefore strictly related to
the conservation of a specific environment as a whole and it would be limiting to associate it to botanical
herbaria, fungal collections, or aquaria. The study of the ecology of a species can afford interesting clues for
drug discovery, as exemplified by exenatide (Byetta), a drug derived from a lizard venom and the first incretin
mimetic introduced into the clinic.
144
The Gila monster (Heloderma suspectun), a poisonous desert reptile from
the American Southwest and Northern Mexico, can withstand long periods of fasting, eating only 3 or 4 times a
year. The physiological bases for this remarkable feeding behavior was traced to the presence of a salivary
hormone (exendin-4) that slow down the digestion and the absorption of food.
145
Exendin-4, a 39 amino acid
peptide, shows an approximately 50% analogy with glucagon-like peptide-1 (GLP-1), a hormone that increases
the production of insulin when blood sugar levels are high. Exendin-4 is more potent than GLP-1 to enhance
glucose-dependent insulin synthesis from pancreatic beta cells, to decrease glucagon production, and to slow
down gastric emptying. Furthermore, exendin-4 has longer duration of action than GLP-1, with a half-life of
224 Natural Products Drug Discovery

Author's personal copy
over 2 h versus less than 1 min for the human hormone, being resistant to enzymatic inactivation by dipeptidyl
pepdidase-IV (DPP-1V). A synthetic form of exendin-4 (exenatide, Byetta) was approved by FDA in April 2005
for the control of type II diabetes in patients whose blood glucose cannot be controlled with oral diabetic agents
(metformin, sulfonylureas, or thiazolidinediones) alone.
144
The wild population of Gila monster is declining
rapidly due to habitat loss and illegal hunting for the pet trade. The project Heloderma has been established to
save the Gila monster and related species from extinction and Eli Lilly, the company that commercializes
Byetta, is making a charitable contribution to this project. Byetta is an interesting example of drug coming from
a threatened species and whose clinical exploitation is actually helping its preservation. The limitation of the
ecological approach to natural products drug discovery is that most targets of high-throughput screens are not
easily translated into observable phenomena that can provide prospecting clues. Thus, while the observation of
a fruit that does not rot can suggest the presence of antibacterial compounds, most drug targets cannot benefit
from this type of observation.
3.08.5.3 Unconventional Natural Products Sources
Plants and microorganisms, especially Actinomycetes, are the most validated sources of natural products drugs,
especially in consideration of the facility of their cultivation or fermentation. Even so, only a fraction of the
known plants and microbial species have been investigated for their pharmaceutical potential and other
biodiversity sources are still largely or completely unexplored and untapped.
146
In general, the taxonomic
and geographical diversity of bioprospecting has constantly increased and now encompasses cyanobacteria,
endophytic fungi, sponges, mollusks, seaweeds, insects, and amphibians. Particularly impressive is the bewil-
dering variety of structurally unique natural products isolated from marine organisms, often with no
counterpart in terrestrial organisms.
147
However, the difficulties of collection and scale-up of marine natural
products are formidable, also because the identity of the actual biological producer is often unknown and its
propagation in a commercial setting unpractical. Thus, it seems well established that, especially in sponges, the
production of secondary metabolites is due to coexisting microorganisms, especially cyanobacteria, and not due
to the their host.
148
The identification and fermentation of these marine microorganisms could represent a
revolutionary twist in marine natural products chemistry, paving the way for the clinical exploitation of an area
of the chemical space distinct from that of terrestrial natural products that has lagged far behind in terms of
pharmaceutical exploitation essentially because of the lack of a sustainable supply.
149
Some environmental
niches are still completely pristine in terms of bioprospecting, with Antartica being a preeminent example.
Despite its harshness, this habitat supports a thriving community of invertebrates and algae that produce very
interesting products, such as the polyketide palmerolide A (43) from the tunicate Synoicum adareanum.
150
Palmerolide A, so named from the Palmer Station on the Antarctic Peninsula in whose vicinity its animal
source was collected, is a potent antimelanoma agent and a one-digit nanomolar inhibitor of V-ATPase, a
vacuolar proton-translocating enzyme that acidifies organelles of both constitutive and regulated secretory
pathways.
150–152
Extremophile microorganisms from a variety of inhospitable terrestrial and marine sources,
such as acidic hot springs (acidophiles), alkaline lakes (halophiles), deep-sea vents (baro- and thermophiles),
polar waters, and alpine lakes (psychrophiles) hold great promise. It is not unreasonable that, just like enzymes
from extremophiles supported the discovery of PCR, also interesting drug leads might come from the study of
their secondary metabolites.
153
Natural Products Drug Discovery 225

Author's personal copy
Apart from these exotic sources, it should also be pointed out that only a fraction of soil microorganisms
can be cultured and have therefore been investigated for the production of secondary metabolites.
154
To
get around this issue, genetic material coding for secondary metabolites can be obtained directly from the
soil and expressed in a host organism. The secondary metabolites obtained so far from environmental DNA
are rather similar to those produced by fermentable microorganisms,
155
but there have been only few
studies of this type and more systematic investigations might lead to uncharted areas of the biological
chemical space. Overall, there is no shortage of areas of the world and habitats where new and unusual
chemodiversity can be discovered and the major limitation of these studies is that, since we know so little
on the ecology of unconventional environments, there are no clues to select in biorational ways the
organisms to study.
3.08.5.4 Edible Plants
Humans are daily exposed to a multitude of secondary metabolites contained in edible plants and spices. These
compounds have accompanied us during evolution, playing a role in the shaping of our genome and making us
not what we eat but rather what our ancestors have eaten.
156
Dietary secondary metabolites are not considered
as nutrients but appear to play a role, still undefined in molecular terms, in the maintenance of health, and there
is therefore great interest in their identification and in the characterization of their biological profile. Dietary
compounds are the basis of the development of highly successful drugs, such as lovastatin (38) and salicylic acid
(44), the archetypal statin and nonsteroid anti-inflammatory drugs, respectively. Lovastatin occurs in the red
yeast of rice (Monascus ruber), an ingredient of Eastern cuisine used to give a red color to the Pekinese duck,
157
while salicylic acid is ubiquitous in plants.
158
Remarkably, the isolation of lovastatin from the dietary mold M.
ruber was reported by Endo 1 year before Merck described its obtaining from Aspergillus terreus.
159
Other
important dietary drug candidates are curcumin (12) from turmeric
160
and capsaicin from hot pepper (13),
54
while traces of pharmaceutical benzodiazepins (including diazepam) occur in common edible plants like
potatoes and cherries.
161
Dietary observations have afforded many clues to drug discovery. The antiasthmatic properties of theophyl-
line (45), a caffeine metabolite and a minor constituent of tea, were discovered because of the improvement of
breathing problems of asthmatic patients who consumed strong black coffee,
162
and resveratrol (46) came under
the limelight because of the alleged protective effect of red wine in the fat-rich French diet (French paradox).
163
Resveratrol, a pleiotropic agent that has raised considerable interest as a sirtuin ligand, was recently granted
orphan drug status for the treatment of encephalomyopathy, a rare disease.
164
Also, negative dietary correla-
tions can afford clues to drug discovery. Thus, the potent immunosuppressant dammarane triterpenoid (47) was
discovered because of epidemiological correlations between the incidence of cancer and the consumption of
palmyrah flour (Borassus flabellifer), a staple food of Sri Lankan Tamils.
165
The major limitation of the many
dietary clues is that the beneficial or detrimental effects of health resist a reductionistic analysis, being the
results of a combination of principles and their bacterial and hepatic metabolites. Anthocyanosides are
remarkable examples. They are the most abundant dietary flavonoids and show a remarkable pattern of activity
in vitro but are also chemical chameleons, varying in structure, polarity, and overall charge according to the pH
of the medium and suffering from an outmost complex enteric and hepatic metabolism as well as entourage
effect in their activity.
166
Anthocyanosides such as cyanidin glucoside (48) have recently raised great interest as
antiobesity agents, due to their inhibiting properties on the differentiation of adipocytes and their lack of
toxicity.
167
226 Natural Products Drug Discovery

Author's personal copy
3.08.5.5 Derivatization, Diverted Total Synthesis, Diversity-Oriented Synthesis, and
Semisynthesis
Because of toxicity, modest activity, poor solubility and stability, or overall unsatisfactory ADMET (absorp-
tion, distribution, metabolism, elimination, toxicology) profile, many natural products are of limited clinical
use as such. Cephalosporin C (49), CPT (14), and curcumin (12) exemplify this situation in terms of
suboptimal potency, toxicity, and oral bioavailability, respectively. However, natural products can be
‘domesticated’ by suitable chemical derivatization. In some cases, chemical modification can revert activity
(iodination of the ultrapotent vanilloid resiniferatoxin (RTX; 50),
168
N-methyl for N-allyl swap in morphine
(51)
169
) or redirect it to unnatural targets, as observed for morphine (51, an opioid agonist) and its acidic
rearrangement product apomorphine (52, a dopamine ligand).
169
However, natural products are often too
complex for straightforward chemical derivatization and the exuberance of functional groups means that their
reactivity is often unpredictable, with the need to develop ad hoc solution of specific and tailored applicability.
For instance, the secondary hydroxyl of phorbol (53), a key element of its pharmacophore, is less reactive
than the adjacent tertiary hydroxyl, which can be esterifed chemoselectively even in the presence of the
primary allylic hydroxyl.
170
Patterns of reactivity like this are difficult to predict and require a careful
preliminary study, with a consequent slowing down of the drug discovery campaign. Furthermore, in complex
natural products, the reactivity of functional groups can be quenched by an unfavorable steric environment, as
exemplified by the endocyclic double bond of paclitaxel (25a), which is resistant to hydrogenation,
171
or the
C-9 tertiary hydroxyl of phorbol (53), which is characterized by total chemical inertness.
172
The manipula-
tion of these cryptic functional groups might be of enormous biological relevance and could provide a
solution to long-standing biological issues, such as the mode of binding of phorbol esters to PKC. Finally,
there are limitations in the extent of the structure–activity relationships that can be studied using the
functionalization pattern of a natural product. This is especially marked for apolar moieties that lack
functional groups or that only bear functional groups redundant for activity. To address these issues, the
concept of diverted total synthesis has been proposed by Wilson and Danishefsky.
173
The most straightfor-
ward way to assemble a complex target is by using a convergent synthesis, where smaller modules are
combined sequentially en route to the target. The reactivity pattern of these small fragments is generally
predictable and by feeding these modified fragments into the pipeline of the synthetic scheme, a full
exploration of the structure–activity relationships can be achieved. Major applications of this strategy were
described in the field of anticancer compounds, using epothilones and radicicol as leads.
173
Natural Products Drug Discovery 227

Author's personal copy
Over the previous years, there has also been an increased interest for the semisynthesis of complex natural
products, with notable achievements by Wender et al.(prostratin(54)fromphorbol(53))
174
and Baran and coworkers
(cortistatin (55) from prednisone (56).
175
By elaborating easily available compounds, semisynthesis can provide a
scalable access to complex structures difficult to source. It requires great ingenuity since synthetic creativity is
constrained by the connectivity and configuration of the starting material. The industrial production of paclitaxel
(25a)
171
and of ecteinascidin-743 (57) are examples of important industrial applications of semisynthesis to the
production of natural products drugs. The marine anticancer compound ecteinascidin-743 (Yondelis), used for the
treatment of soft-tissue sarcoma, was originally isolated from the marine tunicate Ecteinascidia turbinata. Wild harvest
of this organism could not have supported its clinical development, which relied on aquaculture to afford the small
amounts required at that stage. A total synthesis was reported by Corey et al.
176
but the supply problem was eventually
solved by semisynthesis from a related microbial compound, cyanosafracin B (58), from Pseudomonas fluorescens.
177
228 Natural Products Drug Discovery

Author's personal copy
Many natural products are easily available in multigram amounts by isolation and boast a rich decoration of
reactive functional groups as well as complex skeleton amenable to rearrangement. This chemical exuberance
could be coupled to efficient technology platforms like combinatorial chemistry or diversity-oriented synth-
esis
178,179
to expand the pool of natural products and generate new modulators of biological activity.
180
Many
attempts have also been made to combine the quality of natural products and the speed and efficiency of
modern synthetic technologies by using natural products motifs as scaffolds to build combinatorial libraries.
The efficiency of this process is exemplified by the discovery of fexaramine (59), an inhibitor of farnesoid
X-receptor,
181
and of secramine (60), an inhibitor of protein trafficking by the Golgi apparatus.
182
These
molecular probes emerged from synthetic combinatorial libraries built on the 2,2-dimethylbenzopyran motif
183
and on the tetracyclic core of galanthamine.
182
3.08.5.6 Extract Engineering
Crude extracts often contain a series of related compounds that share a common functionality that can make up
for a large proportion of the extract. The crude extracts can be directly treated with a reagent specific for this
functionality, generating a modified ‘secondary’ extract containing semisynthetic compounds that can be
screened for a useful activity. In this way, the exploitable molecular diversity from a given biological source
can be substantially increased. This principle was proposed by Furlan, who investigated the antifungal activity
of a series of natural extracts containing flavones. Noticing the paucity of N–N motifs in natural products
compared to their abundance in drugs, the extract was treated with hydrazine, affording an engineered extract
where the flavone constituents had been converted to their corresponding pyrazoles by remodeling of the
central C ring. Remarkably, while the natural extract lacked antifungal activity, the engineered one showed
interesting activity against human fungal pathogens, traced by bioassay-directed fractionation, to the flavone-
derived pyrazole (61).
184
This ingenuous strategy should be further investigated for its generality and holds
undoubtedly great potential, although not many extracts are amenable to simple engineering.
3.08.5.7 Engineered Biosynthesis (Mutasynthesis, Combinatorial, and Transgenic
Biosynthesis)
The living organisms are just a tiny fraction of those that have inhabited the earth and that went extinct during
evolution. The extraordinary metabolic richness and unicity of living fossils like the gingko tree points to a
chemically exuberant past that we will never be able to recapture. Millions of transient natural products were
evolutionarily deselected along the pathway that eventually led to the natural products of today. Thus,
hydrophobic hopanoid pentacyclic triterpenoids arose early in evolution (Archebacteria) as integral stabilizers
Natural Products Drug Discovery 229

Author's personal copy
of hydrophobic membranes, followed by phytoesterols in plants, and eventually cholesterol in animal cells.
185
The very fact that squalene and squalene oxide can be cyclized in almost 100 different folding patterns to afford
cyclosqualenoids gives a glimpse of the approach followed by nature to optimize natural products and generate
today’s chemodiversity, and of the intrinsic potential of biosynthetic pathways to generate a bewildering array
of different structures.
185
Plants and microorganisms have biogenetic pathways that are expressed only under
certain conditions and there is an enormous hidden chemical diversity apparent only at the genome level. We
might ignore the reasons as to why most folding of squalene and squalene oxide were either never considered
by nature or evolutionarily deselected but, thanks to molecular genetics, we are now in the position to
randomly mutate key biogenetic enzymes, generating natural (since enzyme-derived) products in an unnatural
way (molecular biology) and somehow mimicking evolution (for more details, see Chapter 2.20).
While biosynthetic engineering is still in its infancy, modification of a biosynthetic way by the addition of
suitable building blocks has been pursued since the early studies on -lactam antibiotics, as testified by the
industrial production of penicillin V (phenoxymethylpenicillin, 62) by the addition of phenoxyacetic acid to
fermentation of Penicillum chrysogenum, a process established already in the 1950s.
185
Since the capacity to
produce the natural compounds is retained, precursor-directed biosynthesis leads to a mixture of natural and
unnatural compounds, resulting from competition between the natural building block and its unnatural
analogue. To overcome this limitation, mutasynthesis, which is the use of microorganisms where the produc-
tion of a specific building block is deficient because of an induced genetic mutation, has been developed. By
blocking the biosynthesis of a specific precursor, the production of a complex compound becomes dependent
on the supplementation with that specific precursor, which acts as a sort of metabolic ‘vitamin’. The loose
substrate specificity of many biosynthetic enzymes makes it possible to replace the natural precursor with
modified versions of it. Mutasynthesis is especially suitable for modification of compounds having a modular
structure. Thus, the aminocoumarin hsp90 inhibitor antibiotic chlorobiocin (63) consists of three elements, an
aminocoumarin core, an acylated novobiose moiety, and a 3-prenyl-4-hydroxybenzoyl group (dimethylallyl-
hydroxybenzoic acid, DMAHB). The introduction of the prenyl group is achieved by the dimethylallyl
transferase CloQ and, by using molecular engineering, a strain of Streptomyces roseochromogenes was constructed
where the cloQ gene was inactivated. Supplementation with analogues of DMAHB led to their incorporation
into the biogenetic pathway and to the generation of chlorobiocin analogues.
186
A similar strategy but based on
the shikimate-derived 4,5-dihydroxycyclohex-1-enecarboxylic acid was employed to generate analogues of the
immunosuppressant polyketide rapamycin (64).
187
In combinatorial biosynthesis, genes from different but related biosynthetic pathways are combined to
produced new compounds and this strategy has been particularly successful with polyketides.
188
These
modular compounds represent the single most successful class of natural products drugs, with a lineup of
compounds that encompasses first-in-the-class agents like lovastatin, erythromycin, tetracycline, doxorubicin,
230 Natural Products Drug Discovery

Author's personal copy
amphotericin B, tacrolimus, and avermectin. Polyketides are built from a linear chain of carbon atoms
generated by sequential reactions governed by polyketide synthases (PKSs), basically enzymic complexes
that act like an assembly line tethering a starter unit and growing it. At the end of the process, the chain is
untethered and cyclized by non-PKS enzymes (see Chapters 1.02–1.07). Additional enzymatic reactions
introduce further decorations, such as sugars and methyl groups, while some PKSs also have ketone-modifying
properties. Since genes in a polyketide pathway are always clustered together in contiguous DNA sequences,
their isolation is easy, unlike other biogenetic pathways whose genes are dispersed in different chromosomal
locations and must be isolated one at a time. Thus, a study of the biosynthesis of the polyketide antibiotic
erythromycin (65) has resulted in the identification of some 28 domains. Repositioning the sequence of the
corresponding genes enabled then to produce new ‘unnatural’ natural products.
189
A similar combinatorial
approach was applied to the production of epothilones and to nonribosomal peptides.
190
Natural products can, in principle, be also obtained from a direct biotechnological route, where all the genes
involved in its biosynthesis are expressed in a fermentable host. The transgenic production of the antimalarial
sesquiterpene lactone artemisinin (26) is currently investigated as a cheap alternative to isolation from A. annua
L. or to total synthesis.
191
A biochemical and chemical precursor of artemisinin (artemisinic acid, 66) has been
produced in acceptable yield from the fermentation of an engineered strain of the yeast Saccharomyces cerevisiae
where the production of farnesyl diphosphate was diverted from the triterpenoid sink to the sesquiterpene pool.
The amorphadiene synthase gene and a cytochrome P-450 monooxygenase from A. annua were then expressed
in this engineered yeast, overall resulting in the conversion of farnesyl diphosphate into artemisinic acid.
191
There are clearly several strategies to ‘take the nature out of natural products’ and produce them in a
nonnatural way. Remarkably, these strategies have relevance not only for the mass production of a natural
products drug but also for providing access to natural products-related chemodiversity.
3.08.6 Conclusions
The point is not that natural products will solve all problems. It is that a lot of problems are not being solved
because natural products are not being examined.
S. J. Gould, Chem. Eng. News 13 October 2003, p 103
There is no doubt that natural products represent the best and most validated source to start a drug discovery
campaign to a new druggable target but natural products can be difficult to access efficiently and effectively,
unsuitable for further development due to poor ADMET properties, and plagued by IP issues. In the current
Natural Products Drug Discovery 231

Author's personal copy
scenario of drug discovery, the dwindling use of natural products as pharmaceutical leads seems related to the
intrinsically slower and more resource-intensive nature of natural products research compared to combinator-
ial chemistry and rational (ab initio) drug design. To remain competitive in drug discovery, natural products
research should sharpen its tools by proper methodological evolution, interfacing with the current strategies of
drug discovery, and overall, moving to higher throughput. In general, natural product-based drug discovery
activities should be integrated with complementary technologies, such as combinatorial chemistry and rational
drug discovery, and not be pursued alone in an independent fashion. They should also take advantage of
techniques complementary to bioprospecting, such as derivatization of existing and easily available natural
products, diverted total synthesis, and the high-throughput de novo construction of natural product-like
scaffolds. Natural products have a function in the environment and nature is the functional filter that is lacking
in combinatorial chemistry. A small collection of ‘smart’ compounds like those present in a plant extract or a
fermentation broth will always be more valuable than a collection of randomly assembled synthetic compounds
but the access to these ‘intelligent’ collections should be made technically easier and legally transparent, while
the pharmacokinetic and proprietary profile of natural products could be improved by tailor-made chemical
modification. The transition from paclitaxel (25a) to docetaxel (25b), from artemisinin (26) to artesunate
(67),
192
or from epothilone B (68) to ixabepilone (69),
193
just to mention only recent examples, cogently
demonstrates the success of this approach.
Given a promising natural product lead, there seems to be no difficulty in convincing big pharma to invest in
its chemical derivatization and development. What is getting increasingly difficult is, paradoxically, to convince
corporate decision makers that interesting natural products ligands, hits, leads, and even readymade drugs can
originate from the study of biodiversity and of natural products libraries. It seems therefore logical to end up
with a quotation from Samuel Danishefsky, possibly the most outspoken paladin for natural products in drug
discovery, who, ‘‘at the risk of sounding Neanderthal,’’ urged drug companies to ‘‘get back to the screening of
natural products’’ and ‘‘critically examine the prevailing supposition that synthesizing zillions of compounds at
a time is necessarily going to cut the costs of drug discovery or fill pharma pipelines with new drugs anytime
soon.’’
194
References
1. A. M. Rouhi, Chem. Eng. News 13 October 2003, 77–91.
2. I. Paterson; E. A. Anderson, Science 2005,310, 451–453.
3. W. R. Strohl, Drug Discov. Today 2000,5, 39–41.
4. G. Cragg; D. Newman, Chem. Br. 2001, 22–26.
5. J. Meinwald; T. Eisner, Helv. Chim. Acta 2003,86, 3633–3637.
6. D. J. Newman; G. M. Cragg; K. M. Snader, J. Nat. Prod. 2003,66, 1022–1037.
7. A. M. Rouhi, Chem. Eng. News 13 October 2003, 93–103, 104–106.
8. D. J. Newman; G. M. Cragg; D. G. I. Kingston, Natural Products as Pharmaceuticals and Sources for Lead Structures. In The
Practice of Medicinal Chemistry; 2nd ed.; C. G. Wermuth, Ed.; Academic Press: London; 2003; pp 91–110.
9. M. S. Butler, J. Nat. Prod. 2004,67, 2141–2153.
10. J. Clardy; C. Walsh, Nature 2004,432, 829–837.
11. F. E. Koehnm; G. T. Carter, Nat. Rev. Drug Discov. 2005,4, 206–220.
12. M. S. Butler, Nat. Prod. Rep. 2005,22, 162–195.
13. D. D. Baker; M. Chu; U. Oza; V. Rajgarhia, Nat. Prod. Rep. 2007,24, 1225–1244.
14. J. D. McChesney; S. K. Venkataraman; J. T. Henri, Phytochemistry 2007,68, 2015–2022.
15. A. Saklani; S. K. Kutty, Drug Discov. Today 2008,13, 161–171.
232 Natural Products Drug Discovery

Author's personal copy
16. M. Pucheault, Org. Biomol. Chem. 2008,6, 424–432.
17. M. S. Butler, Nat Prod. Rep. 2008,25, 475–516.
18. O. Potterat; M. Hamburger, Progr. Drug Res. 2008,65, 47–118.
19. D. J. Newman, J. Med. Chem. 2008,51, 2589–2599.
20. P. J. Hajduk; J. R. Huth; S. W. Fesik, J. Med. Chem. 2005,48, 2518–2525.
21. C. P. Adams; V. V. Brantner, Drug Dev. 2006,25, 23–24.
22. I. Kola; J. Landis, Nat. Rev. Drug Discov. 2004,3, 711–716.
23. L. Smith, The Pink Sheet 10 December 2008, 13.
24. N. Blow, Nature 2007,450, 1117–1120.
25. C. Dombrowski, The Pink Sheet 2008,70 (23), 27–28.
26. P. Geborek; E. Nitelius; S. Noltorp; H. Petri; L. Jacobsson; L. Larson; T. Saxne; I. Leden, Ann. Rheum. Dis. 2005,64, 1805–1807.
27. A. Catlin; C. Cowan; S. Heffler; B. Washington, Health Aff. (Millwood) 2007,26, 142–153.
28. A. Johson, The Wall Street Journal 12 December 2007.
29. J. Merril, The Pink Sheet 10 December 2008, p 5.
30. P. Imming; C. Sinning; A. Meyer, Nat. Rev. Drug. Discov. 2006,5, 821–832.
31. K. H. Altmann; J. Gertsch, Nat. Prod. Rep. 2007,24, 327–352.
32. Y. Uehara, Curr. Cancer Drug Targets 2003,3, 325–330.
33. P. Workman, Curr. Cancer Drug Targets 2003,3, 297–300.
34. J. M. McPartland; G. W. Guy, The Evolution of Cannabis and Coevolution with the Cannabinoid Receptor – A Hypothesis. In The
Medicinal Uses of Cannabis and Cananbinoids; G. W. Guy, B. A. Whittle, P. J. Robson, Eds.; Pharmaceutical Press: London,
2004; pp 71–101.
35. E. Jung; A.-L. Veuthney; E. Gasteiger; A. Bairoch, Proteomics 2001,1, 262–268.
36. A. L. Hopkins; C. R. Groom, Nat. Rev. Drug Discov. 2002,1, 727–730.
37. J. P. Overington; B. Al-Lazikani; A. L. Hopkins, Nat. Rev. Drug. Discov. 2006,5, 993–996.
38. R. Breinbauer; I. R. Vetter; H. Waldmann, Angew. Chem. Int. Ed. 2002,41, 2878–2890.
39. B. M. McArdle; M. R. Campitelli; R. J. Quinn, J. Nat. Prod. 2006,69, 14–17.
40. B. A. Halkier; J. Gershenzon, Ann. Rev. Plant Biol. 2006,57, 303–333.
41. E. Block, Angew Chem Int. Ed. 1992,104, 1158–1203.
42. K. C. Nicolaou; W. M. Dai, Ang. Chem. Int. Ed. 1991,30, 1387–1416.
43. M. L. Guzman; R. M. Rossi; L. Karnischky; X. Li; D. R. Peterson; D. S. Howard; C. T. Jordan, Blood 2005,105, 4163–4169.
44. A. Louw; P. Swart; S. S. de Kock; K. J. van der Merwe, Biochem. Pharmacol. 1997,53, 189–197.
45. A. Louw; P. Swart, Endocrinology 1999,140, 2044–2053.
46. P. Dewint; V. Gossye; K. De Bosscher; W. Vanden Berghe; K. Van Beneden; D. Derocer; S. Van Calenbergh; U. Mueller-Ladner;
B. Vander Cruyssen; G. Verbruggen; G. Haegeman; D. Elewaut, J. Immunol. 2008,180, 2608–2615.
47. R. D. Couch; R. G. Browning; T. Honda; G. W. Gribble; D. L. Wright; M. B. Sporn; A. C. Anderson, Bioorg. Med. Chem. Lett.
2005,15, 2215–2219.
48. K. T. Liby; M. M. Yore; M. B. Sporn, Nat. Rev. Cancer 2007,7, 357–369.
49. J. C. Espin; M. T. Garcia-Conesa; F. A. Toma´ s-Barbera´n,Phytochemistry 2007,68, 2986–3008.
50. J. W. Fahey; T. W. Kensler, Chem. Res. Toxicol. 2007,20, 572–576.
51. D. M. Goldstein; N. S. Gray; P. P. Zarrinkar, Nat. Rev. Drug Discov. 2008,7, 391–395.
52. J. Jossang; H. Bel-Kassaoui; A. Jossang; M. Seuleman; A. Nel, J. Org. Chem. 2008,73, 412–417.
53. C. Spickler; M. N. Brunelle; L. Brakier-Gingras, J. Mol. Biol. 1997,273, 586–599.
54. G. Appendino, Capsaicin and Capsaicinoids. In Modern Alkaloids: Structure, Isolation, Synthesis and Biology; E. Fattorusso,
O. Taglialatela-Scafati, Eds.; Wiley: Weinheim, 2007; pp 75–112.
55. G. Appendino; A. Minassi; A. Pagani; A. Ech-Chahad, Curr. Pharm. Design 2008,14, 2–17.
56. S. Sirikantaramas; M. Yanazaki; K. Saito, Proc. Natl. Acad. Sci. U.S.A. 2008,105, 6782–6786.
57. S. Bruzzone; I. Moreschi; C. Usai; L. Guida; G. Damonte; A. Salis; S. Scarfi; E. Millo; A. De Flora; E. Zocchi, Proc. Natl. Acad. Sci.
U.S.A. 2007,104, 5759–5764.
58. V. A. Halim; A. Vess; D. Scheel; S. Rosahl, Plant Biol. (Stuttg.) 2006,8, 307–313.
59. A. A. Begum; S. Leibovitich; P. Migner; F. Zhang, J. Exp. Bot. 2001,52, 1537–1543.
60. E. Cundliffe, Annu. Rev. Microbiol. 1989,43, 207–233.
61. D. A. Hopwood, Mol. Microbiol. 2007,63, 937–940.
62. S. Singh; M. H. Hager; C. Zhang; B. R. Griffith; M. S. Lee; K. Hallenga; J. L. Markley; J. S. Thorson, ACS Chem. Biol. 2006,1,
451–460.
63. T. Henkel; R. M. Brunne; H. Mueller; F. Reichel, Angew. Chem. Int. Ed. Engl. 1999,38, 643–647.
64. X.-H. Xo; G.-M. Yao; Y.-M. Li; J.-H. Lu; C. J. Lin; X. Wang; C.-H. Kong, J. Nat. Prod. 1993,66, 285–288.
65. W. Sneader, Drug Prototypes and their Exploitation; Wiley: Chichester, 1996; p 446.
66. B. Lefkove; B. Govindarajan; J. L. Arbiser, Expert Rev. Anti Infect. Ther. 2007,5, 373–379.
67. M. Yoshida; S. Horinouchi; T. Beppu, Bioassays 1995,17, 423–430.
68. M. Groll; R. Huber, Biochim. Biophys. Acta 2004,1695, 33–44.
69. X. Tao; F. Fan; V. Hoffmann; C. Y. Gao; N. S. Longo; P. Zerfas; P. E. Lpsky, Arthritis Rheum. 2008,58, 1774–1783.
70. P. Jones; C. Steinkueler, Curr. Pharm. Des. 2008,14, 545–561.
71. S. Omura; K. Matsuzaki; T. Fujimoto; K. Kosuge; T. Furuya; S. Fujita; A. Nakagawa, J. Antibiot. (Tokyo) 1991,44, 117–118.
72. D. P. Arya, Top. Heterocycl. Chem. 2006,2, 129–152.
73. C. Lipinski; A. Hopkins, Nature 2004,432, 855–861.
74. M. Feher; J. M. Schmidt, J. Chem. Inf. Comput. Sci. 2003,43, 218–227.
75. G. M. Cragg; D. J. Newman; K. M. Snader, J. Nat. Prod. 1997,60, 52–60.
76. K. P. Mishra; L. Ganjum; M. Sairam; P. K. Banerjee; R. C. Sawhney, Biomed. Pharmacother. 2008,62, 94–98.
Natural Products Drug Discovery 233

Author's personal copy
77. G. R. Elridge; H. C. Vervoort; C. M. Lee; P. A. Cremin; C. T. Williams; S. M. Hart; M. G. Goering; M. O’Neil; M. Johnson; L. Zeng,
Anal. Chem. 2002,74, 3963–3971.
78. H. N. Woller, Mol. Diversity 1999,4, 47–52.
79. M. Hamburger, Chimia 2006,60, 14–18.
80. R. N. Young, Pure Appl. Chem. 1999,71, 1655–1661.
81. A. I. Graul; E. Cruces; L. Revel; N. Serradell; E. Rosa, Drug News Perspect. 2008,21, 44–58.
82. E. K. Wilson, Chem. Eng. News, 13 February 2006, pp 86–87.
83. G. M. Cragg, Med. Res. Rev. 1998,18, 315–331.
84. M. E. Wall; M. C. Wani, Cancer Res. 1995,55, 753–760.
85. E. Hsu, Br. J. Clin. Pharmacol. 2006,61, 666–670.
86. J. Wang; S. M. Soisson; K. Young; W. Shoop; S. Kodali; A. Galgoci; R. Painter; G. Parthasarathy; Y. S. Tang; R. H. S. Cummings;
K. Dorso; M. Motyl; H. Jayasuriya; J. Ondeyka; K. Herath; C. Zhang; L. Hernandez; J. Allocco; A. Basilio; J. R. Tormo;
O. Genilloud; F. Vicente; F. Pelaez; L. Colwell; S. H. Lee; B. Michael; T. Felcetto; C. Gill; L. L. Silver; J. D. Hermes; K. Bartizal;
J. Barrett; D. Schmatz; J. W. Becker; D. Cully; S. B. Singh, Nature 2006,441, 358–361.
87. M. D. Coughlin, Jr.; M. Columb, J. Transn. Law 1993,31, 337–375.
88. Challenges in Negotiating and Implementing ABS Agreements. http://www.iisd.org/pdf/2006/abs_session2_augustine.ppt
89. G. M. Cragg; M. R. Boyd; J. H. Cardellina; M. R. Grever; S. A. Schepartz; K. M. Snader; M. Suffness, Human Medicinal Agents
from Plants; ACS Symposium Series 534; American Chemical Society: Washington, 1993.
90. T. B. Lowinger; B. Riedl; J. Dumas; R. A. Smith, Curr. Pharm. Des. 2002,8, 2269–2278.
91. D. Pytel; T. Sliwinski; T. Poplawski; D. Ferriola; I. Maisterek, Anticancer Agents Med. Chem., 2009,9, 66–76.
92. G. C. Tron; T. Pirali; G. Sorba; F. Pagliai; S. Busacca; A. A. Genazzani, J. Med. Chem. 2006,49, 3033–3044.
93. W. C. Widdison; S. D. Wilhelm; E. E. Cavanagh; K. R. Whiteman; B. A. Leece; Y. Kovtun; V. S. Goldmacher; H. Xie;
R. M. Steeves; R. J. Lutz; R. Zhao; L. Wang; W. A. Bla¨ ttler; R. V. Chari, J. Med. Chem. 2006,49, 4392–4408.
94. M. R. Boyd, J. Ethnopharmacol. 1996,51, 17–25.
95. G. A. Cordell, Phytochemistry 2000,55, 463–480.
96. G. A. Cordell, Phytochemistry 1995,40, 1585–1612.
97. T. D. Mays; K. D. Mazan, J. Ethnopharmacol. 1996,51, 93–102.
98. M. A. Gollin, Nat. Biotechnol. 1999,17, 921–922.
99. W. Snader, Drug Prototypes and Their Exploitation; Wiley: Chichester, 1996.
100. F. Pearce, New Scientist 18 April 2002, p 14.
101. J. Goodman; V. Walsh, The Story of Taxol. Nature and Politics in the Pursuit of an Anti-Cancer Drug; Cambridge University
Press: Cambridge, 2001.
102. A. M. Rouhi, Chem. Eng. News 7 April 1997, 14–29.
103. U. Schueklenk; A. Kleinsmidt, Dev. World Bioeth. 2006,6, 112–134.
104. S. L. Pimm; G. J. Russellm; J. L. Gittleman; T. M. Brooks, Science 1995,269, 347–350.
105. F. Pearce, New Scientist 29 June 2006, p 11.
106. S. J. Mickel, Curr. Opin. Drug Discov. Devel. 2004,7, 869–881.
107. M. Saleem; M. S. Ali; S. Hussain; A. Jabbar; M. Ashraf; Y. S. Lee, Nat. Prod. Rep. 2007,24, 1142–1152.
108. J. W. Daly, J. Nat. Prod. 2004,67, 1211–1215.
109. Analyticon Discovery. http://www.ac-discovery.com/english/go.html
110. F. W. Sertuerner, Trommsdorff’s J. Pharm. 1805,13,234.
111. R. Mechoulam; S. Ben-Shabat, Nat. Prod. Rep. 1999,16, 131–143.
112. M. P. Barnes, Expert Opin. Pharmacother. 2006,7, 607–615.
113. R. G. Pertwee, Br. J. Pharmacol. 2008,153, 199–215.
114. R. Verpoorte, Drug Discov. Today 1998,3, 232–238.
115. M. Zhu; J. D. Phillipson; P. M. Greengrass; N. E. Bowery; Y. Cai, Phytochemistry 1999,44, 441–447.
116. G. M. Rishton, Drug Discov. Today 1997,2, 382–384.
117. X. Ma; C. Tan; D. Zhu; D. R. Gang; P. Xiao, J. Ethnopharmacol. 2007,113, 15–34.
118. G. A. Cordell, Phytochemistry 2000,55, 463–480.
119. N. CanMiddlesworth; R. J. P. Cannell, Natural Products Isolation. In Methods in Biotechnology; R. J. P. Cannell, Ed.; Humana
Press: Totowa, NJ, 1998; Vol. 4, pp 279–327.
120. H. J. Kim; I. Baburin; S. Khom; S. Hering; M. Hamburger, Planta Med. 2008,74, 521–526.
121. P. L. Smith; K. N. Maooney; R. G. Pothen; J. Clardy; D. E. Clapham, J. Biol. Chem. 2006,281, 29897–29904.
122. S. L. Schreiber, Science 2000,287, 1964–1969.
123. L. A. Marcaurelle; C. W. Johannes, Prog. Drug Res. 2008,66, 189–216.
124. B. L. Bray, Nat. Rev. Drug Discov. 2003,2, 587–593.
125. J. Marco-Contelles; M. do Carmo Carreiras; C. Rodrı´guez; M. Villaroya; A. G. Garcia, Chem. Rev. 2006,106, 116–133.
126. M. S. Butler; D. J. Newman, Prog. Drug Res. 2008,65, 3–44.
127. G. Liu; M. Grifman; J. Macdonald; P. Moller; F. Wong-Stall; Q. X. Li, J. Endocrinol. 2007,194, 569–578.
128. K. Leuner; V. Kazanski; M. Mu¨ ller; K. Essin; B. Henke; M. Gollasch; C. Harteneck; W. E. Muller, FASEB J. 2007,21, 4101–4111.
129. P. Evers, SCRIP 21 May 2008, pp 31–34.
130. M. J. Plotkin, Medicine Quest: In Search of Nature’s Healing Secrets; Viking: New York, 2000.
131. R. W. Spjut; A. R. E. Perdue, Cancer Treat. Rep. 1976,60, 979–985.
132. Y. Gaillard; G. Pepin, J. Chromatogr. B 1999,733, 181–229.
133. E. E. Bar; D. Stearns, Expert Opin. Investig. Drugs 2008,17, 185–195.
134. B. M. Olivera; R. W. Teichert, Mol. Interv. 2007,7, 251–260.
135. M. Heinrich; S. Gibbons, J. Pharm. Pharmacol. 2001,53, 425–432.
136. W. F. Li; J. C. Jiang; J. Chen, Arch. Med. Res. 2008,39, 246–251.
137. B. M. Schmidt; D. M. Ribnicky; P. E. Lipsky; I. Raskin, Nat. Chem. Biol. 2007,3, 360–366.
234 Natural Products Drug Discovery

Author's personal copy
138. G. Gross, Hautarzt 2008,59, 31–35.
139. M. Heinrich; P. Bremner, Curr. Drug Targets 2006,7, 239–245.
140. G. Bo, Clin. Microbiol. Infect. 2000,6(Suppl. 3), 6–9.
141. W. Sneader, Drug Prototypes and their Exploitation; Wiley: Chichester, 1996; pp 489–491.
142. J. L. Hartwell, Lloydia 1967,30, 379–426 (and the other seven articles in the series).
143. B. C. Elford; M. F. Roberts; J. D. Phillipson; R. J. Wilson, Trans. R. Soc. Trop. Med. Hyg. 1987,81, 434–436.
144. C. Triplitt; E. J. Chiquette, Am. Pharm. Assoc. 2003,46, 44–52.
145. J. Eng; W. A. Kleinman; L. Singh; G. Singh; J. P. Raufman, J. Biol. Chem. 1992,15, 7402–7405.
146. A. Harvey, Drug Disc. Today 2000,7, 294–300.
147. T. E. Adrian, Curr. Pharm. Des. 2007,13, 3417–3426.
148. T. L. Simmons; R. C. Coates; B. R. Clark; N. Eugene; D. Gonzalez; E. Esquenazi; P. C. M. Dorrestein; W. H. Gerwick, Proc. Natl.
Acad. Sci. U.S.A. 2008,105, 4587–4594.
149. M. Gordaliza, Clin. Transl. Oncol. 2007,9, 767–776.
150. T. Divabalange; C. D. Amsler; J. B. McClintock; B. J. Baker, J. Am. Chem. Soc. 2006,128, 5630–5631.
151. X. Jiang; B. Liu; S. Lebreton; J. K. Brabander, J. Am. Chem. Soc. 2007,129, 6386–6387.
152. K. C. Nicolaou; R. Guduru; Y. P. Sun; B. Banerji; D. Y. Chen, Angew. Chem. Int. Ed. Engl. 2007,46, 5896–5900.
153. C. Chiraldi; M. De Rosa, Trends Biotechnol. 2002,20, 515–521.
154. C. A. Roessner; A. L. Scott, Annu. Rev. Microbiol. 1996,50, 467–490.
155. S. F. Brady; C. J. Chao; J. Clardy, Appl. Environ. Microbiol. 2004,70, 6865–6870.
156. G. Appendino; O. Taglialatela-Scafati, Drug-Like Compounds from Food Plants and Spices. In Dietary Supplements of Plant
Origin; M. Maffei, Ed.; Taylor & Francis: London, 2003; pp 43–74.
157. A. Endo, J. Antibiot. 1979,32, 852–854.
158. I. Raskin, Plant Physiol. 1992,99, 799–803.
159. A. W. Alberts; J. Chen; G. Kuron; V. Hunt, Proc. Natl. Acad. Sci. U.S.A. 1980,77, 3957–3961.
160. A. Goel; A. B. Kunnumakkara; B. B. Aggarwal, Biochem. Pharmacol. 2008,75, 787–809.
161. J. H. Medina; C. Pena; M. Piva; C. Wolfman; M. L. de Stein; C. Wasowski; C. Da Cunha; I. Izquierdo; A. C. Paladini, Mol.
Neurobiol. 1992,6, 377–386.
162. H. Salter, Asthma. Its Pathology and Treatment; Churchill: London, 1860; p 181.
163. J. C. Espin; M. T. Garcia-Conesa; F. A. Toma´ s-Barbera´n,Phytochemistry 2007,68, 2986–3008.
164. SCRIP 3362, 16 May 2008, p 19.
165. L. Re´ve´ sz; P. Hiestand; L. La Vecchia; R. Naef; H. U. Naegeli; L. Overer; H. J. Roth, Bioorg. Med. Chem. Lett. 1999,9,
1521–1526.
166. F. Galvano; L. la Fauci; P. Vitaglione; V. Fogliano; L. Vanella; C. Felgines, Ann. Ist. Super. Sanita
`2007,43, 382–393.
167. T. Tsuda, J. Agric. Food Chem. 2008,56, 642–646.
168. P. Wahlm; C. Foged; S. Tullin; C. Thomsen, Mol. Pharmacol. 2001,59, 9–15.
169. W. Sneader, Drug Prototypes and their Exploitation; Wiley: Chichester, 1996; pp 85–89.
170. N. Ma´ rquez; M. A. Calzado; G. Sa´ nchez-Duffhues; M. Pe´ rez; A. Minassi; A. Pagani; G. Appen dino; L. Diaz;
M. A. Mun
˜oz-Ferna´ ndez; E. Mun
˜oz, Biochem. Pharmacol. 2008,75, 1370–1380.
171. D. G. I. Kingston; P. G. Jagtap; H. Yuan; L. Samala, Fortschr. Chem. Org. Naturst. 2002,84, 53–225.
172. G. Appendino; A. Bertolino; A. Minassi; R. Annunziata; A. Szallasi; L. De Petrocellis; V. Di Marzo, Eur. J. Org. Chem. 2004,
3413–3421.
173. R. M. Wilson; S. J. Danishefsky, J. Org. Chem. 2006,71, 8329–8351.
174. P. A. Wender; J. M. Kee; J. M. Warrington, Science 2008,320, 649–652.
175. R. A. Shenvi; C. A. Guerrero; C. C. Li; P. S. Baran, J. Am. Chem. Soc. 2008,130, 7241–7243.
176. E. J. Corey; D. Y. Gin; R. S. Kania, J. Am. Chem. Soc. 1996,118, 9202–9203.
177. C. Cuevas; M. Pe´ rez; M. J. Martı´n; J. L. Chicharro; C. Fernandez-Rivas; M. Flores; A. Francesch; P. Gallego; M. Zarzuelo; F. de
la Calle; J. Garcı´a; C. Polanco; I. Rodrı´guez; I. Manzanares, Org. Lett. 2000,2, 2545–2548.
178. R. J. Spandl; A. Bender; D. R. Spring, Org. Biomol. Chem. 2008,6, 1149–1158.
179. L. A. Marcaurelle; C. W. Johannes, Prog. Drug Res. 2008,66, 189–216.
180. G. Appendino; G. C. Tron; T. Jarevang; O. Sterner, Org. Lett. 2001,3, 1609–1612.
181. M. Downes; M. A. Verdecia; A. J. Roecker; R. Hughes; J. B. Hogenesch; H. R. Kast-Woelbern; M. E. Bowman; J. L. Ferrer;
A. M. Anisfeld; P. A. Edwards; J. M. Rosenfeld; J. G. Alvarez; J. P. Noel; K. C. Nicolaou; R. M. Evans, Mol. Cell. 2003,11,
1079–1092.
182. H. E. Pelish; N. J. Westwood; Y. Feng; T. Kirchhausen; M. D. Shair, J. Am. Chem. Soc. 2001,123, 6740–6741.
183. K. C. Nicolaou; J. A. Pfefferkorn; A. J. Roecker; G. Q. Cao; S. Barluenga, J. Am. Chem. Soc. 2000,122, 9939–9953.
184. S. N. Lo´ pez; I. A. Ramallo; M. Gonzalez Sierra; S. A. Zacchino; R. L. E. Furlan, Proc. Natl. Acad. Sci. U.S.A. 2007,104, 441–444.
185. J. Kennedy, Nat. Prod. Rep. 2008,25, 25–34.
186. U. Galm; S. Heller; S. Shapiro; M. Page; M. S. M. Li; L. Heide, Antimicrob. Agents Chemother. 2004,48, 1307–1312.
187. K. J. Weissman, Trends Biotechnol. 2007,25, 139–142.
188. J. K. Borchardt, Mod. Drug Discov. July/August 1999, 22–29.
189. Y. Volcegursky; Z. Hu; R. McDaniel, Mol. Microbiol. 2000,37, 752–762.
190. D. E. Cane; C. T. Walsh; C. Khosla, Science 1998,282, 63–68.
191. D. K. Ro; E. M. paradise; M. Ouellet; K. J. Fischer; K. L. Newman; J. M. Ndungu; K. A. Ho; R. A. Eachus; T. S. Ham; J. Kirby;
M. C. Chang; S. T. Withers; Y. Shiba; R. Sarpong; J. D. Keasling, Nature 2006,440, 940–943.
192. G. Li; X. Guo; R. Jin; Z. Wang; H. Jian; Z. Li, J. Tradit. Chin. Med. 1982,2, 125–130.
193. F. Y. Lee; R. Barzilleri; C. R. Fairchild; S. H. Kim; B. H. Long; C. Reventos-Suarez; G. D. Vite; W. C. Rose; R. A. Kramer, Clin.
Cancer Res. 2001,7, 1429–1437.
194. Quoted in S. Borman, Chem. Eng. News 14 January 2002, 23–24.
Natural Products Drug Discovery 235

Author's personal copy
Biographical Sketches
Giovanni Appendino was born in Carmagnola, Italy, in 1995. After graduating from the University
of Torino in 1979, he did post-Laurea work with Professor Pierre De Clercq (University of Gent,
Belgium), working on the total synthesis of gibberellic acids. In 1983, he became lecturer and in
1998 associated professor at his alma mater. Since 2000, he is full professor of organic chemistry at
the Universita
`del Piemonte Orientale, Faculty of Pharmacy and since 2006, chief scientific
adviser of Indena S.p.A., Milano. Professor Appendino’s research interests are in the realm of
bioactive natural products (isolation, chemical modification, and total synthesis). He has published
over 250 original articles in this area and in 1991 he received the Rho
ˆne–Poulenc Rorer Award of
the Phytochemical Society of Europe for his studies on isoprenoids.
Gabriele Fontana was born in Magenta, Italy, in 1967. After he graduated from the University
of Milano in 1992 and obtained his Ph.D. in Chemistry in 1996, he was research assistant at
the University of Newcastle Upon Tyne (UK) till 1998 under the guidance of Professor
Roger J. Griffin. He then moved to Glaxo-Wellcome, Italy, as medicinal chemistry scientist
under the direction of Dr. Romano di Fabio. In September 2000 he joined Indena SpA, Milan,
Italy, where he became head of medicinal chemistry in 2008.
Federica Pollastro was born in Novara, Italy, in 1976. After obtaining her Laurea Diploma in 2006
at the Universita
`del Piemonte Orientale, Faculty of Pharmacy, she is currently a Ph.D. student in
Professor Appendino’s group in Novara, working on the medicinal chemistry of bioactive natural
products.
236 Natural Products Drug Discovery