Rethinking production of Taxol

W
(paclitaxel) using endophyte
biotechnology

Souvik Kusari
1
, Satpal Singh
2
, and Chelliah Jayabaskaran
2
1
Institute of Environmental Research (INFU), Department of Chemistry and Chemical Biology, Chair of Environmental Chemistry
and Analytical Chemistry, TU Dortmund, Otto-Hahn-Str. 6, D-44221 Dortmund, Germany
2
Department of Biochemistry, Indian Institute of Science (IISc), Bangalore 560012, Karnataka, India
Taxol
W
(generic name paclitaxel) represents one of the
most clinically valuable natural products known to man-
kind in the recent past. More than two decades have
elapsed since the notable discovery of the rst Taxol
W
-
producing endophytic fungus, which was followed by a
plethora of reports on other endophytes possessing simi-
lar biosynthetic potential. However, industrial-scale Tax-
ol
W
production using fungal endophytes, although
seemingly promising, has not seen the light of the day.
In this opinion article, we embark on the current state of
knowledge on Taxol
W
biosynthesis focusing on the chem-
ical ecology of its producers, and ask whether it is actually
possible to produce Taxol
W
using endophyte biotechnol-
ogy. The key problems that have prevented the exploita-
tion of potent endophytic fungi by industrial bioprocesses
for sustained production of Taxol
W
are discussed.
Taxol
W
, an interesting case in point for revisiting
endophyte biotechnology
We recently highlighted the current bottlenecks in exploit-
ing a promising group of microbes called endophytic micro-
organisms (or endophytes) using red biotechnology, which
are capable of producing pharmaceutically-relevant second-
ary metabolites [1]. On the one hand, we are increasingly
gaining a deeper understanding of how endophytes engage
in bi-, tri-, and multipartite interactions with their host
plants as well as with other associated organisms (fungi,
bacteria, or viruses) andendosymbionts, under the selection
pressures of various biotic (such as pathogens and feeders)
and abiotic factors (such as precursors of plant/endophyte
secondary metabolites and environmental conditions) in
order to produce certain value-added natural products.
On the other hand, we have failed to translate these
amazing discoveries into successful industrial bioprocesses
for sustained production of desirable compounds using en-
dophyte biotechnology. We present the example of the
blockbuster drug Taxol
1
(generic name paclitaxel) to dis-
cuss the trade-offs of attempting to translate potent endo-
phytic Taxol
1
producers into industrial microbial factories.
Chemical ecology of taxane production by yew plants
Taxol
1
, belonging to a class of complex diterpenoids called
taxanes and possessing an unusual oxytane ring together
with a tricyclic core, is a blockbuster anticancer drug.
Being unique in its mode of action of halting the prolifera-
tion of cancer cells [2,3], the drug was approved by the US
FDA to treat a variety of tumors including breast, ovarian,
and AIDS-related Kaposis sarcoma, among others. It was
originally isolated from the bark of Pacic Yew, Taxus
brevifolia [4]. By and large, the production of Taxol
1
and other taxanes is conned to a narrow taxonomic group
of higher plants belonging to the genus Taxus (family
Taxaceae, syn. Coniferales). Among the other four genera
of this family, namely Amenotaxus, Autrotaxus, Pseudo-
taxus, and Torreya [5], only Autrotaxus [6] and Pseudo-
taxus [7] contain some simpler taxanes. Among the conifers
apart from the genus Taxus only two other species have
been reportedly claimed to produce taxanes. These include
a close cousin of Taxus called Cephalotaxus (Cephalotax-
aceae) [7] and also Podocarpus gracilor Pilger (Podocarpa-
ceae) [8]. With the exception of the genus Taxus, the
occurrence of this diterpenoid in other reported gymno-
sperm taxa has not been extensively studied and seems
conned to only an exceptionally limited number of species.
Interestingly, the presence of paclitaxel and other taxanes
has also been shown in an angiosperm from the family
Betulaceae, namely Corylus avellena L. [9,10]. Although
the molecular basis of taxane production in this angio-
sperm has not received adequate attention (see [11] for a
recent report on the transcriptome analysis of a Taxol
1
-
producing endophyte harboring this plant), it might imply
either an independent evolution of Taxol
1
biosynthesis in
unrelated taxonomic groups or some type of evolutionary
relatedness in the chemical ecology of these different taxa.
From the nigh-exclusive occurrence of this class of
molecules in yew and the evolutionary and taxonomic
uniqueness of Taxaceae and more importantly Taxus,
Opinion
0167-7799/
2014 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.tibtech.2014.03.011
DOI of companion article: http://dx.doi.org/10.1016/j.tibtech.2014.03.009.

This is a companion Opinion article to: Kusari, S. et al. (2014) Biotechnological

potential of plant-associated endophytic fungi: hope versus hype. Trends. Biotechnol.
(http://dx.doi.org/10.1016/j.tibtech.2014.03.009).
Corresponding authors: Kusari, S. (souvik.kusari@infu.tu-dortmund.de,
souvik.kusari@gmail.com); Jayabaskaran, C. (cjb@biochem.iisc.ernet.in).
Keywords: endophytic fungi; red biotechnology; plantmicrobe interaction; Taxol
1
;
endophyte biotechnology; industrial bioprocess; secondary metabolites; biosynthetic
pathway; Taxol
1
biosynthetic pathway; genetic engineering; co-cultivation; bio-
reactor design.
304 Trends in Biotechnology, June 2014, Vol. 32, No. 6
together with the fact that most of the secondary metabo-
lites of plants and fungi are produced as ecosystem perfor-
mance-enhancing agents [12], it is possible that before the
dispersion of the progenitor members of this genus, the
evolution of the biosynthetic module of this unique mole-
cule would already have been well underway or even
achieved. The variety of different modications to the basic
scaffold seems to stem from the reported promiscuity of
many of the determining enzymes, and from the richness
and independent evolution of cytochrome P450 monooxy-
genases (constituting the majority of pathway genes) in
geographically isolated Taxus plants.
After the huge clinical success of Taxol
1
, a number of
studies for screening the high-yielding Taxus species were
conducted at geographically and ecologically diverse sites
all over the world. The results obtained in these studies
expanded our understanding of the role of genetic and
environmental factors inuencing the type and concentra-
tion of taxanes in different Taxus species, varieties, culti-
vars, organs, and tissues of various Taxus plants. Further,
parameters like species, age of the plant/organ, and vari-
ous biotic (pathogens, herbivores, endophytes, symbionts,
etc.) and abiotic (seasonal variations, altitude of the site,
soil composition and pH, heavy metals, temperature, etc.)
factors were shown to play a signicant role in determining
the taxane prole of the Taxus plant tissues. The examples
of T. canadensis (from Canada) exhibiting a somewhat
unique chemical composition [7] and T. mairei (from South-
ern China) possessing a high content of 7-xylosyltaxanes
[13] might point to such ecosystem impositions. Interest-
ingly, many cultivars of T. canadensis have been reported
to have chemical proles different to other Taxus species
and also observed to be particularly prone to grazing by
animals [7,14]. This may indicate an intricate, yet unprov-
en, connection between chemical composition (such as type
and composition of taxanes) and susceptibility to grazing in
a particular ecological niche. Moreover, because most Tax-
us plants are protected from insect herbivory and grazing
animals, it has been rationalized that the exposure of the
plants to fungal pathogens might have necessitated the
evolution of biosyntheses of taxanes. Fleshy and colored
aril of Taxus plants, strikingly, is the only aerial organ
lacking taxanes and might, thus, represent a purposeful
exception for attracting birds for seed dispersal.
Plants respond to other ecosystem partners over time by
ne-tuning their metabolism and modifying their physiol-
ogy and biochemistry to establish and execute costbenet
relationships for optimal ecosystem performance. In this
context, approximately 500 types of taxanes reported from
different Taxus species and varieties [7] would, from an
ecological standpoint, probably allow the different Taxus
plants to engage in multitrophic interactions with other
ecosystem partners in diverse ecological niches. One such
important but rather unexplored trophic dimension is
provided by their interactions with the resident endophytic
fungi as well as with other ecosystem partners mediated by
endophytes [15]. Reports of the Taxol
1
content in different
Taxus plants and tissues correlating to the number of
endophytic fungal species highlight such an interplay
[16,17]. In addition, the secondary metabolites produced
by endophytic fungi have already been designated to be
directly or indirectly responsible for altering the host plant
defense chemistry, as well as for growth promotion, antag-
onism towards insects, herbivores, and pathogens [12,16
18]. For instance, endophytes can enhance the availability
of nitrogen in host plant tissues and apoplast, thereby
directing the overall ux towards increased/continuous
biosynthesis of bioactive alkaloids by the host plants
([18] and references therein). A number of reports on Taxus
cell cultures have shown signicantly enhanced taxane
production upon co-culturing [12] or elicitation with an
endophytic fungus, its culture, or even a puried fungal
molecule like an oligosaccharide [16,17,19].
A direct connection between Taxol
1
production by Tax-
us and an endophytic fungus was demonstrated in a recent
experimental setup that further corroborated these nd-
ings on the effect of multitrophic crosstalk on Taxol
1
production [20]. The authors demonstrated that a Tax-
ol
1
-producing Paraconiothyrium sp. induced the in planta
expression of Taxol
1
biosynthetic pathway genes of T. x
media. Interestingly, the fungus selectively inhibited
growth of Taxus-specic pathogenic fungi when challenged
with them. This indicates, as concluded by the authors
themselves, a means for the resident Taxol
1
-producing
Paraconiothyrium sp. to save its own metabolic resources
by inducing Taxol
1
production by the host to thwart its
competitors. Utilizing the same endophyte, these authors
have further demonstrated another aspect of multitrophic
interaction by uncovering the positive synergistic effect of
endophytic fungi harbored in the same plant but not
producing Taxol
1
(such as Alternaria sp. and Phomopsis
sp.) on the Taxol
1
production of Paraconiothyrium sp. [20].
Such complex interplay of ecosystem partners seems to
provide an important dimension to the production of plant
natural products such as Taxol
1
, which depends on the
species involved and their genotypes, in combination with
other higher-order-trophic interactions [18].
In the light of these ecological community interactions
together with the background information on the occur-
rence of hundreds of different taxane molecules most of
them still functionally undened it is compelling to draw
the conclusion that many, if not all, of these structures
would rather fulll still uncovered ecosystem-imposed
requirements of the individual Taxus plants. This is espe-
cially important within a costbenet context in geograph-
ically distinct and ecologically imposing habitats. The
partners involved in such multitrophic interactions could,
at least theoretically, be segregated and reassembled to
achieve their functional integration into a biotechnology-
based toolkit to positively alter the taxane content of Taxus
plants, by means of tissue or cell cultures and most deci-
sively by in vitro cultured Taxol
1
-producing endophytic
fungi. The easiest way to achieve this is, of course, nding
the inducing-organisms engaged in ecological crosstalk
with the producer or their metabolites responsible for such
a production enhancement and utilizing these in the cul-
ture and associated processes. Co-culturing of the ecolog-
ically interacting organisms has been demonstrated to this
end [12,16,17,20]. However, development of fully stream-
lined and robust biotechnological methodologies and pro-
cess technologies to achieve the desired production scales
and economics remains a daunting challenge.
Opinion Trends in Biotechnology June 2014, Vol. 32, No. 6
305
Taxol
W
-producing endophytic fungi
The huge clinical and pharmaceutical success of Taxol
1
as
an anticancer drug resulted in an enormous demand for this
drug the world over, which in turn led to its supply crisis.
This especially ensued from unprecedented yew cutting,
disappointingly low amounts of the drug in slow-growing
yew trees, and the laborious and slow process of Taxol
1
extraction, and therefore prompted the discovery of alter-
native sources of this valuable compound [21]. The elabo-
ratedtotal synthesis did not appear commerciallyviableand
semi-synthesis from renewable yew parts such as needles
was not very productive and subject to environmental,
genetic, and biological variables. Plant cell cultures of the
yew nally provided some hope and have signicantly con-
tributed to the worlds Taxol
1
supply since the 1990s [22].
However, due to the low yield and cost-related issues cou-
pled with factors like sensitivity to shear stress and lengthy
culture duration, the search for alternative microbial
sources of Taxol
1
was considered imperative.
In 1993, an endophytic fungus, Taxomyces andreanae,
was isolated from the bark of Taxus brevifolia and was
shown to produce Taxol
1
under in vitro axenic culture
conditions [23]. This discovery was projected as the dawn of
a new era of endophyte biotechnology with billions of
dollars worth of global market for Taxol
1
already in place,
and agreements were immediately underway among lead-
ing pharmaceutical companies to explore the possibility of
fungal Taxol
1
production through industrial fermentation
[21]. The inability of the fungus to show reproducible high-
titer yields of Taxol
1
in axenic cultures, thus not amenable
to industrial scale-up, led to the disappointing failure in
delivering the promises of this highly heralded discovery.
However, Taxol
1
-producing endophytic fungi harbored in
other Taxus species and even in non-Taxus plants includ-
ing many angiosperms, have been regularly reported
[21,24]. At the present time, around 200 endophytic fungi
belonging to more than 40 fungal genera from several
different orders representing mostly Ascomycota and Deu-
teromycota, with only a few from Basidiomycota and Zygo-
mycota, have been reported to produce Taxol
1
[21,24].
Many endophytic fungi are added to the list every year
underlining the fact that only a tiny fraction of an estimat-
ed one million (or more) endophytic fungal species has been
cultured and screened [25]. Undeniably, none of these
discoveries have been successfully translated into indus-
trial bioprocesses so far.
Taxol
W
production by endophytic fungi: variable,
unstable yields and ecological considerations
Numerous reports are available on the pronounced vari-
ability in Taxol
1
production (nanogram to milligram scale
per liter media) from various endophytic fungal isolates
across different batch cultures. This, combined with a
bewildering loss of production after repeated cycles of
subculturing [26] and the lack of a comprehensive under-
standing of endophyte biology, physiology, and molecular
and chemo-ecological aspects vis-a -vis their secondary me-
tabolite production, has led to disagreements in the scien-
tic community over the ability of endophytes to produce
Taxol
1
independent of their host plant [26,27]. As many of
these fungi might have a multinucleate hyphal phenotype,
it is tempting to note that, if the genes encoding compo-
nents of a particular secondary metabolite biosynthetic
pathway are lost or not maintained in adequate copy
number due to aging after repeated subculture, a compro-
mise in the target secondary metabolite biosynthesis might
be expected. On the one hand, extra-chromosomal mainte-
nance of the Taxol
1
biosynthetic pathway by these fungi, if
responsible for any production loss, could be ruled out
based on the reproducible Taxol
1
pathway gene amplica-
tions from many of them even after repeated subculturing
(Table 1). On the other hand, repeated subculturing, un-
derstandably, might alter the growth-based requirements
of a fungus coupled to its developmental program, which
might itself result in overall fungal physiological repro-
gramming and regulatory collapse of secondary metabolite
production. The reported revival of Taxol
1
production [28]
as well as the stimulatory/inductive effect of host plant
components on Taxol
1
production by some endophytic
fungi [26] point towards a need-based Taxol
1
production
scenario. All Taxol
1
-producing fungi are naturally resis-
tant to Taxol
1
, which has been shown to be a strong
fungicide against a plethora of fungal phytopathogens.
The production of Taxol
1
by endophytic fungi might thus
represent a means to thwart attack by invading fungi to
keep plants healthy for an unhindered access to their
apoplastic space. Soliman and Raizada, working with
Paraconiothyrium sp. isolated from T. x media, recently
obtained results in agreement with such a hypothesis [20].
The fungus showed higher Taxol
1
production upon treat-
ment with the host bark extracts and more importantly,
when co-cultured with one or more endophytic fungi not
capable of taxane production but isolated from the same
host plant bark. Interestingly, co-culturing with both the
Alternaria sp. and Phomopsis sp. resulted in an eightfold
increase in Taxol
1
production by the fungus, which was
almost double compared to what was obtained when co-
cultured only with Alternaria sp. Such a trophic interaction
between different fungi sharing the same ecological niche
demonstrates an ecosystem-based crosstalk for driving the
Taxol
1
production involving system-level coordination.
Molecular basis of Taxol
W
production by Taxus plants
and Taxol
W
-producing endophytic fungi
The molecular pathway of Taxol
1
biosynthesis in different
Taxus plants has been well characterized at both native
and recombinant levels with the discovery of close to 20
different enzymatic steps spatially organized in plastids,
endoplasmic reticulum, and cytosol [2931]. However, the
molecular signatures of Taxol
1
biosynthesis in Taxol
1
-
producing endophytic fungi remains largely ill-dened and
an unsolved riddle. Several groups have independently
attempted to screen many of these fungi through PCR-
based approaches to seek these biosynthetic blueprints
using primers designed from the Taxol
1
biosynthesis gene
sequences of different Taxus plants available in the
databases (homology-based approach; see Table 1)
[21,24,26,32,33]. Indeed, reports on the PCR amplication
and cloning of many genes of this pathway from several
Taxol
1
-producing endophytic fungi (Table 1) facilitate a
decisive re-evaluation of their true biosynthetic potential,
and in turn their potential as alternative and sustainable
Opinion
Trends in Biotechnology June 2014, Vol. 32, No. 6
306
sources of Taxol
1
. More than two decades have passed
since the celebrated discovery of Taxol
1
-producing T.
andreanae and no imminent breakthrough in achieving
their industrial and commercial utilization seems in sight
despite discovery and validation of a plethora of these
fungi. The aforementioned points attest to the mysterious
lifestyles of endophytes (alternating between endophyte-
pathogen-epiphyte lifestyles), their complex and varying
physiology under various environmental and culture con-
ditions, and our inadequate knowledge about their bio-
chemistry, molecular controls, and regulatory networks.
The consequence of reports that a large number of unre-
lated fungal taxa isolated from taxonomically and ecolog-
ically diverse plant species produce Taxol
1
casts doubts on
whether Taxol
1
produced by endophytes is a biosynthetic
product or an adduct (carry-over from the host plant).
However, many of these endophytic fungi that produce
Taxol
1
in vitro were isolated from taxonomically diverse
non-Taxus species that do not contain this compound, thus
negating the notion that Taxol
1
in these fungi might just
be a carry-over from the host plant tissues.
The most important clue to this apparent paradox of
endophytes producing Taxol
1
with their endogenous bio-
synthetic capabilities might lie in the very origin of Taxol
1
biosynthesis in these unique microorganisms. Horizontal
gene transfer (HGT) is often hypothesized to explain relat-
ed genes being present in taxonomically distant organisms.
However, HGT cannot account for co-opting such a complex
pathway orchestrated in different subcellular compart-
ments with genes possibly scattered over different plant
chromosomes. As discussed above, even the maintenance
of this pathway as an extra-chromosomal module would
not appear to be true. The reported differences in the plant
and endophytic fungal Taxol
1
biosynthetic genes, TS (tax-
adiene synthase) and BAPT [baccatin III 13-O-(3-amino-3-
phenylpropanoyl) transferase], further refute the HGT
theory [32]. Because TS catalyzes the so-called committed
step of this diterpenoid-derived pathway in plants
(Figure 1), the molecular basis and evolutionary origin
of this biosynthetic step could account for the generally
accepted and experimentally observed phenomenon of
Taxol
1
production by endophytic fungi only upon specic
induction. Indeed, the reported positive selection pressure-
mediated changes in TS [34,35] might point towards the
possibility of gain-of-function gene transfer events at an
early stage from a fungus to plant even involving co-
evolution, and also among different fungi with diversity
arising from gene duplication and divergence under differ-
ent habitat-imposed selection pressures.
Whether the Taxol
1
biosynthetic pathway is conserved
among plants and endophytic fungi especially with regard
to gene clustering and regulation is an interesting and
important question. If this was true, much of our current
understanding about this pathway from plants would be
applicable to Taxol
1
-producing endophytic fungi for any
sort of genetic manipulation and regulation. Reports of the
discovery of several plantTaxol
1
biosynthetic genes in
many Taxol
1
-producing endophytic fungi even hint in part
towards an overall related pathway in these fungi (Table
1). However, as is known for the biosynthesis of many
complex secondary metabolites (for instance, gibberellins),
the plant and fungal pathways might differ substantially
and might even represent convergent or parallel evolution
[11,36,37]. Perhaps not surprisingly then, the Taxol
1
bio-
synthetic pathway in plants is known to have many enzy-
matic steps, which divert the ux away from Taxol
1
[38].
By contrast, there have been contradicting reports on the
independent Taxol
1
biosynthetic capacity of endophytes
for which the presumed Taxol
1
-producing endophytes
were examined in detail [27]. The search for the possible
genetic signatures of this pathway using even state-of-the-
art tools and techniques, however, led to the conclusion
Table 1. Taxol
W
biosynthetic pathway genes reported from endophytic fungi
a,b
Gene name Molecule
type
Fungus Host plant GenBank
acc. no.
Fermentation
method
Refs
TS cDNA Fusarium solani Taxus celebica HM113487 SMF
TS gDNA Taxomyces andreanae Taxus brevifolia SMF [26]
TS gDNA Fusarium redolens Taxus baccata subsp. wallichiana SMF [32]
TS gDNA Gibberella intermedia Taxus x media KC337345 SMF [33]
TS gDNA Mucor rouxianus Taxus chinensis SMF [46]
TAT cDNA Ozonium sp. BT2 Taxus chinensis var. mairei AY960682 SMF
10bH gDNA
cDNA
Ozonium sp. BT2 Taxus chinensis var. mairei AY836677
AY907826
SMF [47]
13aH cDNA Fusarium solani Taxus celebica EF626531 SMF [48]
DBAT gDNA Fusarium solani Taxus celebica GU392264 SMF
DBAT gDNA Cladosporium cladosporoides MD2 Taxus x media EU375527 SMF [49]
DBAT gDNA Aspergillus candidus MD3 Taxus x media EU883596 SMF [50]
DBAT gDNA Fusarium redolens Taxus baccata subsp. wallichiana SMF [32]
BAPT gDNA Taxomyces andreanae Taxus brevifolia SMF [26]
BAPT gDNA Colletotrichum gloeosporioides Taxus x media KC337344 SMF [33]
BAPT gDNA Guignardia mangiferae Taxus x media KC337343 SMF [33]
BAPT gDNA Fusarium redolens Taxus baccata subsp. wallichiana KC924919 SMF [32]
a
Abbreviations: TS, taxa-4(5),11(12)-diene synthase; TAT, taxa-4(5),11(12)-diene-5a-ol-O-acetyltranseferase; T10bH, taxane-10b-hydroxylase; T13aH, taxa-4(5),11(12)-diene-
13a-hydroxylase; DBAT, 10-deacetylbaccatin III-O-acetyltransferase; BAPT, baccatin III 13-O-(3-amino-3-phenylpropanoyl) transferase; SMF, submerged fermentation;
cDNA, complementary DNA; gDNA, genomic DNA.
b
See also [11] for some recent putative homologues of this pathway froma PenicilliumaurantiogriseumNRRL 62431 strain isolated fromhazelnut plant where extensive in
silico analysis of the fungal genome sequence was performed.
Opinion Trends in Biotechnology June 2014, Vol. 32, No. 6
307
that no such pathway exists in these fungi. Interestingly,
more recent ndings based on genome mining [11] revealed
that such a scenario does not seem to be true.
Concluding remarks and future perspectives
Genetic engineering of endophytic fungi known to produce
Taxol
1
, both by gene overexpression and random muta-
genesis coupled with genome shufing, have been
attempted in only a very limited number of fungal isolates.
In Ozonium sp. EFY-21 isolated from T. chinensis var.
mairei, overexpression of Taxus TS gene under a fungal
specic promoter resulted in about vefold increase in
Taxol
1
production as compared to control [39]. Multiple
mutagenesis of Nodulisporum sylviforme provided the
strain NCEU-1 from which three hereditarily stable
strains were obtained by mutagenesis. Protoplasts (round
fungal cells generated from spores and lacking the cell
wall) generated from these and fused randomly nally led
to three strains that showed an increase in Taxol
1
yield by
31, 64, and 45% over the control, respectively [40]. The
reported Taxol
1
pathway metabolic engineering
approaches in Escherichia coli [4143] and Saccharomyces
cerevisiae [44,45] have mostly focused on taxadiene engi-
neering. Reported attempts to engineer the Taxol
1
biosyn-
thetic pathway beyond taxadiene encountered metabolic
bottlenecks as observed by total absence or insignicant
yields of any intermediate beyond taxadiene. This was
shown for seven consecutive gene transfers in S. cereviseae,
which is the highest number of steps (for the Taxol
1
pathway) engineered in a heterologous host so far [45].
Notwithstanding these unsuccessful endeavors, however,
the several hundred milligrams per liter (reaching 1 g/l)
yields of taxadiene obtained in few such attempts together
with reports of biotransformation of intermediate taxanes
by several microbial enzymes provide some strategies
worth exploring to realize sustained Taxol
1
supply using
endophyte (and related microbial) biotechnology. This
would be especially interesting when supplemented with
MVA/MEP Pathway
Paclitaxel
OPP
H
H
OH
H
H
H
HO
HO
HO
HO
AcO
AcO
AcO
OH
OH
O
O
O
O
O
O
O
O AcO
H
O
O
O
NH
OH
OH
OH
AcO
O
O
O
O
O
O
H
OH
OH
AcO
AcO
OH
OH
OCcA
NH
2
NH
2
NH
2
NH
2
O
O
O
H HO
O
O
O
O
OH
H
H
H
CAc
CAc
H
OPP
OPP
OPP
Isopenteyldiphosphate
Taxa-4(5),11(12)-diene
Taxa-4(5),11(12)-diene-5-ol
Taxa-4(5),11(12)-diene-5-acetate
Taxa-4(5),11(12)-diene-10-ol-5-acetate
Farnesyldiphosphate
Isopenteyldiphosphate
Geranylgeranyldiphosphate
10-Deacetylbaccan III Baccan III
-Phenylalanine CoA
-Phenylalanine
3-N-debenzoyl-2-deoxypaclitaxel
-Phenylalanine
GGPPS
T10H
TAT
DBTNBT
PAM
*
T5H
DBAT
BAPT
TS
+

TRENDS in Biotechnology
Figure 1. Prevalent consensus biosynthetic route for Taxol
1
in Taxus species. Abbreviations: MVA, mevalonic acid; MEP, 2-C-methyl-D-erythritol-4-phosphate; GGPPS,
geranylgeranyldiphosphate synthase; TS, taxa-4(5),11(12)-diene synthase that catalyzes the committed step of this pathway; T5aH, taxa-4(5),11(12)-diene-5a-hydroxylase;
TAT, taxa-4(5),11(12)-diene-5a-ol-O-acetyltranseferase; T10bH, taxane-10b-hydroxylase; z, oxytane ring formation and branch migration enzymes including taxane 2a-O-
benzoyltransferase (T2BT or DBBT = debenzoyltaxane-2
0
-a-O-benzoyltransferase) as well as C-13 hydroxylation and steps taking pathway flux towards non-Taxol
1
-type
molecules; DBAT, 10-deacetylbaccatin III-O-acetyltransferase; BAPT, baccatin III 13-O-(3-amino-3-phenylpropanoyl) transferase; DBTNBT, 3
0
-N-debenzoyl-2
0
-deoxytaxol-N-
benzoyltransferase which follows hydroxylation in the side chain by an unknown enzyme; PAM, phenylalanineaminomutase; *, b-phenylalanine coenzyme A ligase.
Multiple arrows imply more than one biosynthetic step. The Taxol
1
biosynthetic pathway is proposed to have about 20 different enzymatic steps in Taxus plants.
Opinion
Trends in Biotechnology June 2014, Vol. 32, No. 6
308
Box 1. Sustained Taxol
W
supply using endophytes
Genetic engineering approaches can be used to engineer the Taxol
1
biosynthetic pathway (Figure IA). This includes overexpression of the
important or all of the Taxol
1
biosynthetic pathway genes and their
promoter modulation, including other complimentary genetic engi-
neering approaches such as epigenetic engineering involving tran-
scription factors or chromatin modifier elements, and gene silencing.
Taxol
1
-producing endophytes and other heterologous hosts including
simpler organisms such as yeast and bacteria could serve as ideal
candidates for such an attempt. Taking advantage of inter-organismal
effects could also lead to greater Taxol
1
/taxane yields (Figure IB). This
strategy may involve co-cultivation of two or more organisms for
elicitation of Taxol
1
biosynthesis, Taxol
1
pathway-intermediate bio-
transformation, or utilization of metabolic dead ends/high yield Taxol
1
pathway-intermediates fromone organismby another organismas the
exogenously supplied substrates, and even combinatorial Taxol
1
/
taxane biosynthesis by different organisms. Various molecular-,
biotechnological-, and bioprocess-related methodologies, tools, tech-
niques, and optimization strategies, both stand-alone or in combina-
tion, are known or proposed to affect the Taxol
1
/taxane yields of a
given endophyte or a recombinant heterologous host (Figure IC). These
would seem to prove especially relevant when used in combination
with the genetic engineering and inter-organismal combinatorial
approaches. For example, a heterologous host engineered for the
Taxol
1
biosynthetic pathway gene(s) could be optimized for an
inducible Taxol
1
production, a greater cellular release of Taxol
1
,
two-phase growth and production cycles, and silencing or modulation
of its own Taxol
1
pathway negative feedback, flux-diversionary, or
dead-end metabolite metabolic steps.
(A)
TS GGPPS T5H TAT T10H DBAT PAM etc.
Metabolic engineering of Taxol

biosynthec pathway
Heterologous hosts
Flux balance (pathway and protein engineering)
Enzyme engineering
Co-culture and culture opmizaon
Bioreactor design
Removal of metabolic bolenecks/dead ends
Strain improvement (gene overexpression,
mutagenesis, genome shuing. gene silencing,
engineering, and physiochemical opmizaon
for increased product excreon)
Superior Taxol

yields ???
Endophyc fungi
(B)
(C)
d i
TRENDS in Biotechnology
Figure I. Possible strategies for achieving sustained Taxol
1
supply using endophytes. (A) Genetic engineering approaches involving the Taxol
1
biosynthetic pathway.
(B) Inter-organismal contribution towards greater Taxol
1
/taxane yields. (C) Bioprocessing-based strategies for increasing Taxol
1
yield.
Opinion Trends in Biotechnology June 2014, Vol. 32, No. 6
309
contributions from heterologous hosts and other optimiza-
tion methodologies and tools such as intracellular com-
partment optimization, storage and efux modulation, and
control of pathway regulatory elements (Box 1). However,
delineation of the molecular mechanisms of Taxol
1
bio-
synthesis and regulation thereof remains a prerequisite for
all such endeavors. Most notably, as seen from the Taxol
1
biosynthetic pathway of Taxus sp., there seems to be an
obvious hurdle in engineering such a lengthy and complex
pathway in its entirety in heterologous hosts. Transforma-
tion of genes of the entire pathway is a challenge and more
importantly, regulation of Taxol
1
production encompass-
ing epigenetic modulation and signaling crosstalk itself
remains a poorly understood topic. Taxol
1
-producing en-
dophytic fungi, therefore, still present a viable and long-
term target, despite many unanswered questions (Box 2).
Acknowledgments
Research in the laboratory of S.K. (INFU, TUDortmund) is supported in
part by the International Bureau (IB) of the German Federal Ministry of
Education and Research (BMBF/DLR), Germany, the Ministry of
Innovation, Science, Research, and Technology of the State of North
Rhine-Westphalia, Germany, the German Academic Exchange Service
(DAAD; Welcome to Africa initiative), and the German Research
Foundation (Deutsche Forschungsgemeinschaft, DFG). S.K. is a Visiting
Researcher at the Department of Plant Sciences, University of Oxford,
South Parks Road, Oxford OX1 15 3RB, United Kingdom. S.K. gratefully
acknowledges M. Spiteller for approving and authorizing, Gail M.
Preston for hosting, and TU Dortmund for supporting his stay at the
University of Oxford. Research in the C.J. laboratory (IISc Bangalore) is
supported by grants fromthe Department of Biotechnology (DBT), India
and the Council of Scientic & Industrial Research (CSIR), India. S.S.
thanks CSIR and DBT for fellowships. We thank Bhagat Singh for
assistance in preparation of Figure 1. We apologize to the numerous
investigators whose publications could not be cited here owing to space
constraints.
References
1 Kusari, S. et al. (2014) Biotechnological potential of plant-associated
endophytic fungi: hope versus hype. Trends. Biotechnol. 32, http://
dx.doi.org/10.1016/j.tibtech.2014.03.009
2 Schiff, P.B. et al. (1979) Promotion of microtubule assembly in vitro by
Taxol. Nature 277, 665667
3 Schiff, P.B. and Horwitz, S.B. (1980) Taxol stabilizes microtubules in
mouse broblast cells. Proc. Natl. Acad. Sci. U.S.A. 77, 15611565
4 Wani, M.C. et al. (1971) Plant antitumor agents. VI. The isolation and
structure of taxol, a novel antileukemic and antitumor agent from
Taxus brevifolia. J. Am. Chem. Soc. 93, 23252327
5 Cope, E.A. (1998) Taxaceae: the genera and cultivated species. Bot.
Rev. 64, 291322
6 Gueritte-Voegelein, F. et al. (1987) Taxol and derivatives: a biogenetic
hypothesis. J. Nat. Prod. 50, 918
7 Wang, Y-F. et al. (2011) Natural taxanes: developments since 1928.
Chem. Rev. 111, 76527709
8 Stahlhut, R. et al. (1999) The occurrence of antitumor diterpene Taxol
in Podocarpus gracilor Pilger (Podocarpaceae). Biochem. Syst. Ecol. 27,
613622
9 Service, R. (2000) Hazel tree offers a newsource of cancer drug. Science
288, 2728
10 Qaderi, A. et al. (2012) Hazel (Corylus avellana L.) as a new source of
Taxol and taxanes. J. Med. Plants 11, 6677
11 Yang, Y. et al. (2014) Genome sequencing and analysis of the paclitaxel-
producing endophytic fungus Penicillium aurantiogriseum NRRL
62431. BMC Genomics 15, 69
12 Li, Y-C. et al. (2009) Paclitaxel production using co-culture of Taxus
suspension cells and paclitaxel-producing endophytic fungi in a co-
bioreactor. Appl. Microbiol. Biotechnol. 83, 233239
13 Hao, D.C. (2008) Phylogenetic relationships of the genus Taxus
inferred from chloroplast intergenic spacer and nuclear coding DNA.
Biol. Pharm. Bull. 31, 260265
14 Windels, S.K. and Flaspohler, D.J. (2011) The ecology of Canada Yew
(Taxus canadensis Marsh.): a review. Botany 89, 117
15 Kusari, S. et al. (2013) Untapped mutualistic paradigms linking host
plant and endophytic fungal production of similar bioactive secondary
metabolites. Phytochemistry 91, 8187
16 Soliman, S.S.M. et al. (2013) A fungal endophyte induces transcription
of genes encoding a redundant fungicide pathway in its host plant.
BMC Plant Biol. 13, 93
17 Li, Y-C. and Tao, W-Y. (2009) Interactions of Taxol-producing
endophytic fungus with its host (Taxus spp.) during Taxol
accumulation. Cell Biol. Int. 33, 106112
18 Hartley, S.E. and Gange, A.C. (2009) Impacts of plant symbiotic fungi
on herbivores: mutualism in a multitrophic context. Annu. Rev.
Entomol. 54, 323342
19 Yuan, Y-J. et al. (2001) Signal transduction pathway for oxidative burst
and taxol production in suspension cultures of Taxus chinensis var.
mairei induced by oligosaccharide from Fusarium oxysprum. Enzyme
Microb. Technol. 29, 372379
20 Soliman, S.S.M. and Raizada, M.N. (2013) Interactions between co-
inhabiting fungi elicit synthesis of Taxol froman endophytic fungus in
host Taxus plants. Front. Microbiol. 4, http://dx.doi.org/10.3389/
fmicb.2013.00003
21 Flores-Bustamante, Z.R. et al. (2010) Microbial paclitaxel: advances
and perspectives. J. Antibiot. (Tokyo) 63, 460467
22 Christen, A.A. et al. (1989) Cell cultures as a means to produce Taxol.
Proc. Am. Assoc. Cancer Res. 30, 566
23 Stierle, A. et al. (1993) Taxol and taxane production by Taxomyces
andreaenae, an endophytic fungus of Pacic yew. Science 260,
214216
24 Hao, X. et al. (2013) Taxol producing fungi. In Natural Products
(Ramawat, K.G. and Merillon, J.M., eds), pp. 27972812, Springer-
Verlag
25 Suryanarayanan, T.S. et al. (2009) Fungal endophytes and
bioprospecting. Fungal Biol. Rev. 23, 919
26 Staniek, A. et al. (2009) Taxomyces andreanae: a presumed Paclitaxel
producer demystied. Planta Med. 75, 15611566
27 Heinig, U. et al. (2013) Getting to the bottom of Taxol biosynthesis by
fungi. Fungal Divers. 60, 161170
28 Li, J.Y. et al. (1998) Stimulation of Taxol production in liquid culture of
Pestalotiopsis microspora. Mycol. Res. 102, 461464
29 Jennewein, S. and Croteau, R. (2001) Taxol biosynthesis: molecular
genetics and biotechnological applications. Appl. Microbiol. Biotechnol.
57, 1319
30 Walker, K. and Croteau, R. (2001) Taxol biosynthetic genes.
Phytochemistry 58, 17
31 Croteau, R.B. et al. (2006) Taxol biosynthesis and molecular genetics.
Phytochem. Rev. 5, 7597
32 Garyali, S. et al. (2013) Taxol production by an endophytic fungus,
Fusarium redolens, isolated from Himalayan Yew. J. Microbiol.
Biotechnol. 23, 13721380
33 Xiong, Z-Q. et al. (2013) Diversity of endophytic fungi and screening of
fungal paclitaxel producer from Anglojap yew, Taxus x media. BMC
Microbiol. 13, 71
Box 2. Outstanding questions
Is the Taxol
1
biosynthesis in endophytic fungi similar to that of
Taxus plants and what is its origin?
Why do endophytic fungi tend to lose their Taxol
1
production
potential upon extensive propagation?
Do these endophytic fungi produce Taxol
1
in planta and, if so,
how does that differ fromtheir in vitro production, especially with
regard to its biosynthetic regulation?
What is the molecular basis of the reported elicitation of
endophyte Taxol
1
biosynthesis by host plant components?
Do the host plant Taxol
1
pathway intermediates play a part in
fungal Taxol
1
biosynthesis?
How much is the storage, release, and compartmentalization of
Taxol
1
biosynthesis in fungi affected by metabolic flux dynamics
connecting primary and secondary metabolism or vice versa?
Opinion
Trends in Biotechnology June 2014, Vol. 32, No. 6
310
34 Hao, D.C. et al. (2009) Molecular evolution of paclitaxel biosynthetic
genes TS and DBAT of Taxus species. Genetica 135, 123135
35 Koksal, M. et al. (2011) Taxadiene synthase structure and evolution of
modular architecture in terpene biosynthesis. Nature 469, 116120
36 Tudzynski, B. (2005) Gibberellin biosynthesis in fungi: genes, enzymes,
evolution, and impact on biotechnology. Appl. Microbiol. Biotechnol.
66, 597611
37 Hedden, P. et al. (2001) Gibberellin biosynthesis in plants and fungi: a
case of convergent evolution? J. Plant Growth Regul. 20, 319331
38 Frense, D. (2007) Taxanes: perspectives for biotechnological
production. Appl. Microbiol. Biotechnol. 73, 12331240
39 Wei, Y. et al. (2012) Engineering Taxol biosynthetic pathway for
improving Taxol yield in Taxol-producing endophytic fungus EFY-21
(Ozonium sp.). Afr. J. Biotechnol. 11, 90949101
40 Kai, Z. et al. (2008) Screening and breeding of high Taxol producing
fungi by genome shufing. Sci. China C: Life Sci. 51, 222231
41 Ajikuar, P.K. et al. (2010) Isoprenoid pathway optimization for Taxol
precursor overproduction in Escherichia coli. Science 330, 7073
42 Huang, Q. et al. (2001) Engineering Escherichia coli for the synthesis of
Taxadiene, a key intermediate in the biosynthesis of Taxol. Biorg. Med.
Chem. 9, 22372242
43 Meng, H. et al. (2011) In silico analysis and experimental improvement
of taxadiene heterologous biosynthesis in Escherichia coli. Biotechnol.
Bioprocess Eng. 16, 205215
44 Engels, B. et al. (2008) Metabolic engineering of taxadiene biosynthesis
in yeast as a rst step towards Taxol (paclitaxel) production. Metab.
Eng. 10, 201206
45 DeJong, J.H.M. et al. (2005) Genetic engineering of Taxol biosynthetic
genes in Saccharamyces cereviseae. Biotechnol. Bioeng. 93, 221224
46 Miao, Z. et al. (2009) Newendophytic taxane production fungus fromT.
chinensis. Appl. Biochem. Microbiol. 45, 8186
47 Guo, B.H. et al. (2006) An endophytic Taxol-producing fungus BT2
isolated from Taxus chinensis var. mairei. Afr. J. Biotechnol. 5,
875877
48 Chakravarthi, B.V.S.K. et al. (2008) Production of paclitaxel by
Fusarium solani isolated from Taxus celebica. J. Biosci. 33, 259267
49 Zhang, P. et al. (2009) An endophytic Taxol-producing fungus from
Taxus media, Cladosporiumcladosporioides MD2. Curr. Microbiol. 59,
227232
50 Zhang, P. et al. (2009) An endophytic Taxol-producing fungus from
Taxus x media, Aspergillus candidus MD3. FEMS Microbiol. Lett. 293,
155159
Opinion Trends in Biotechnology June 2014, Vol. 32, No. 6
311