A comprehensive and engaging overview of the type III family of polyketide synthases

A comprehensive and engaging overview of the type III family ofpolyketide synthasesKenji Watanabe1, Alex P Praseuth1 and Clay CC Wang1,2

Customizing biosynthesis of naturalproducts to yield biologically

active derivatives has captivated scientists in the field of

biosynthetic research. To substantiate this goal, there are scores

of obstacles to consider. To create novel metabolites by

mutating amino acid residues in wild-typeenzymes,a researcher

must broaden the range of the enzymes substrate tolerance and

increase its turnover rate during reaction catalysis. In the past

decade, numerous gene clusters responsible for the

biosynthesis of notable natural products have been identified

from a variety of organisms. Several genes coding for type III

polyketide synthases, particularly the chalcone synthase

superfamilyenzymes, were recently uncovered and expressed in

E. coli. Furthermore, it was observed and reported how these

recombinant enzymes are capable of producing essential

metabolites in vitro. Three of the type III polyketide synthases,

chalcone synthase, octaketide synthase and pentaketide

chromone synthase, have been characterized and their active

sites subjected to rational engineering for biosynthetic

production of their analogs.Because theyareencoded ina single

open reading frame and are post-translationally small in size,

type III polyketide synthases are ideal targets for protein

engineering. The relative ease with which these genes are

expressed makes molecular biological manipulation to obtain

mutated enzymes more procurable, ameliorating analysis of its

biosynthetic pathway. In summary, time devoted to modification

of biosynthetic proteins and unravelling of the detailed reaction

mechanisms involved in biosynthesis will be shortened, paving

the way for a much wider scope for metabolic engineers in future.

This review focuses on the use of chalcone synthase, octaketide

synthase and pentaketide chromone synthase for rational

biosynthetic engineering to generate molecular diversity and

pursue innovative, biologically potent compounds.

Addresses1 Department of Pharmacology and Pharmaceutical Sciences,

University of Southern California 1985 Zonal Ave PSC 718 Los Angeles,

California 90033, USA2 Department of Chemistry, University of Southern California 1985

Zonal Ave PSC 718 Los Angeles, California 90033, USA

Corresponding author: Watanabe, Kenji ([email protected]);

Wang, Clay CC ([email protected])

Current Opinion in Chemical Biology 2007, 11:279–286

This review comes from a themed issue on

Molecular diversity

Edited by Gregory A Weiss and Richard Roberts

Available online 7th May 2007

1367-5931/$ – see front matter

# 2006 Elsevier Ltd. All rights reserved.

DOI 10.1016/j.cbpa.2006.11.041

www.sciencedirect.com

Introduction
Varying structurally, polyketides are secondary metab-
olites produced by polyketide synthases (PKSs) encoded

in the chromosomes of a variety of organisms including

bacteria, fungi, marine organisms and plants. These

natural products are valuable antibiotics, anticancer drugs

and immunosuppressants. Recently, researchers have

focused much attention on the ability of PKSs to syn-

thesize complex chemical scaffolds, and the possibility of

engineering them to produce new antimicrobial agents

has been contemplated. Depending on the overall

approach, PKSs are categorized into three classes: types

I, II and III. All three PKS classes are easily distinguished

by their physical composition. Type I is recognized for

the size of their multi-domain enzymes (Figure 1a) and

type II for its many mono-functional subunits (Figure 1b),

respectively, whereas type III PKSs consist of only a

single protein (Figure 2), with a molecular mass ranging

between 80 to 90 kDa in their native state of a homo-

dimeric form acting in an iterative fashion [1��].

In past reports, natural products chemists were able to

isolate several type III polyketide scaffolds, such as

chalcone from plants, and studies have identified the

chalcone synthase (CHS) as an enzyme catalyzing one

of the many steps in the flavonoid biosynthetic pathway

[1��]. These secondary metabolites commonly found in

plants contain polyphenol substituents, affording them

the ability to act as free radical scavengers or a phytoa-

lexin chemical defense against pathogen attack and other

external stressors [2]. Along with these studies, type III

polyketides have also exhibited antimicrobial activity

against a spectrum of parasitic microorganisms in plants.

Plants generally carry these biosynthetic pathways, which

have evolved over time for self-preservation. These

bioactive compounds are effective chemical weapons

for defense against any assault from microbial foes or

competing foliage within its vicinity.

Ascertained by Noel’s group in 1999, the crystal structure

of CHS from alfalfa authenticated the existence of a type

III PKS and provided valuable information regarding its

structure [3]. This finding gave rise to the field of bio-

synthetic engineering with a focus on this family of

enzymes [4]. Using the crystal structure of CHS as a

template, Noel and colleagues were able to compare and

ascertain the active site of stilbene synthase (STS) for

successful site-directed mutagenesis, rationally changing

out key amino acid residues within the pocket of this

equally functional enzyme. These findings have shed

light on the detailed reaction mechanism involved in


mailto:[email protected]

mailto:[email protected]

http://dx.doi.org/10.1016/j.cbpa.2006.11.041

280 Combinatorial chemistry and molecular diversity

Figure 1

Organization of type I and type II polyketide synthases. Biosynthesis of (a) 6-deoxyerythronolide B (erythromycin aglycone) and (b) tetracenomycin

F1 by type I and type II polyketide synthases, respectively. Abbreviations for proteins or catalytic domains are as follows: ACP, acyl carrier protein;

AT, acyltransferase; DEBS, 6-deoxyerythronolide B synthase; DH, dehydratase; ER, enoylreductase; KR, ketoreductase; KS, ketosynthase;

LD, loading domain; M, module; MAT, malonyl-CoA ACP acyltransferase; TE, thioesterase.

the assembly of this class of natural products. Such find-

ings have also paved the way for the field of natural

products engineering by way of introducing a simple

mutation for in vitro conversion.

Further data was gathered when two additional gene

clusters encoding type III PKSs, octaketide synthase

(OKS) and pentaketide chromone synthase (PCS), were

uncovered by Abe et al. from plants traditionally used for

medicinal purposes, Aloe arborescens [5�,6��]. Both

proteins share about 60% amino acid sequence homology

with CHS and 92% sequence homology with each other.

Exploiting the general homology of type III PKSs, Abe

and colleagues selected amino acid residues in the active

site of OKS that were comparable to those found in CHS


to demonstrate how a simple mutation of its polyketide

biosynthetic machinery can control the chain length of

the end product. Given the ease of modifying a relatively

short open reading frame that codes for type III PKS and

taking advantage of readily available molecular biological

techniques, there has been mounting interest from inves-

tigators for this novel but rapidly developing science.

Moreover, the availability of structural information on

CHS reveals substantial amino acid sequence homology

with its counterpart, compelling researchers to assess the

limitations in which biologically active compounds and

unnatural natural compounds are biosynthesized when

using CHS as the exemplary enzyme. Focus and empha-

sis will be placed on generating molecularly diverse plant

type III polyketides through rational engineering of the


A comprehensive and engaging overview of the type III family of polyketide synthases Watanabe, Praseuth and Wang 281

Figure 2

Chemical structures and biosynthetic schematics of type III polyketides, (a) triacetic acid lactone, (b) naringenin chalcone, (c) resveratrol and

(d) benzalacetone.

CHS and OKS active sites. Innovative approaches to

engineering these remarkable proteins to broaden and

strengthen the potential diversity of such compounds will

also be explored.

Diverse schematics for the assembly of typeIII polyketidesFrom previous studies, type III PKSs are known to accept

a variety of CoA thioester molecules for use as starting

units (acetyl-, p-coumaroyl and isovaleroyl-CoA) and cat-

alyze several elongation reactions by loading acyl-CoA as

extender units [1��]. To date, type III PKSs possess four

enzyme releasing mechanisms where the polyketide

chain is terminated. These enzymes are capable of cat-

alyzing linear chain elongation followed by three cyclor-

elease avenues, thereby generating a compound

containing a 6-membered ring. Examples include 2-pyr-

one synthase (2-PS) [7], which produces triacetic acid

lactone, while intramolecular Claisen and aldol conden-

sation mechanisms are employed by CHS and STS to

produce naringenin chalcone and resveratrol, respectively

(Figure 2a–c) [8–12]. Additionally, decarboxylation has

been implicated in polyketide chain termination by ben-


zalacetone synthase (BAS) and is exclusive of cyclization

for the release of benzalacetone (Figure 2d) [13]. From

the above description, these enzymes from this super-

family show evidence for potential divergence of type III

polyketides. Using the diversity of this class of com-

pounds, engineers can benefit and logically generate

additional compounds with potentially improved bioac-

tivity. During the past few years, engineering of biosyn-

thetic pathway for production of more diverse and

controlled architectures of this series of compounds has

been attempted, aiding chemists in further understanding

the reaction mechanisms employed by these enzymes

[14,15��].

Engineering of CHS to incite divergence oftype III polyketidesThe biosynthesis of naringenin chalcone by CHS is

initiated by binding of p-coumaroyl-CoA in the active

pocket followed by three condensation reactions, where

p-coumaroyl-CoA is elongated at the expense of three

malonyl-CoAs, and a decarboxylation reaction following

each step [16]. More notably, this enzyme is capable of

catalyzing a cyclization reaction of a linear polyketide



intermediate via an intramolecular Claisen condensation

reaction yielding naringenin chalcones where STS

employs an aldol condensation reaction for cyclorelease

of resveratrol. Considering these notable differences and

following extensive inquiry of the active site of CHS,

Noel et al. recognized the potential for functionally

engineering CHS and endowing this enzyme with the

equivalent catalytic capability of STS via site-directed

mutagenesis [4]. They targeted suspect amino acid resi-

dues in the catalytic cavity of CHS and made the follow-

ing substitutions: Val98 to Leu (V98L), Thr131 to Ser

(T131S), Ser133 to Thr (S133T), Gly134 to Thr (G134T),

Val135 to Pro (V135P), Met137 to Leu (M137L), Met158

to Gly (M158G) and Tyr160 to Phe (Y160F). These

residues were judiciously replaced with amino acid resi-

dues identified in the active site of STS. Empirically,

engineering the active site of CHS with residues to

epitomize the active site of STS produced resveratrol

in place of naringenin chalcone, demonstrating the new

Figure 3

Product specificity and modelled structures of the active sites in (a) wild-typ


inclination of CHS toward aldo condensation contrary to

the wild-type enzyme. Markedly, kinetic studies using

p-coumaroyl-CoA as a substrate provided a comparable

kcat value of 0.14 min�1 and a Km value of 1.1 mM, which

is more than double that of wild-type STS (0.12 min�1,

0.48 mM). Despite the decline of substrate specificity for

p-coumaroyl-CoA, it was evident that engineering of CHS

produced the expected compound by means of simple

mutagenesis technology.

By mutating three amino acid residues in the catalytic

pocket of CHS (T197L, G256L and S338I), a six-

membered lactone ring was generated and production

of TAL in place of naringenin chalcone when supplied

with acetyl- and malonyl-CoA as the starter and building

unit was observed [17]. By supplying either p-coumaroyl-

CoA or acetyl-CoA as the starter unit for functional

assessment of amino acid sequence divergence in the

active sites of CHS and 2-PS, Noel’s team was able to

e OKS and (b) mutated OKSs.



locate the above three amino acid residues for targeted

mutagenesis and catalytic conversion of CHS to mimic

2-PS. Comparing the crystal structure of CHS with 2-PS

complexed with the reaction intermediate acetoacetyl-

CoA and the triple mutant, it was clear that p-coumaroyl-

CoA was not its inherent starter unit. In fact, engineering

of the protein physically impoverished the catalytic

pocket of the mutagenic enzyme for accepting p-coumar-

oyl-CoA as a starting unit. Furthermore, both the mutated

enzyme and 2-PS exhibited an indistinguishable kinetic

profile. As a result, mutation of CHS could provide a

functional enzyme indistinguishable to 2-PS, allowing

researchers the opportunity to generate a compelling

divergence in this class of compounds.

Controlling chain length of type III polyketidesby means of biosynthetic engineeringIn this section, we describe biosynthetic engineering for

products with preferential chain length through use of

mutated OKS, PCS and CHS. The degree of difficulty

that is involved in synthesizing and catalyzing a carbon-

carbon bond forming reaction for the assembly of a

Figure 4

Product specificity, crystal and modelled structure of the active sites in (a) w

in the text are numbered. The figure of PCS crystal structure was prepared


complex molecule is well recognized by chemists who

work in the field. A contemporary approach to alleviate

such a challenge is to enlist enzymes that can simply and

efficiently catalyze such formation reactions and produce

complex macromolecular products naturally. The ability

to manipulate such pathways and control the number of

elongation steps implicated in producing a natural poly-

ketide product will assuage scientists’ efforts to create

novel and diverse therapeutic candidates.

Obtained from its crystal structure, the amino acid

sequence of the active site of CHS was exploited for

alignment analysis of CHS, OKS and PCS. Abe et al.rationalized and imposed a single point mutation of an

amino acid residue at position 197 for each of the three

enzymes with various amino acid residues at that location:

T197 for CHS, G197 for OKS and M197 for PCS [5�,6��].Abe et al. demonstrated how a single point mutation in the

active site of OKS can influence the chain length of the

secondary metabolite of OKS. While the wild-type

enzyme is known to biosynthesize two octaketides at

the cost of eight malonyl-CoAs (Figure 3a), the following

ild-type PCS and (b) mutated PCS. The catalytic residues described

by using information deposited in the Protein Data Bank (2D3M).



polyketides were observed in vitro when OKS was sub-

jected to the indicated amino acid substitutions: triketide

(G197W), tetraketide (G197W), pentaketide (G197M),

hexaketide (G197T) and heptaketide (G197A)

(Figure 3b). The two octaketides, SEK4 and 4b, were

isolated when genes encoding for the essential protein

component of actinorhodin type II PKS were expressed

[18,19]. Known to turnover five molecules of malonyl-

CoA and yield a pentaketide chromone (Figure 4a), wild-

type PCS was a candidate for biosynthetic engineering

and subjected to a single point mutation (M197G) exhib-

ited the ability to biosynthesize SEK4 and 4b (Figure 4b).

These findings suggest how the amino acid residue at

position 197 in the active site could govern the chain

length of the polyketide. The space dictated by the

bulkiness of the amino acid (glycine versus alanine) in

the catalytic pocket at the noted position can either

provide or decrease indispensable room for additional

Figure 5

Product specificity, crystal and modelled structure of the active sites in (a) w

in the text are numbered. The figure of CHS crystal structure was prepared


elongation steps, determining the size of the secondary

metabolite of the macroenzyme.

A leucine at position 256 in the active site for both OKS

and PCS could influence the substrate preference for

malonyl-CoA as a starting unit, while a glycine residue

located in the same position and found in the catalytic

pocket of CHScould possibly compel the enzyme to

readily accept p-coumaroyl-CoA as a starting unit

[20,21] (Figure 5a). Furthermore, studies of the crystal

structure of CHS have implicated serine 338 as playing a

role in dictating the chain elongation threshold. Exchange

of this amino acid with valine in the active sites of OKS

and PCS substantiated this hypothesis. Accordingly, CHS

containing a double point mutation (G256L/S338V) not

only accepted p-coumaroyl-CoA but also biosynthesized

octaketides SEK4 and 4b, exhausting eight malonyl-

CoAs. Moreover, when subjected to a triple mutation

ild-type CHS and (b) mutated CHS. The catalytic residues described

by using information deposited in the Protein Data Bank (G256F).



(T197G/G256L/S338V), CHS production of both octake-

tides increased considerably [22�] (Figure 5b). These

remarkable results, exhibited by enzymes of this super-

family, could provide a divergence for varying the chain

length of type III polyketide scaffolds by means of a

simple mutation.

Conclusions and outlookDuring the past decade, significant progress has been

achieved by researchers in understanding the fundamen-

tal aspects of secondary metabolite production involving

type III polyketide biosynthetic mechanisms. Unravel-

ling the crystal structure of CHS has fostered increasing

interest in this field of research and researchers are now

making attempts at deciphering the reaction mechanisms

executed by this unique family of enzymes [23]. From

the putative reaction mechanism and exploiting the

straightforwardness of modifying a relatively short, single

open reading frame coding for type III PKS, rationally

engineering this family of proteins has successfully pro-

vided a divergence of type III polyketide scaffolds.

Therefore, it will be possible to generate more diverse

and novel compounds with desirable pharmacological

profiles through rational engineering of their biosynthetic

machinery.

Future work in this field should focus on improving the

productivity of these synthetic enzymes. Results

described in this review were obtained from in vitroexperimental work, meaning productivity was highly

dependant on protein activity, in particular the kcat value

of the enzyme. However, even wild-type PKSs such as

CHS, STS, OKS and PCS, which we mentioned in this

review, exhibit relatively low kcat values. To use these

enzymes as biosynthetic tools for expanding the diversity

and efficient production of more noteworthy molecules, it

will be necessary to construct a higher turnover enzyme.

Such an enzyme could be realized by utilizing the method

of directed evolution to increase its kcat value. Another

approach is to establish an in vivo biosynthetic system

capable of stabilizing its activity for long-term production,

which will necessitate transplantation of either or both

starting and building unit biosynthetic pathways into a

heterologous expression cell-line. These challenges, if

accomplished, will arm us with innovational means of

generating molecularly diverse type III polyketides.

References and recommended readingPapers of particular interest, published within the period of review,have been highlighted as:

� of special interest�� of outstanding interest

1.��

Austin MB, Noel JP: The chalcone synthase superfamily of typeIII polyketide. synthases. Nat Prod Rep 2003, 20:79-110.

A detailed review spotlighting structural analyses and biochemical studiesof type III polyketide synthases, giving a full appreciation of the reactionmechanisms, molecular diversity of this class of natural products, andviable future directions toward generating functionally diverse enzymes.


2. Schroder J: Comprehensive Natural Products Chemistry.edn 2. Oxford: Elsevier; 1999:. 749–771.

3. Ferrer JL, Jez JM, Bowman ME, Dixon RA, Noel JP:Structure of chalcone synthase and the molecularbasis of plant polyketide biosynthesis. Nat Struct Biol 1999,6:775-784.

4. Austin MB, Bowman ME, Ferrer JL, Schroder J, Noel JP: An aldolswitch discovered in stilbene synthases mediates cyclizationspecificity of type III polyketide synthases. Chem Biol 2004,11:1179-1194.

5.�

Abe I, Utsmi Y, Oguro S, Morita H, Sano Y, Noguchi H: A plant typeIII polyketide synthase that produces pentaketide chromone.J Am Chem Soc 2005, 127:1362-1363.

These authors describe a novel plant-specific type III PKS capable ofcatalyzing the formation of a pentaketide chromone using five moleculesof malonyl-CoA. The engineering of these enzymes by point mutationprovide them with the ability to produce two forms of octaketides usingeight molecules of malonyl-CoA.

6.��

Abe I, Oguro S, Utsmi Y, Sano Y, Noguchi H: Engineeredbiosynthesis of plant polyketides: chain length control in anoctaketide-producing plant type III polyketide synthase.J Am Chem Soc 2005, 127:12709-12716.

In this work, plant cDNA coding an octaketide synthase was cloned andsequenced. A single amino acid residue in the active site of this synthaseis reported to influence the polyketide chain length and product speci-ficity. This excellent example of biosynthetic engineering generated avariety of polyketides with varying chain length creating a divergenceof type III polyketides by simply applying rudimentary mutagenesistechnology.

7. Eckermann S, Schroder G, Schmidt J, Strack D, Edrada RA,Helariutta Y, Elomaa P, Kotilainen M, Kilpelainen I, Proksch Pet al.: New pathway to polyketides in plant. Nature 1998,396:387-390.

8. Tropf S, Karcher B, Schroder G, Schroder J: Reactionmechanisms of homodimeric plant polyketide synthases(stilbene and chalocone synthases). J Biol Chem 1995,270:7922-7928.

9. Suh DY, Fukuma K, Kagami J, Yamazaki Y, Shibuya M, Ebizuka Y,Sankawa U: Identification of amino acid residues important inthe cyclization reactions of chalcone and stilbene synthases.Biochem J 2000, 350:229-235.

10. Abe I, Morita H, Nomura A, Noguchi H: Substrate specificityof chalcone synthase: enzymatic formation of unnaturalpolyketides from synthetic cinnamoyl-CoA analogues.J Am Chem Soc 2000, 122:11242-11243.

11. Morita H, Noguchi H, Schroder J, Abe I: Novel polyketidessynthesized with a higher plants stilbene synthase. Eur JBiochem 2001, 268:3759-3766.

12. Abe I, Utsmi Y, Oguro S, Noguchi H: The first plant type IIIpolyketide synthase that catalyzes formation of aromaticheptaketide. FEBS Lett 2004, 562:171-176.

13. Borejsza-Wysocki W, Hrazdina G: Aromatic polyketidesynthases: purification, characterization, and antibodydevelopment to benzalacetone synthase from raspberryfruits. Plant Physiol 1996, 110:791-799.

14. Jez JM, Bowman ME, Noel JP: Expanding the biosyntheticrepertoire of plant type III polyketide synthases by alteringstarter molecule specificity. Proc Natl Acad Sci USA 2002,99:5319-5324.

15.��

Austin MB, Izumikawa M, Bowman ME, Udwary DW, Ferrer JL,Moore BS, Noel JP: Crystal structure of a bacterial oftype III polyketide synthases and enzymatic control ofreactive polyketide intermediates. J Biol Chem 2004,279:45162-45174.

Elucidation of the crystal structure of the type III PKS from this bacteriawas reported; successive mutagenic and biochemical studies havedelineated the detailed reaction mechanism to reveal both intramolecularClaisen and aldol condensation reactions for the production of 1,3,6,8-tetrahydroxynaphthalene. It is the first bacterial type III PKS crystalstructure elucidated.

16. Jez JM, Ferrer JL, Bowman ME, Dixon RA, Noel JP: Dissection ofmalonyl-coenzyme A decarboxylation from polyketide



formation in the reaction mechanism of a plant polyketidesynthase. Biochemistry 2000, 39:890-902.

17. Jez JM, Austin MB, Ferrer JL, Bowman ME, Schroder J, Noel JP:Structure control of polyketide formation in plant-specificpolyketide synthases. Chem Biol 2000, 7:919-930.

18. Fu H, Ebert-Khosla S, Hopwood DA, Khosla C: Engineeredbiosynthesis of novel polyketides: dissection of the catalyticspecificity of the act ketoreductase. J Am Chem Soc 1994,116:4166-4170.

19. Fu H, Hopwood DA, Khosla C: Engineered biosynthesis of novelpolyketides: evidence for temporal, but regiospecific, controlof cyclization of an aromatic polyketide precursor. Chem Biol1994, 1:205-210.

20. Jez JM, Bowman ME, Noel JP: Structure-guided programmingof polyketide chain-length determination in chalconesynthase. Biochemistry 2001, 40:14829-14838.


21. Abe I, Watanabe T, Lou W, Noguchi H: Active site residuesgoverning substrate selectivity and polyketide chain length inaloesone synthase. FEBS J 2006, 273:208-218.

22.�

Abe I, Watanabe T, Morita H, Kohno T, Noguchi H: Engineeredbiosynthesis of plant. polyketides: manipulation of chalconesynthase. Org Lett 2006, 8:499-502.

This paper explores the engineering of a chalcone synthase capable ofaccepting malonyl-CoA as a starting unit in place of p-coumaroyl-CoA,and successfully biosynthesizing two octacetides from malonyl-CoAs inplace of chalcone. Moreover, in this study, a notable increase in octake-tide-forming activity was observed because of an additional mutation.This paper also highlights some influential amino acid residues situated inthe active site that governs product specificity.

23. Austin MB, Saito T, Bowman ME, Haydock S, Kato A, Moore BS,Kay RR, Noel JP: Biosynthesis of Dictyostelium discoideumdifferentiation-inducing factor by a hybrid type I fatty acid-type III polyketide synthase. Nat Chem Biol 2006, 2:494-502.


Documents

A comprehensive and engaging overview of the type III family of polyketide synthases