Genome-based deletion analysis reveals the prenyl xanthone biosynthesis pathway in Aspergillus nidulans.

Abstract

Xanthones are a class of molecules that bind to a number of drug targets and possess a myriad of biological properties. An understanding of xanthone biosynthesis at the genetic level should facilitate engineering of second-generation molecules and increasing production of first-generation compounds. The filamentous fungus Aspergillus nidulans has been found to produce two prenylated xanthones, shamixanthone and emericellin, and we report the discovery of two more, variecoxanthone A and epishamixanthone. Using targeted deletions that we created, we determined that a cluster of 10 genes including a polyketide synthase gene, mdpG, is required for prenyl xanthone biosynthesis. mdpG was shown to be required for the synthesis of the anthraquinone emodin, monodictyphenone, and related compounds, and our data indicate that emodin and monodictyphenone are precursors of prenyl xanthones. Isolation of intermediate compounds from the deletion strains provided valuable clues as to the biosynthetic pathway, but no genes accounting for the prenylations were located within the cluster. To find the genes responsible for prenylation, we identified and deleted seven putative prenyltransferases in the A. nidulans genome. We found that two prenyltransferase genes, distant from the cluster, were necessary for prenyl xanthone synthesis. These genes belong to the fungal indole prenyltransferase family that had previously been shown to be responsible for the prenylation of amino acid derivatives. In addition, another prenyl xanthone biosynthesis gene is proximal to one of the prenyltransferase genes. Our data, in aggregate, allow us to propose a complete biosynthetic pathway for the A. nidulans xanthones.

Full text

Page 1

Published: February 25, 2011
r2011 American Chemical Society
4010
dx.doi.org/10.1021/ja1096682|J. Am. Chem. Soc. 2011, 133, 4010–4017
ARTICLE
pubs.acs.org/JACS
Genome-Based Deletion Analysis Reveals the Prenyl Xanthone
Biosynthesis Pathway in Aspergillus nidulans
James F. Sanchez,†Ruth Entwistle,‡Jui-Hsiang Hung,§Junko Yaegashi,†Sofina Jain,†Yi-Ming Chiang,†,||
Clay C. C. Wang,*,†,^and Berl R. Oakley*,‡
†DepartmentofPharmacologyandPharmaceuticalSciences,UniversityofSouthernCalifornia,SchoolofPharmacy,1985ZonalAvenue,
Los Angeles, California 90089, United States
‡Department of Molecular Biosciences, University of Kansas, 1200 Sunnyside Avenue, Lawrence, Kansas 66045, United States
§Department of Biotechnology, Chia Nan University of Pharmacy and Science, Tainan 71710, Taiwan
)
Graduate Institute of Pharmaceutical Science, Chia Nan University of Pharmacy and Science, Tainan 71710, Taiwan
^Department of Chemistry, University of Southern California, College of Letters, Arts, and Sciences, Los Angeles, California 90089,
United States
b
S Supporting Information
’INTRODUCTION
Xanthones are polyphenolic compounds produced by higher
plants,lichens,andfungi.Incommontheysharea9H-xanthen-9-
one scaffold, but the class is highly diverse owing to functiona-
lization with a range of substituents at various positions.1As a
consequence, the xanthone core has been described as a “privi-
leged structure,”2with members of this group exhibiting the
potential to bind to a variety of targets. Indeed, xanthones have
proven to be an important class of secondary metabolites. Over
250 of them have been shown to possess biological activities,
including antimicrobial, antioxidant, cytotoxic, and neurophar-
macological activities.1
In many instances xanthones are functionalized with prenyl
groups, commonly the five carbon dimethylallyl moieties. Exam-
ples of prenylated xanthones include mangostanin and R-man-
gostin, which are strongly inhibitory to both sensitive and
methicillin-resistant strains of Staphylococcus aureus.3γ-Mangostin
also displays potent inhibitory activity against mediators of
prostaglandin release, COX-1 and COX-2.4Other prenylated
xanthones have been found to be strongly active against Bacillus
subtilis, Staphylococcus faecalis, Staphylococcus typhi, and Candida
glabrata.5By themselves, prenyl groups are important contribu-
tors to the outstanding structural diversity of natural products.6
Prenylated secondary metabolites isolated from plants, bacteria,
andfungidisplayalargevarietyofmedicinalproperties,including
antitumor, antiretroviral, and psychotrophic activities, that are
often distinct from their nonprenylated precursors.7
Molecular genetic analysis of the biosynthesis of prenylated and
nonprenylated xanthones, as well as prenylations in general, would
greatly advance our understanding of fungal secondary metabolite
biosynthesis and, by relating gene sequence to function, facilitate
Received:October 27, 2010
ABSTRACT: Xanthones are a class of molecules that bind to a
number of drug targets and possess a myriad of biological
properties. An understanding of xanthone biosynthesis at the
genetic level should facilitate engineering of second-generation
molecules and increasing production of first-generation com-
pounds. The filamentous fungus Aspergillus nidulans has been
found to produce two prenylated xanthones, shamixanthone
and emericellin, and we report the discovery of two more,
variecoxanthone A and epishamixanthone. Using targeted dele-
tions that we created, we determined that a cluster of 10 genes including a polyketide synthase gene, mdpG, is required for prenyl
xanthone biosynthesis. mdpG was shown to be required for the synthesis of the anthraquinone emodin, monodictyphenone, and
related compounds, and our data indicate that emodin and monodictyphenone are precursors of prenyl xanthones. Isolation of
intermediate compounds from the deletion strains provided valuable clues as to the biosynthetic pathway, but no genes accounting
for the prenylations were located within the cluster. To find the genes responsible for prenylation, we identified and deleted seven
putative prenyltransferases in the A. nidulans genome. We found that two prenyltransferase genes, distant from the cluster, were
necessary for prenyl xanthone synthesis. These genes belong to the fungal indole prenyltransferase family that had previously been
shown to be responsible for the prenylation of amino acid derivatives. In addition, another prenyl xanthone biosynthesis gene is
proximal to one of the prenyltransferase genes. Our data, in aggregate, allow us to propose a complete biosynthetic pathway for the
A. nidulans xanthones.

Page 2

4011
dx.doi.org/10.1021/ja1096682 |J. Am. Chem. Soc. 2011, 133, 4010–4017
Journal of the American Chemical Society
ARTICLE
prediction of the function of homologous genes. Moreover, identi-
fication of the genes involved in the production of particular
secondarymetabolitesoffersthepossibilityofenhancedproduction
through replacement of their native promoters with strong or
inducible promoters. In addition, semisynthesis, mutasynthesis,
and chemoenzymatic synthesis that lead to second-generation
compoundswithimprovedpharmacodynamicandpharmacokinetic
properties should greatly benefit from identification and manipula-
tion of the genes involved in secondary metabolite production.
The filamentous fungus Aspergillus nidulans is, in principle, an
excellent organism in which to study the biosynthesis of prenylated
xanthones. It is known to produce two prenylated xanthones,
shamixanthone (2) and emericellin (3) (Figure 1).8,9Its genome
has been sequenced and is reasonably well annotated,10,11and
development of an efficient gene targeting system12,13has greatly
facilitated the targeted deletion of secondary metabolite genes.14,15
We have taken advantage of these attributes to identify and
characterizetheprenylxanthonebiosynthesispathwayinA.nidulans.
Wepreviouslyidentified,usingasetofdeletionsofthearomatic,
nonreducing PKS genes in a chromatin remodeling mutant back-
ground, a PKS gene, mdpG, that is responsible for synthesis of the
related polyketides monodictyphenone and emodin.16,17Here, we
employed the same set of deletions in the investigation of prenyl
xanthone synthesis and found that, again, mdpG is responsible.
Deletion of genes flanking mdpG revealed that the prenylated
xanthone biosynthetic pathway is complex, involving the products
ofninegenesthatclusterwithmdpG.Identificationofintermediate
compounds that accumulate in the deletion strains provides
important clues as to the steps in the xanthone biosynthesis
pathway. Many of these intermediates have not previously been
identified from A. nidulans and are of potential medical interest.
Surprisinglyandinterestingly,however,theprenyltransferasegenes
required for prenyl xanthone biosynthesis are not clustered with
the PKS gene. Bioinformatic analysis of the A. nidulans genome
allowedustoidentifyputativeprenyltransferasegeneselsewherein
the genome whose products might be required for prenylation of
thexanthones.Bydeletingeachofthesecandidategenes,wefound
twogenes that arerequiredfor prenylationof xanthones. They are
distant from the mdpG cluster, and it is unlikely that they could
have been identified by traditional approaches. Interestingly, the
products are members of a family of prenyltransferases that are
recognized to prenylate amino acids and their derivatives, as
Figure1. Prenylxanthonesandcompoundsthatemergedfromthestudyoftargetedgenedeletions.Thecompoundsareasfollows:variecoxanthoneA,
1; shamixanthone, 2; emericellin, 3; epishamixanthone, 4; 2,ω-dihydroxyemodin, 5; ω-hydroxyemodin, 6; emodin, 7; 9H-xanthen-9-one, 8-hydroxy-
1-(hydroxymethyl)-3-methyl-, 8; paeciloxanthone, 9; endocrocin, 10; aloe-emodin, 11; chrysophanol, 12; cichorine, 13; austinol, 14; dehydroaustinol,
15; monodictyphenone, 16; and 1(3H)-isobenzofuranone, 3-(2,6-dihydroxyphenyl)-4-hydroxy-6-methyl-, 17.

Page 3

4012
dx.doi.org/10.1021/ja1096682 |J. Am. Chem. Soc. 2011, 133, 4010–4017
Journal of the American Chemical Society
ARTICLE
opposed to polyketides. Lastly, we discovered a gene proximal to
oneof the prenyltransferase genesthat,upondeletion, leadsto the
loss of two of the four prenyl xanthones that we identified. Our
data, in aggregate, allowed us to unravel the complex xanthone
biosynthesispathwayinA.nidulans,andtheyhighlighttheutilityof
genomics coupled with powerful molecular genetic methods for
determining biosynthetic pathways, especially when the cluster is
complex and critical genes are distant from the primary cluster.
’RESULTS
Characterization of Prenylated Xanthones from A. nidu-
lans.LO2026, an A. nidulans strain carrying a deletion of the stcJ
gene (stcJΔ), which is required for the biosynthesis of the
carcinogenic secondary metabolite sterigmatocystin,18was culti-
vated on Yeast Extract Sucrose (YES) agar for 5 days at 37 ?C.
The rationale for the use of an stcJΔ strain is that elimination of
sterigmatocystin frees up the common polyketide precursor
malonyl-CoA and facilitates detection and isolation of other
metabolites. The strain also carries nkuAΔ to facilitate gene
targeting.12Reverse-phase LC/MS analysis of the crude organic
extract revealed the presence of several late-eluting, nonpolar
metabolites (Figure 2). For full characterization, we cultivated
the strain on YES plates at a larger scale and purified the
compoundsusingflashchromatography,HPLC,andpreparatory
TLC when necessary. NMR characterization revealed the mole-
cules to be shamixanthone (2) and emericellin (3), as well as, in
lesser amounts, two prenylated xanthones not previously ob-
served from this species, variecoxanthone A (1)19and epi-
shamixanthone (4).20(Figure 1; NMR data available in the
SupportingInformation,StructuralCharacterizationandFigures
S3-S15). We also found that other solid media including Yeast
Agar Glucose (YAG) and shredded wheat are conductive to
xanthone production.
Analysis of Prenyl Xanthone Biogenesis through Tar-
geted Deletions. The structural similarity of the prenyl
xanthones (Figure 1) suggests a common biosynthetic origin,
and the presence of aromatic rings is indicative of formation
catalyzed by a nonreduced polyketide synthase (NR-PKS).21
Analysis of the sequenced genome of A. nidulans suggests there
are 12 NR-PKS genes in this species. The products of two of the
genes, stcA encoding the sterigmatocystin PKS and wA, a spore
pigmentPKS,werealreadydetermined,22,23leadingustofocuson
the 10 remaining NR-PKSs. In a previous project, we discovered
that deletion of cclA, a bre2 homologue orchestrating histone H3
lysine 4 methylation, resulted in the synthesis of related aromatic
compounds including monodictyphenone and emodin under
conditions in which these molecules were normally not
observed.16For that project we deleted the 10 annotated NR-
PKS genes in a cclAΔ background and discovered that deletion of
one NR-PKS gene, AN0150.4 (using the nomenclature of the
Central Aspergillus Data Repository, CADRE, http://www.
cadre-genomes.org.uk/, and the Aspergillus genome database,
http://www.aspgd.org/)inacclAΔbackground,resultedintheelimi-
nationofthesecompounds.17WedesignatedAN0150.4mdpGand
thePKSitencodesMdpG.(SeeTable1forour genedesignations
andthecorrespondingannotationsfromCADRE/AspGDandthe
Broad Institute Aspergillus Comparative Database.) For the
current project we utilized the same 10 NR-PKS deletant strains
tolearnwhetheranyofthemwererequiredforthesynthesisofthe
prenyl xanthones in A. nidulans. We were surprised to discover
that, in addition to monodictyphenone and its related molecules,
the xanthones disappeared in the mdpGΔ strain (Figure 2) and
not in the other deletant strains (data not shown). This reveals
that MdpG is the PKS responsible for the synthesis of not only
emodin, monodictyphenone, and related compounds but also the
xanthones.
Next, we tried to identify additional genes involved in A.
nidulans prenyl xanthone biosynthesis. Facilitating the search,
secondary metabolism genes in A. nidulans are usually clustered,
prompting us to focus on the genes surrounding mdpG
(AN10039 to AN0153) (Figure 3). To study the synthesis of
the xanthones as it would occur naturally, without the influence
of a potentially powerful chromatin remodeling mutation, we
deletedthesegenesintheLO2026backgroundcarryingthewild-
type cclA gene (cclAþ). All deletions were verified by diagnostic
PCR (using three different primer sets for each gene) (see
Materials and Methods).
Attempted deletion of AN0153, homologous to a DNA-
binding protein, failed to yield any viable transformants, in
agreement with previous observations.17This gene appears to
be essential for viability and unrelated to secondary metabolism
and was thus excluded from further study here. Deletion of the
other genes significantly reduced or eliminated prenyl xanthone
formation except for AN10039, mdpEΔ, and mdpIΔ (Figures 3
and 4). Two differences were observed in relation to our
investigation using cclAΔ: In the cclAΔ background, deletion
ofthetranscriptionalregulatorgenemdpEeliminatesthesynthe-
sis of monodictyphenone and related compounds; however, for
the prenyl xanthones this gene does not appear consequential
(Figure3).Additionally,themonooxygenasegenemdpDwasnot
found to be important for the production of monodictyphenone
and analogous products in the cclAΔ background, whereas it
plays a crucial role in prenyl xanthone formation.
The deletant strains that reduced prenyl xanthone production
were examined for the presence of any new metabolites that
mightrepresent intermediatesinthebiosynthetic pathwayofthe
xanthones. Extracts from strains carrying deletions of mdpG,
mdpA, and mdpF contained no obvious intermediates from the
prenyl xanthone biosynthesis pathways. A strain carrying a
deletion of mdpH did not display new metabolites in significant
amounts, but a compound with the mass and retention time of
endocrocin(10),recentlyidentifiedfrom anmdpH/cclA double-
deletant strain,17was observed.
Figure 2. HPLCprofilesofextractsfromstcJΔ;stcJΔ,cclAΔ;andstcJΔ,
cclAΔ, mdpGΔ strains, as detected by UV absorbance at 254 nm.
Numbering of peaks correspond to the compounds in Figure 1. Shami-
xanthone 2 and emericellin 3 elute at the same retention time.

Page 4

4013
dx.doi.org/10.1021/ja1096682 |J. Am. Chem. Soc. 2011, 133, 4010–4017
Journal of the American Chemical Society
ARTICLE
The mdpLΔ strain, the extract of which displayed the greatest
number of significant peaks in the chromatogram, was grown in
largescaleonYESplates,andthemajormetaboliteswereisolated
by flash chromatography and HPLC and characterized primarily
using NMR. The compounds were determined to be 2,ω-
dihydroxyemodin (5), ω-hydroxyemodin (6), emodin (7),
aloe-emodin (11), and chrysophanol (12) (Figure 4; NMR data
shown in the Supporting Information). Also, mdpCΔ, mdpJΔ,
and mdpKΔ extracts each exhibited5, 6, and 7, as determinedby
their masses and retention times in comparison to the mdpLΔ
extract, but little or no 11 or 12. Chromatograms of mdpBΔ
extracts similarly displayed the three metabolites, along with a
number of small, indistinct peaks.
The mdpDΔ extract also yielded 5, 6, and 7 but with addition
of monodictyphenone (16), a core xanthone structure (8), and
the C-prenylated analog paeciloxanthone (9). Finally, the poly-
ketide cichorine (13) and the meroterpenoids austinol (14) and
dehydroaustinol (15) could be detected in all strains tested,
indicating they are unrelated to prenyl xanthone biosynthesis.
Identification of Two Genes Required for Prenylation of
the Xanthone Scaffold. Although our results revealed a gene
cluster responsible for synthesis of A. nidulans prenyl xanthones,
the lack of a prenyltransferase gene in this cluster made it clear
that the required prenyltransferase gene(s) must be located
elsewhere in the genome. On the basis of homology to known
genes of this class,24we were able to identify several putative
prenyltransferase genes in the A. nidulans genome: AN6784.4,
AN8514.4 [tdiB25], AN10289.4, AN11080.4, AN11194.4, and
AN11202.4. We deleted each of these genes, and the deletant
strains were cultured on YES plates. LC/MS analysis revealed
that deletion of AN6784.4, but not the other prenyltransferase
genes, resulted in the elimination of 2-4 (Figure 5a) and
accumulation of a compound which upon large-scale isolation
andcharacterizationwasconfirmedtobevariecoxanthoneA(1).
Because AN6784.4 is a prenyltransferase gene required for
synthesis of the xanthones, we hereby designate it xptA
(xanthone prenyltransferase A).
Since the A. nidulans xanthones contain two prenyl groups
(except for variecoxanthone A, which only has one), it was
Table 1. Gene Designations of the Text and the Corresponding Annotations from Two Websitesa
gene designationputative function CADRE/AspGD annotationBroad annotation
AN10039
mdpA
mdpB
mdpC
mdpD
mdpE
mdpF
mdpG
mdpH
mdpI
mdpJ
mdpK
mdpL
AN0153
xptA
xptB
xptC
histidine acid phosphatase
regulatory gene
dehydratase
ketoreductase
monooxygenase
regulatory gene
Zn-dependent hydrolase
polyketide synthase
hypothetical protein
acyl-CoA synthase
glutathione S transferase
oxidoreductase
Baeyer-Villiger oxidase
MYB DNA binding protein
prenyltransferase
prenyltransferase
GMC oxidoreductase
AN10039.4
AN10021.4
AN10049.4
AN0146.4
AN0147.4
AN0148.4
AN0149.4
AN0150.4
AN10022.4
AN10035.4
AN10038.4
AN10044.4
AN10023.4
AN0153.4
AN6784.4
unannotated
AN7998.4
unannotated
ANID_10021.1
ANID_10049.1
ANID_00146.1
ANID_00147.1
ANID_00148.1
ANID_00149.1
ANID_00150.1
ANID_11847.1 þ ANID_11848.1
ANID_10035.1
ANID_10038.1
ANID_10044.1
ANID_10023.1
ANID_00153.1
ANID_06784.1
ANID_12402.1 þ ANID_12430.1
ANID_07998.1
aCADRE (http://www.cadre-genomes.org.uk) which uses the same designations as the Aspergillus genome database (AspGD, http://www.aspgd.org)
and the Broad Institute Aspergillus Comparative Database (http://www.broadinstitute.org/annotation/genome/aspergillus_group/MultiHome.
html). Putative functions are from BLAST searches performed previously16or for this study.
Figure 3. (Top) Organization ofthegenes surrounding the PKSmdpG
involvedinprenylxanthonebiosynthesis.Eacharrowindicatesgenesize
and direction of transcription. On the basis of a set of deletions we
created and analyzed, genes shown in black are involved in prenyl
xanthone biosynthesis while those in gray are not. (Middle) Organiza-
tion of the genes for monodictyphenone biosynthesis in a cclAΔ
background (based on ref 17). AN0147 and AN0148 are circled to
emphasize that AN0147 is unnecessary for monodictyphenone genera-
tion but required for xanthone synthesis, whereas AN0148 is necessary
for monodictyphenone generation but not required for xanthone
synthesis. (Bottom) Organization of the genes outside of the mdpG
cluster that are involved in prenyl xanthone synthesis. The genes that
were successfully deleted were AN10039 through mdpL, xptA, xptC,
AN7999, and xptB.

Page 5

4014
dx.doi.org/10.1021/ja1096682 |J. Am. Chem. Soc. 2011, 133, 4010–4017
Journal of the American Chemical Society
ARTICLE
apparent that an additional prenyltransferase gene was required.
We hypothesized that the genome must contain an additional,
unannotated prenyltransferase gene. We performed a local
BLAST search of the Broad Institute Aspergillus Comparative
Database(http://www.broadinstitute.org/annotation/genome/
aspergillus_group/MultiHome.html)usingAN6784.4asaquery
sequence. We identified two annotated partial gene fragments
designated ANID_12402.1 and ANID_12430.1. Inspection of
the two fragments revealed that the 30-end of ANID_12402.1
overlapswiththe50-endofANID_12430.1by∼600bp.Further,
the Web site predicted that the sequence of ANID_12402 is
prematurely truncated at its 30-end. For these reasons we
concluded that the nonoverlapping portion of ANID_12430
was in fact the continuation of ANID_12402 and that the 50-end
of ANID_12402 and the 30-end of ANID_12430 are the
boundariesof asinglegene.Indeed, thesequenceofthis putative
gene is found intact in CADRE/AspGD but without annotation.
Deletion of this gene led to the elimination of the xanthones and
accumulation of several metabolites, including 7, 12, 16, and a
putative monodictyphenone derivative (17) (Figure 5a). We
designate this new gene xptB (xanthone prenyltransferase B).
IdentificationofanAdditionalPrenylXanthoneBiosynthe-
sis Gene. The data were suggestive of a biosynthesis in which
emericellin (3) is a precursor to shamixanthone (2) and epi-
shamixanthone (4) (see Discussion). Even after identification of
xptA and xptB it was clear that a gene encoding the enzyme
responsible for the transformation from emericellin to shami-
xanthone and epishamixanthone was required but not yet identi-
fied. Given thetendencyoffungalsecondarymetabolite genesto
exist in clusters, we reasoned that the missing gene might reside
proximatetoxptAorxptB.Ananalysisoftheputativefunctionof
the genes neighboring xptA revealed none that were likely
involved in secondary metabolism; however, AN7998.4 and
AN7999.4 upstream of xptB were possible candidates. Deletion
of AN7998.4 yielded a metabolite profile that continued to
display variecoxanthone A (1) and 3 but lacked 2 and 4
(Figure 5b and 5c), revealing that AN7998.4 is the missing gene
in the biosynthesis of the two prenyl xanthones. We designate
AN7998.4 as xptC. The amino acid sequence indicates that it is a
member of the glucose-methanol-choline (GMC) oxidore-
ductase superfamily, a broad class that includes cellobiose
dehydrogenase,cholinedehydrogenase,andmethanoloxidase.26
Deletion of AN7999.4 had no obvious effect on prenyl xanthone
biosynthesis.
’DISCUSSION
We used a combination of genomics, efficient gene targeting,
and natural products chemistry to elucidate the complex prenyl
xanthone biosynthesis pathway in A. nidulans, the first such
pathway to be deciphered in a fungus. This pathway is complex,
involving the products of 10 clustered genes and three genes
located apart from the main cluster. The genes mdpA through
mdpLarelocatedontherightarmofChromosomeVIII,whereas
xptA is located on the right arm of Chromosome I and xptB and
xptCarefoundontheleftarmofChromosomeII.Inallcasesthe
genesare∼0.5Mbfromatelomere.Otheruncommonexamples
of separated fungal secondary metabolism genes include the
genes for T-toxin biosynthesis in Cochliobolus heterostrophus,
with 9 genes in two unlinked loci, and the nonclustered acetyl-
transferasegeneTri101involvedintrichothecenebiosynthesisin
Fusarium species, having a different evolutionary history from
other trichothecene genes.27,28Without genomics and efficient
gene targeting it would have been difficult, if not impossible, to
identify all of the genes in the pathway. In particular, it would
haveprobablybeenextremelydifficulttoidentifythethreegenes
that are separated in the genome from the bulk of the xanthone
Figure 4. HPLC extracts of strains in the cluster as detected by UV absorbance at 254 nm. ω-Hydroxyemodin 6 and the unrelated metabolite austinol
14 elute at the same retention time.

Page 6

4015
dx.doi.org/10.1021/ja1096682 |J. Am. Chem. Soc. 2011, 133, 4010–4017
Journal of the American Chemical Society
ARTICLE
biosynthesis genes in the absence of the ability to identify
candidate genes and to delete them en masse.
The two prenyltransferases, encoded by the xptA and xptB
genes, are particularly interesting. They are both homologous to
members of a class of enzymes known as fungal indole
prenyltransferases.29For example, XptA is 27% identical at the
amino acid level to SirD from Leptosphaeria maculans30and 24%
identical to FgaPT2 from Aspergillus fumigatus.31XptB is 29%
and 27% identical to these enzymes, respectively, and XptA
and XptB are 37% identical to each other. Unlike many other
prenyltransferase enzymes, the fungal indole prenyltransferases
aresoluble,lack(N/D)DxxDmotifs,andcontinuetobeactivein
the absence of Mg2þ.29The fungal indole prenyltransferases are
capable of regular and reverse C-prenylations as well as N- and
O-prenylations. Their solubility may be useful for chemoenzy-
maticsynthesisintermsofease ofmanipulationandefficiencyof
catalysis. All members of this class studied in detail to date
prenylate compounds of amino acid origin, whereas our data
reveal that XptA and XptB clearly prenylate polyketides. There
werehintsfromtheliteraturethatthisclassmighthavearangeof
substrates extending to polyketides. For example, the X-ray
structure of dimethylallyl tryptophan synthase (DMATS) from
A. fumigatus revealed a common architecture (but no significant
primary amino acid similarity) with the bacterial enzyme NphB
that catalyzes addition of a 10-carbon geranyl moiety to a
polyketide-based aromatic scaffold.32In addition, a gene in
Penicillium aethiopicum that is homologous to indole prenyl-
transferases is proximal to a polyketide gene necessary for the
formation of viridicatumtoxin and was proposed to contribute to
addition of a geranyl group to that metabolite.33Our findings
provide the first direct evidence, however, that this class of
prenyltransferases can prenylate polyketides, and the possibility
must be considered that XptA and XptB may prenylate polyke-
tide substrates beyond those reported in this study.
Our deletions reveal the genes required for prenyl xanthone
synthesis, but correlation of the deleted genes with the inter-
mediates that accumulate in the deletion strains allows us to
propose a biosynthetic pathway for the A. nidulans prenyl
xanthones (Figure 1). We previously detailed the synthesis of
monodictyphenone (16) in a cclAΔ background, which was
dependent on many of the same secondary metabolite genes
also responsible for the prenyl xanthones.17We propose that 16
is an intermediate en route to prenyl xanthones (Figure 6). One
newobservation,however,isthatfromthemdpLΔstrainwehave
now isolated chrysophanol (12).
Earlierworkbyothersusedisotopicprecursorstoproposethe
biosynthesis of 2 and a similar metabolite, tajixanthone, in
Aspergillus variecolor.34The labeling patterns from [1-13C]- and
[2-13C]-acetateincorporationsuggestedanoctaketideprecursor,
Figure 6. Proposed biosynthetic pathway of prenyl xanthones.
Figure 5. (a) HPLC extracts of the prenyltransferase deletant strains
xptAΔ and xptBΔ as detected by UV absorption at 254 nm. (b) HPLC
extractofxptCΔ.(c)MassspectraofthemajorLC/MSpeakfrom(top)
stcJΔ control extract and (bottom) xptCΔ extract.

Page 7

4016
dx.doi.org/10.1021/ja1096682 |J. Am. Chem. Soc. 2011, 133, 4010–4017
Journal of the American Chemical Society
ARTICLE
whichcould condenseintoananthrone andpossiblyautooxidize
to an anthraquinone. Interestingly, the earlier work also found
thatisotopicallylabeled12wasincorporatedintotajixanthone;it
was proposed that xanthoneformationproceeds from anoctake-
tideto7andfurtherto12.Becauseourearlieranalysis17indicates
that 12is notinthedirect biosyntheticpathway from7to16, we
propose that it is a shunt metabolite; more specifically, the mdpL
knockout may yield an unstable intermediate that transforms
to chrysophanol. 7 might not be converted to 12 when the
biosynthesis is uninterrupted. An alternative explanation for
incorporation of 12into tajixanthone is that 12 is metabolized to
7at alow rate through anendogenousmechanism. Aloe-emodin
(11), the ω-hydroxylated analog of 12, may be a side product
arising from oxidation by an endogenous cytochrome P450
enzyme.35
Elucidatingtheenzymaticconversionof7to16iscomplicated
bythefactthattheprocessappearstobedependentonfiveofthe
clustered genes, mdpB, mdpC, mdpJ, mdpK, and mdpL, and that
the metabolite profiles from each of the corresponding deletion
strains are similar. As a further complication, the presence of
several metabolites in a deletant strain (for example, five major
compounds from mdpLΔ)makes it difficult to ascribe theirroles
in the biosynthesis and their relation to each other. We refer the
readertoref17foradiscussionofthepossiblefunctionsofmdpB,
mdpC, mdpJ, mdpK, and mdpL in monodictyphenone biosynthe-
sis. Our data allow us, however, to deduce the remainder of
the prenyl xanthone biosynthesis pathway with considerable
confidence.
Inthisstudyweshowconclusivelythatfourgenes,mdpD,xptA,
xptB, and xptC are, at a minimum, responsible for conversion of
monodictyphenonetotheprenylxanthones.Itiseasiesttodeduce
the steps of the biosynthetic pathway by working back from the
finalproducts,shamixanthone(2)andepishamixanthone(4).The
xptC deletion does not accumulate 2 or 4 but does accumulate
emericellin (3) and variecoxanthone A (1). This allows us to
deduce that XptC is required for conversion of 3 or 1 to 2 and 4.
The xptA deletion strains accumulate 1 but not 3, 2, or 4. This
result allows us to deduce that XptA is required for the C-pre-
nylation of 1 to form 3 and also allows us to deduce that 3 is the
compound that is converted by XptC to 2 and 4. Given the
similarity of XptCto oxidativeenzymes,wesuggest thatoxidation
of the primary alcohol of 3 to an aldehyde, followed by ene
cyclization,19may yield 2 and its epimer 4. From the facts
that XptA catalyzes the C-prenylation of 1 and no prenylated
xanthonesarefoundinxptBdeletionstrains,wededucethatXptB
catalyzes the O-prenylation required for formation of 1 and that
the O-prenylation occurs before the C-prenylation and is a
prerequisite for the C-prenylation. From the mdpD deletant we
isolated compound 8, which contains the xanthone core but does
notbearfunctionalityatthe2position,atwhichthefinalproducts
feature the O-prenyl group. Oxidation of the xanthone core at the
C-2 position is necessary for O-prenylation, so we postulate that
MdpD, homologous to known monooxygenases, catalyzes oxida-
tion prior to O-prenylation by XptB. In the absence of the C-2
oxygen, XptA is able to yield the C-prenylated product 9 in small
amounts. In the xptB deletant strain, we were able to detect
compound 17, which we suspect to be a shunt metabolite arising
from reduction of the carbonyl group of monodictyphenone(16)
and condensation with the carboxylic acid.
In addition to assisting in elucidating the biosynthetic path-
way, many of the shunt products and intermediate molecules are
bioactive. For instance, 11 is cytotoxic to neuroblastoma, Ewing
sarcoma, and pPNET cell lines and inhibits neuroblastoma
tumors in vivo but is not toxic to normal cells nor animal
models.369 displays antimicrobial and antiacetylcholineesterase
activity and is also cytotoxic against hepG2 cells.37Compound 8
has antifungal and antibacterial activities.38These findings in-
dicate that a benefit of targeted deletions of genes is the
accumulation of useful quantities of biosynthetic intermediates
that may themselves possess noteworthy biological properties.
’CONCLUSION
Previous work demonstrated that a cluster of 10 genes was
responsible for several related compounds that emerged in a
chromatin deletant strain of A. nidulans, including mono-
dictyphenone.17The current work reveals that four prenyl
xanthones rely on most of these genes for their synthesis but in
addition require a monooxygenase gene within the cluster and
three genes outside of the cluster, including two belonging to a
family of indole prenyltransferase genes and one homologous to
GMC oxidoreductase genes. The combined data allow us to
propose that these latter genes are involved in the later stages of
prenyl xanthone formation to complete a biosynthesis in which
monodictyphenone is a precursor.
’MATERIALS AND METHODS
Generation of Fusion PCR Fragments, A. nidulans Proto-
plasting, and Transformation. All gene deletions were carried out
according to the gene targeting procedures of Szewczyk et al.13Two
∼1000 base pair fragments upstream and downstream of every targeted
genewereamplifiedfromA.nidulansgenomicDNAusingPCR.Primers
used in this study are listed in Table S1, Supporting Information. The
two amplified flanking sequences and an A. fumigatus pyrG selectable
marker cassette were fused together by PCR using nested primers. A.
nidulans strains in this study are displayed in Table S2, Supporting
Information. Protoplast generation and transformation were performed
as described.13The strain LO2026 carrying a deletion of the stcJ gene
that eliminates sterigmatocystin production was used as the recipient
strain. Diagnostic PCR of the deletant strains was carried out using
the external primers from the first round of PCR. The difference in the
size between the gene replaced by the selective marker and the native
gene allowed us to establish if the transformants carried correct gene
replacements. For further verification, diagnostic PCR was performed
two more times, using one of the external primers and a primer located
inside the marker gene, then the other external primer and an internal
primer. In these instances, the deletants yielded the PCR product of the
expected size whereas no product was present in nondeletants.
Media and Cultivation of Strains. YES media was prepared by
combining 20 g of yeast extract, 120 g of sucrose, 20 g of agar, and 2 mL
oftraceelementsolution15in1LH2O.ForLC/MSscreening,sporesof
LO2026(thecontrolstrain)andthreestrainsofeachgenedeletantwere
individually cultivated (1 ? 107spores) on 10 ? 150 mm Petri dishes
containing YES agar and cultivated at 37 ?C for 5 days. The agar was
chopped into ∼2 cm2pieces, and the material was extracted using
sonication with methanol, followed by 1:1 methanol:dichloromethane.
Theorganicsolventswereremovedinvacuo,andtheremainingmaterial
was partitioned between H2O (25 mL) and ethyl acetate (25 mL ? 2).
The combined ethyl acetate layers were evaporated, and the crude
material was redissolved at a concentration of 20 mg/mL in DMSO and
then diluted 5-fold in methanol.
LC/MS Analysis. LC/MS was carried out using a ThermoFinnigan
LCQ Advantage ion trap mass spectrometer with an RP C18 column
(AlltechPrevail;2.1?100mmwitha3μmparticlesize)ataflowrateof
125 μL/min and monitored by a UV detector at 254 nm. The solvent

Page 8

4017
dx.doi.org/10.1021/ja1096682 |J. Am. Chem. Soc. 2011, 133, 4010–4017
Journal of the American Chemical Society
ARTICLE
gradient was 95% MeCN-H2O (solvent B) in 5% MeCN-H2O
(solvent A) both containing 0.05% formic acid:0% B from 0 to 5 min,
0to100%Bfrom5to35min,100%Bfrom35to40min,100%Bto0%
B from 40 to 45 min, and reequilibration with 0% B from 45 to 50 min.
IsolationofMetabolites.TheLO2026(stcJΔ),LO3387(mdpLΔ),
LO3337 (mdpDΔ), LO3896 (xptAΔ), and LO4178 (xptBΔ) strains
were each cultivated in 25 ? 150 mm Petri dishes containing 2 L YES
mediaandextracted inthesame manneras above.The crudematerialwas
subjected to silica gel column chromatography, using ethyl acetate and
hexanes as the eluent. Thematerials were furtherseparatedbypreparative
HPLC [Phenomenex Luna 5 μm C18 (2), 250 ? 21.2 mm] with a flow
rate of 5.0 mL/min and measured by a UV detector at 250 nm.
Shamixanthone (2), emericellin (3), epishamixanthone (4), and paecilo-
xanthone (9) required further purification using preparative TLC. See
Supporting Information for more details about isolation.
Structural Characterization. See the Supporting Information
under Structural Characterization and Figures S3-S15.
’ASSOCIATED CONTENT
b
S
Supporting Information.
used in this study, detailed structural characterization, purifica-
tion methods, and complete refs 10 and 11. This material is
available free of charge via the Internet at http://pubs.acs.org.
A nidulans strains and primers
’AUTHOR INFORMATION
Corresponding Author
boakley@ku.edu; clayw@usc.edu
’ACKNOWLEDGMENT
This work was supported by grant PO1GM084077 from the
NationalInstituteofGeneralMedicalSciencesandbytheKansas
University Endowment Fund.
’REFERENCES
(1) El-Seedi, H. R.; El-Barbary, M. A.; El-Ghorab, D. M. H.; Bohlin,
L.; Borg-Karlson, A. K.; Goransson, U.; Verpoorte, R. Curr. Med. Chem.
2010, 17, 854–901.
(2) Lesch, B.; Brase, S. Angew. Chem., Int. Ed. 2004, 43, 115–118.
(3) Panthong, K.; Pongcharoen, W.; Phongpaichit, S.; Taylor, W. C.
Phytochemistry 2006, 67, 999–1004.
(4) Nakatani, K.; Nakahata, N.; Arakawa, T.; Yasuda, H.; Ohizumi,
Y. Biochem. Pharmacol. 2002, 63, 73–79.
(5) Saleem,M.;Nazir,M.;Ali,M.S.;Hussain,H.;Lee,Y.S.;Riaz,N.;
Jabbar, A. Nat. Prod. Rep. 2010, 27, 238–254.
(6) Heide, L. Curr. Opin. Chem. Biol. 2009, 13, 171–179.
(7) Botta,B.; Vitali, A.; Menendez, P.; Misiti,D.; Delle Monache, G.
Curr. Med. Chem. 2005, 12, 713–739.
(8) (a) Marquez-Fernandez, O.; Trigos, A.; Ramos-Balderas, J. L.;
Viniegra-Gonzalez, G.; Deising, H. B.; Aguirre, J. Eukaryot. Cell 2007,
6, 710–720. (b) Marquez-Fernandez, O.; Trigos, A.; Ramos-Balderas,
J. L.; Viniegra-Gonzalez, G.; Deising, H. B.; Aguirre, J. Eukaryot. Cell
2007, 6, 710.
(9) Ishida, M.; Hamasaki, T.; Hatsuda, Y. Agric. Biol. Chem. 1975,
39, 2181–2184.
(10) Galagan, J. E.; et al. Nature 2005, 438, 1105–1115.
(11) Wortman, J. R.; et al. Fungal Genet. Biol. 2009, 46, S2–S13.
(12) Nayak, T.; Szewczyk, E.; Oakley, C. E.; Osmani, A.; Ukil, L.;
Murray, S. L.; Hynes, M. J.; Osmani, S. A.; Oakley, B. R. Genetics 2006,
172, 1557–1566.
(13) Szewczyk,E.;Nayak,T.;Oakley,C.E.;Edgerton,H.;Xiong,Y.;
Taheri-Talesh, N.; Osmani, S. A.; Oakley, B. R. Nat. Protoc. 2006,
1, 3111–3120.
(14) Chiang, Y. M.; Szewczyk, E.; Nayak, T.; Davidson, A. D.;
Sanchez, J. F.; Lo, H. C.; Ho, W. Y.; Simityan, H.; Kuo, E.; Praseuth,
A.; Watanabe, K.; Oakley, B. R.; Wang, C. C. C. Chem. Biol. 2008,
15, 527–532.
(15) Sanchez, J. F.; Chiang, Y. M.; Szewczyk, E.; Davidson, A. D.;
Ahuja, M.; Oakley, C. E.; Bok, J. W.; Keller, N.; Oakley, B. R.; Wang,
C. C. C. Mol. Biosyst. 2010, 6, 587–593.
(16) Bok, J. W.; Chiang, Y. M.; Szewczyk, E.; Reyes-Dominguez, Y.;
Davidson, A. D.; Sanchez, J. F.; Lo, H. C.; Watanabe, K.; Strauss, J.;
Oakley, B. R.; Wang, C. C. C.; Keller, N. P. Nat. Chem. Biol. 2009,
5, 462–464.
(17) Chiang, Y. M.; Szewczyk, E.; Davidson, A. D.; Entwistle, R.;
Keller, N. P.; Wang, C. C. C.; Oakley, B. R. Appl. Environ. Microb. 2010,
76, 2067–2074.
(18) Brown, D. W.; Adams, T. H.; Keller, N. P. Proc. Natl. Acad. Sci.
U.S.A. 1996, 93, 14873–14877.
(19) Chexal,K.K.;Holker,J.S.E.;Simpson,T.J.;Young,K.J.Chem.
Soc., Perkin Trans. 1 1975, 543–548.
(20) Ishida, M.; Hamasaki, T.; Hatsuda, Y.; Fukuyama, K.; Tsukihara,
T.; Katsube, Y. Agric. Biol. Chem. 1976, 40, 1051–1052.
(21) Cox, R. J. Org. Biomol. Chem. 2007, 5, 2010–2026.
(22) Watanabe,A.;Fujii,I.;Sankawa,U.;Mayorga,M.E.;Timberlake,
W. E.; Ebizuka, Y. Tetrahedron Lett. 1999, 40, 91–94.
(23) Minto,R.E.;Townsend,C.A.Chem.Rev.1997,97,2537–2555.
(24) von Dohren, H. Fungal Genet. Biol. 2009, 46, S45–S52.
(25) Bok, J. W.; Hoffmeister, D.; Maggio-Hall, L. A.; Murillo, R.;
Glasner, J. D.; Keller, N. P. Chem. Biol. 2006, 13, 31–37.
(26) Cavener, D. R. J. Mol. Biol. 1992, 223, 811–814.
(27) Inderbitzin,P.;Asvarak,T.;Turgeon,B.G.Mol.Plant-Microbe
Interact. 2010, 23, 458–472.
(28) O’Donnel, K.; Kistler, H. C.; Tacke, B. K.; Casper, H. H. Proc.
Natl. Acad. Sci. U.S.A. 2000, 93, 7905–7910.
(29) Steffan, N.; Grundmann, A.; Yin, W. B.; Kremer, A.; Li, S. M.
Curr. Med. Chem. 2009, 16, 218–231.
(30) Kremer, A.; Li, S. M. Microbiology (UK) 2010, 156, 278–286.
(31) Unsold,I. A.;Li,S.
151, 1499–1505.
(32) Metzger, U.; Schall, C.; Zocher, G.; Unsold, I.; Stec, E.; Li,
S. M.; Heide, L.; Stehle, T. Proc. Natl. Acad. Sci. U.S.A. 2009,
106, 14309–14314.
(33) Chooi, Y. H.; Cacho, R.; Tang, Y. Chem. Biol. 2010, 17, 483-
494.
(34) Ahmed,S.A.;Bardshiri,E.;McIntyre,C.R.;Simpson,T.J.Aust.
J. Chem. 1992, 45, 249–274.
(35) Kelly, D. E.; Krasevec, N.; Mullins, J.; Nelson, D. R. Fungal
Genet. Biol. 2009, 46, S53–S61.
(36) Pecere, T.; Gazzola, M. V.; Mucignat, C.; Parolin, C.; Dalla
Vecchia,F.;Cavaggioni,A.;Basso,G.;Diaspro,A.;Salvato,B.;Carli,M.;
Palu, G. Cancer Res. 2000, 60, 2800–2804.
(37) Wen, L.; Lin, Y. C.; She, Z. G.; Du, D. S.; Chan, W. L.; Zheng,
Z. H. J. Asian Nat. Prod. Res. 2008, 10, 133–137.
(38) Hein, S. M.; Gloer, J. B.; Koster, B.; Malloch, D. J. Nat. Prod.
1998, 61, 1566–1567.
M. Microbiology (UK)2005,