1 running title: sorghum 3-deoxyanthocyanidin biosynthesis

3-Deoxyanthocyanidins are the unique phytoalexins synthesized by sorghum in response to fungal inoculation. They are structurally related to anthocyanins but the fi nal steps of their pathogen-inducible biosynthesis are not fully understood. We have identifi ed new fl avonoid structural genes from the recently completed sorghum BTx623 genome sequence. The biochemical functions of the different expressed sorghum genes were established in planta by complementation in the appropriate Arabidopsis transparent testa mutants. There is a family of nine chalcone synthase genes which are all inducible by fungal inoculation in sorghum seedlings. Specifi c dihydrofl avonol 4-reductase (DFR) genes responsive to conditions which stimulated anthocyanin accumulation ( SbDFR1 ) or 3-deoxyanthocyanidin production ( SbDFR3 ) were identifi ed. Recombinant SbDFR1 and SbDFR3 were found to function as typical DFRs by accepting dihydrofl avonol substrates. On the other hand, both DFRs showed substantially lower but detectable NADPH-dependent activities toward fl avanones. Reduction of fl avanones to fl avan-4-ols is a reaction step required for 3-deoxyanthocyanidin production. Flavanone 3-hydroxylase (F3H) converts fl avanones to dihydrofl avonols for anthocyanin biosynthesis. In sorghum seedlings, expression of two F3H genes was either absent or strongly suppressed during the accumulation of 3-deoxyanthocyanidins. Under such conditions, most fl avanones are expected to be reduced by the pathogen-induced SbDFR3 for the formation of fl avan-4-ols. Our work also revealed that 3-deoxyanthocyanidin accumulation and SbDFR3 expression were induced by methyl jasmonate treatment in sorghum roots but the stimulation effects were antagonized by salicylic acid.

Keywords: Anthocyanin • 3-Deoxyanthocyanidins • Dihydrofl avonol 4-reductase • Jasmonic acid • Phytoalexin • Sorghum .

Abbreviations : ANS , anthocyanidin synthase ; CHS , chalcone synthase ; CHI , chalcone isomerase ; DFR , dihydrofl avonol 4-reductase ; DOX , 2-oxoglutarate-dependent oxygenase ; F3H , fl avanone 3-hydroxylase ; IPTG , isopropyl- β - D -thiogalactopyranoside ; JA , jasmonic acid ; LC-MS , liquid chromatography–mass spectrometry ; Me-JA , methyl jasmonate ; MRM , multiple-reaction monitoring ; MS , Murashige and Skoog ; MS/MS , tandem mass spectrometry ; RT–PCR , reverse transcription–PCR ; SA , salicylic acid.

The nucleotide sequence reported in this paper has been submitted to NCBI GenBank under the accession number GU320740 ( SbF3H2 ).

Introduction

3-Deoxyanthocyanins (orange-red coloration) are structurally similar to anthocyanins except for the absence of C-3 hydroxy-lation. They are the major pigments in sinningia ( Sinningia

cardinalis ) fl owers ( Winefi eld et al. 2005 ) and are found in silk tissues of certain maize lines ( Halbwirth et al. 2003 ). Sorghum ( Sorghum bicolor L.) is considered to be the only dietary source for 3-deoxyanthocyanins which are present in large quantities up to 10 mg g − 1 in the bran of some cultivars ( Awika et al. 2004 ). Many plants use the products of secondary metabolism to protect themselves from pathogens. In sorghum, this response is an active process which results in rapid accumulation of high levels of 3-deoxyanthocyanidin phytoalexins in infected tissues ( Snyder and Nicholson 1990 ). We identifi ed these compounds as luteolinidin, apigeninidin and their methoxylated derivatives, which are all present as aglycones ( Lo et al. 1996 , Wharton et al. 2000 ). In leaves, these phytoalexins fi rst appear in the cells under fungal attack, where they accumulate in cytoplasmic inclusion bodies ( Snyder and Nicholson 1990 ). The inclusions migrate to the site of attempted penetration, become pigmented and ultimately

Molecular Dissection of the Pathogen-Inducible 3-Deoxyanthocyanidin Biosynthesis Pathway in Sorghum Hongjia Liu 1 , 2 , 4 , Yegang Du 1 , 4 , Hung Chu 1 , Chun Hat Shih 1 , Yu Wai Wong 1 , Mingfu Wang 1 , Ivan K Chu 3 , Yuezhi Tao 2 and Clive Lo 1 , ∗ 1 School of Biological Sciences, The University of Hong Kong, Pokfulam Road, Hong Kong, China 2 Crop Molecular Breeding Center, Zhejiang Academy of Agricultural Sciences, Hangzhou, Zhejiang, China 3 Department of Chemistry, The University of Hong Kong, Pokfulam Road, Hong Kong, China 4 These authors contributed equally to this work ∗ Corresponding author: E-mail, [email protected] ; Fax, + 852-2858-3477 (Received April 30, 2010; Acepted May 31, 2010)

Plant Cell Physiol. 51(7): 1173–1185 (2010) doi:10.1093/pcp/pcq080, available online at www.pcp.oxfordjournals.org© The Author 2010. Published by Oxford University Press on behalf of Japanese Society of Plant Physiologists.All rights reserved. For permissions, please email: [email protected]

1173Plant Cell Physiol. 51(7): 1173–1185 (2010) doi:10.1093/pcp/pcq080 © The Author 2010.

Regular Paper

Downloaded from https://academic.oup.com/pcp/article-abstract/51/7/1173/1912614by gueston 31 January 2018

release their contents, killing both the plant cell and the fungus. We also demonstrated that the 3-deoxyanthocyanidin phytoalexins represent a signifi cant component of resistance against the fungal pathogen Colletotrichum sublineolum ( Lo et al. 1999 ), the causal agent of anthracnose which is a major disease affecting the worldwide production of sorghum.

Flavonoids as dietary constituents are increasingly popular because of their health benefi cial properties as antioxidant and anticancer agents. Anthocyanins from different sources have been shown to suppress cancer cell proliferation and induce apotosis ( Yi et al. 2005 , Zhang et al. 2005 ). We and other laboratories have recently reported that the 3-deoxyan-thocyanins have strong antiproliferative activities against several human cancer cell lines ( Shih et al. 2007 , Yang et al. 2009 ). In fact, they were shown to be more cytotoxic than their anthocyanidin analogs ( Shih et al. 2007 ). In addition, sorghum extracts rich in 3-deoxyanthocyanins were demonstrated to induce phase II enzymes ( Yang et al. 2009 ), which is considered as an indicator of protection against carcinogens in animal cells ( Gao et al. 2006 ). Moreover, the 3-deoxyanthocyanidins show higher color stability to pH, temperature and light changes than the anthocyanidins and their glycosides ( Awika et al. 2004 ). They have absorption maxima lower than 480 nm,

allowing them to be used as a potential replacement of artifi cial yellow and orange pigments ( Awika et al. 2004 ). Thus, the 3-deoxyanthocyanidins also represent a new source of natural colorants with nutraceutical values.

The health benefi ts of 3-deoxyanthocyanidins to both plants and humans highlight the importance of elucidating their route of biosynthesis. Flavonoid biosynthesis begins with a condensation reaction catalyzed by chalcone synthase (CHS) converting phenylpropanoid-CoA and malonyl-CoA into chalcones, followed by isomerization into fl avanones by chalcone isomerase (CHI). It is from naringenin that the anthocyanidin and 3-deoxyanthocyanidin pathways are believed to diverge from each other ( Fig. 1 ). In the antho-cyanidin pathway, fl avanone-3-hydroxylase (F3H) catalyzes the C-3 hydroxylation of fl avanones, followed by an NADPH-dependent reduction of the C-4 carbonyl group by dihydro-fl avonol 4-reductase (DFR). Removal of the resulting hydroxyl group then occurs via the anthocyanidin synthase (ANS)-catalyzed reaction. In sorghum, it has been demon-strated that fungal inoculation resulted in de novo synthesis of 3-deoxyanthocyanidins. Thus, infected seedlings incorporated [ 14 C]phenylalanine into 3-deoxyanthocyanidins within 48 h post-inoculation ( Wharton and Nicholson 2000 ). Activities of the key branch-point enzymes phenylalanine ammonia-lyase

3 x Malonyl CoA + Phenylpropanoid CoA

tt4CHS

Chalcones

O+

OH

HO

OH

R

F3H

O+

OH

HO

OH

OH

R

O

OH

HO

OH

OH

R

OH

DFR

O

OH

HO

OH

R

OH

tt5

O

OH

HO

OH

O

R

O

OH

HO

OH

O

OH

R

DihydroflavonolsFlavanones

?

DFR

Flavan-4-ols Flavan-3,4-diols

3-Deoxyanthocyanidins

34

34

3

34

3

4

3

Anthocyanidins

ANS

CHI

tt3

tt6

tt18

Fig. 1 Biosynthesis pathway of anthocyanidins and 3-anthocyanidins in sorghum. CHS is the committed enzyme for fl avonoid biosynthesis. Flavanones, which are isomerized by CHI from chalcones, are the common substrates for the formation of different classes of fl avonoids, including anthocyanidins and 3-deoxyanthocyanidins. The sequential activities of F3H, DFR and ANS convert fl avanones to anthocyanidins. A pathogen-inducible DFR gene ( SbDFR3 ) described in the present study is involved in the reduction of fl avanones to fl avan-4-ols. The fi nal step leading to 3-deoxyanthocyanin formation remains uncharacterized. Arabidopsis tt mutations in the different enzymatic steps are indicated.

1174

H. Liu et al.

Plant Cell Physiol. 51(7): 1173–1185 (2010) doi:10.1093/pcp/pcq080 © The Author 2010.


and CHS and their gene expression were induced within 6 h of fungal inoculation ( Hipskind et al. 1996 ). However, the molecular components responsible for the fi nal steps in the pathogen-inducible 3-deoxyanthocyanidin pathway in sorghum remain unidentifi ed. DFR-type reduction and ANS-type dehydrogenation/dehydroxylation reactions similar to those in the anthocyanidin pathway have been proposed for the synthesis of 3-deoxyanthocyanidins, presumably without the involvement of F3H ( Fig. 1 ). In Northern hybridization experiments using maize DFR and ANS cDNAs as probes, Hipskind et al. (1996) demonstrated that the homologous genes in sorghum were not expressed during the pathogen-inducible accumulation of 3-deoxyanthocyanidins, leading to their conclusion that 3-deoxyanthocyanidin biosynthesis does not occur via a pathway similar to that for the normal anthocyanins.

In this study, we attempted to characterize a collection of key fl avonoid structural genes (CHS, CHI, F3H, DFR and ANS) which were retrieved by bioinformatics searches against the sorghum (cultivar BTx623) genome sequence. To establish their in planta biochemical functions, the sorghum genes were transformed to Arabidopsis transparent testa ( tt ) mutants defective in the corresponding enzymatic steps for comple-mentation analysis. Their gene expression was investigated in sorghum seedlings following fungal inoculation or illumination. We have identifi ed specifi c CHS- and DFR-encoding genes which are up-regulated during the pathogen-induced 3-deoxyanthocyanidin accumulation. Recombinant enzyme assays demonstrated that both the light-inducible and pathogen-inducible DFR proteins were able to convert fl a-vanones to fl avan-4-ols, which are the immediate precursors for 3-deoxyanthocyanidins. On the other hand, the expression of anthocyanidin structural genes was found to be strongly suppressed during 3-deoxyanthocyanidin biosynthesis. Our work also revealed that 3-deoxyanthocyanidin accumulation can be induced in sorghum roots by methyl jasmonate (MeJA) but its stimulation effect is antagonized by salicylic acid (SA) treatment.

Results

Genome organization of fl avonoid structural genes in sorghum Previously, we isolated seven CHS genes ( SbCHS1–SbCHS7 ) from a sorghum genomic bacterial artifi cial chromosome (BAC) library derived from BTx623 ( Lo et al. 2002 ), the same cultivar which was used for the genome sequencing project ( Paterson et al. 2008 ). The sorghum CHS genes are highly conserved, including the coding and untranslated regions. They all have a single intron and their encoded proteins share at least 97.5 % amino acid sequence identity ( Lo et al. 2002 ). Analysis of the recently completed BTx623 genome sequence data annotated at Phytozome v4.0 ( http://www.phytozome.net/index.php ) revealed that all these genes are located on chromosome 5

within a region of approximately 152 kb ( Table 1 ). In particular, three tandem copies of SbCHS2 ( SbCHS2a , SbCHS2b and SbCHS2c ) were found in a 22 kb region. They are completely identical in their putative promoter ( > 2 kb upstream of the initiation codon), coding region, intron and the 5 ′ - and 3 ′ -untranslated regions. On the other hand, two sorghum genes ( A1-a and A1-b ) homologous to the maize DFR-encoding gene A1 were identifi ed by previous comparative mapping analyses of the sorghum and maize genomes ( Chen et al. 1997 , Chen et al. 1998 ). In the BTx623 genome, we found that A1-a ( SbDFR1 ) and A1-b ( SbDFR2 ) are located on chromosome 3 ( Table 1 ) and both genes contain three introns. Further bioinformatics searches retrieved two additional DFR genes which are located on different chromosomes: SbDFR3 (no introns) on chromosome 4 and SbDFR4 (three introns) on chromosome 9.

Sorghum genes encoding CHI, F3H and ANS have not been reported in the literature previously. Using the corresponding maize sequences, we were able to identify the homologous sequences from the BTx623 genome. For example, two genes located on chromosome 6 were found to encode F3H-like proteins ( Table 1 ). Both SbF3H1 and SbF3H2 show around 90 % sequence identity to maize F3H (NCBI accession No. AAA91227). However, SbF3H2 was not annotated as a gene locus in the Phytozome v4.0 database. We determined the coding sequence through the isolation of a full-length cDNA by reverse transcription–PCR (RT–PCR; data not shown). On the other hand, both CHI and ANS appear to be encoded by single genes: SbCHI (three introns) on chromosome 1 and SbANS (no introns) on chromosome 4, and their encoded pro-teins are 91 and 82 % identical to maize CHI (NCBI accession No. Q08704) and A2 (ANS) (NCBI accession No, CAA39022), respectively.

Expression of fl avonoid structural genes in sorghum seedlings The de novo biosynthesis of 3-deoxyanthocyanidin phytoalex-ins in sorghum is a light-independent response to attempted pathogen infection ( Weiergang et al. 1996 ). In this study, RT–PCR analysis was fi rst performed to examine the expression of the different fl avonoid structural genes described above in inoculated sorghum seedlings kept in darkness or in light-grown non-infected seedlings ( Fig. 2 ). The availability of the complete genome sequence data has allowed the design of gene-specifi c primers derived from the 3 ′ -untranslated regions for the differ-ent families of fl avonoid structural genes in sorghum. The maize fungal pathogen Cochliobolus heterostrophus , which is known to cause rapid and intense phytoalexin accumulation in sor-ghum mesocotyls ( Hipskind et al. 1996 ), was used for the inocu-lation experiments. Following light exposure, anthocyanin pigments accumulate in the mesocotyl tissues of some sor-ghum cultivars, e.g. DK 46 ( Lo and Nicholson 1998 ), but not in BTx623 ( Lo et al. 1999 ). Pigmentation phenotypes of the different sorghum cultivars used in this study are shown in Table 2 .

1175

Sorghum 3-deoxyanthocyanidin biosynthesis



http://www.phytozome.net/index.php

Table 1 Genome analysis of fl avonoid structural genes in sorghum (BTx623)

Gene name Position Gene identifi er mRNA (NCBI)

Chalcone synthase

SbCHS5 Chr 5: 49347424. .49351196 ( + ) Sb05g020150.1 XM_002449571

SbCHS4 Chr 5: 49368387. .49370616 ( − ) Sb05g020160.1 XM_002450825

SbCHS2a Chr 5: 49381777. .49383147 ( − ) Sb05g020170.1 XM_002450826

SbCHS2b Chr 5: 49392232. .49393602 ( − ) Sb05g020180.1 XM_002450827

SbCHS2c Chr 5: 49402563. .49404240 ( − ) Sb05g020190.1 XM_002450828

SbCHS1 Chr 5: 49414226. .49416106 ( − ) Sb05g020200.1 XM_002450874

SbCHS3 Chr 5: 49430465. .49432213 ( − ) Sb05g020210.1 XM_002450830

SbCHS7 Chr 5: 49473660. .49475306 ( − ) Sb05g020220.1 XM_002450831

SbCHS6 Chr 5: 49496992. .49500070 ( − ) Sb05g020230.1 XM_002450832

Chalcone isomerase

SbCHI Chr 1: 2669008. .2670703 ( + ) Sb01g003330.1 XM_002463631

Flavanone 3-hydroxylase

SbF3H1 Chr 6: 60106108. .60107732 ( − ) Sb06g031790.1 XM_002448658

SbF3H2 Chr 6: 60109782. .60111313 ( − ) Not assigned GU320740

Dihydrofl avonol 4-reductase

SbDFR1 ( A1-a ) Chr 3: 57039148. .57040668 ( + ) Sb03g028880.1 XM_002455970

SbDFR2 ( A1-b ) Chr 3: 57050278. .57051815 ( + ) Sb03g028890.1 XM_002455971

SbDFR3 Chr 4: 4074924. .4076251 ( + ) Sb04g004290.1 XM_002451541

SbDFR4 Chr 9: 4191437. .4193565 ( − ) Sb09g003710.1 XM_002440548

Anthocyanidin synthase

SbANS Chr 4: 83230. .84739 ( + ) Sb04g000260.1 XM_002451291

SbCHS1

D/I

12 24 4812 24 4812 24 48 h

L D

L D D/I

12 24 4812 24 4812 24 48 h

L

12 24 48 h

SbF3H1SbCHS2

SbF3H2SbCHS3

SbDFR1SbCHS4

SbDFR3SbCHS5

SbANS

SbCHS7

SbCHS6Actin

BTx623 DK46SbCHI

BTx623

A B

Fig. 2 Semi-quantitative RT–PCR (28 cycles) expression analysis of fl avonoid structural genes in sorghum seedlings. Four-day-old etiolated seedlings were used for the experiments. RNA samples were prepared from mesocotyl tissues collected at the indicated time points (hours) following illumination or inoculation ( C . heterostrophus ). BTx623 seedlings produce the pathogen-induced 3-deoxyanthocyanidins but do not synthesize the light-induced anthocyanin pigments. DK46 is a pigmented cultivar which accumulates anthocyanin in mesocotyls after light exposure. Actin expression was used as the internal control. L, light exposure; D, dark incubation; D/I, dark incubation and inoculation.

1176

H. Liu et al.



As shown in Fig. 2A , induced expression of all the CHS genes was detected in BTx623 mesocotyl tissues after the fungal inoculation. On the other hand, light-up-regulated expression was only detected for SbCHS5 which is expressed at slightly lower levels in the dark. Similar to SbCHS5 , expression of the single-copy gene SbCHI was induced by either light or pathogen inoculation in the BTx623 seedlings. When the genes encoding the different downstream fl avonoid enzymes (F3H, DFR and ANS) were examined, none of them was found to be expressed in the illuminated BTx623 seedlings ( Fig. 2B ), consistent with the absence of light-induced anthocyanin accumulation in the mesocotyls. In contrast, SbDFR3 expression was strongly induced in the inoculated BTx623 seedlings. On the other hand, expression of SbF3H1 , SbF3H2 , SbDFR1 and SbANS was activated in the pigmented mesocotyls of DK46 seedlings fol-lowing illumination ( Fig. 2B ). However, SbDFR2 and SbDFR4 expression was not detectable in the seedlings of either sor-ghum cultivar under our experimental conditions (data not shown). Previously, we have demonstrated that the light-induced anthocyanin pigmentation in sorghum mesocotyls was reduced by fungal inoculation ( Lo and Nicholson 1998 ). Here, we further examined the expression of the individual F3H-, DFR- and ANS-encoding genes in DK46 seedlings which were exposed to both stimuli of light and fungal inoculation simultaneously. Using real-time quantitative RT–PCR analysis, the light-inducible expression of SbF3H1 , SbF3H2 , SbDFR1 and

SbANS was found to be suppressed substantially ( > 80 % ) after fungal inoculation ( Fig. 3 ). In contrast, the pathogen-inducible SbDFR3 expression was not affected in the light-grown inocu-lated DK46 seedlings which produced the 3-deoxyanthocyanidin phytoalexins. Expression of SbDFR2 and SbDFR4 was not detected (data not shown).

Induction of 3-deoxyanthocyanidin production and SbDFR3 expression in sorghum roots by MeJA During our investigations on the effects of different plant signaling compounds on defense gene expression, we observed the appearance of orange pigmentation in roots of the sorghum cultivar DK18 seedlings within 24 h of MeJA treatment ( Fig. 4A ). Methanol extractions were then performed on the root tissues for metabolite analysis. As shown in Fig. 4B , the accumulation of apigeninidin in the MeJA-treated roots was confi rmed by tandem mass spectrometry (MS/MS) analysis. Interestingly, the MeJA-induced 3-deoxyanthocyanidin accu-mulation was suppressed by co-treatment with SA ( Fig. 4A ). The expression levels of the genes encoding the fl avonoid downstream enzymes (F3H, DFR and ANS) were then examined in seedling roots following the different chemical treatments. RT–PCR experiments revealed the induction of SbDFR3 expression in the MeJA-treated, pigmented roots ( Fig. 4C ). In contrast, expression of SbF3H1 , SbF3H2 , SbDFR1 , SbDFR2 , SbDFR4 and SbANS was not activated under the same

Table 2 Pigmentations in seedlings of the different sorghum cultivars analyzed in this study

Treatment BTx623 DK46 DK18

Light None Anthocyanin a None

Dark None None None

Inoculation (light) 3-Deoxyanthocyanidins a Anthocyanin a (reduced) + 3-deoxyanthocyanidins a 3-Deoxyanthocyanidins a

Inoculation (dark) 3-Deoxyanthocyanidins a 3-Deoxyanthocyanidins a 3-Deoxyanthocyanidins a

MeJA root treatment None None 3-Deoxyanthocyanidins b

All seedlings were planted in the dark to allow elongation of mesocotyls before light and inoculation treatments. a Detected in mesocotyls. b Detected in roots.

0

20

40

60

80

100

120

SbF3H1SbDFR1SbANS SbDFR3SbF3H2

Rel

ativ

e ex

pre

ssio

nL

evel

s (%

)

Light (24h) Light + inoculation (24h)

Fig. 3 Real-time RT–PCR analysis of relative expression levels of sorghum genes encoding the fl avonoid downstream enzymes (F3H, DFR and ANS) in inoculated DK46 seedlings placed under light. Mesocotyl tissues were collected for RNA extraction after 24 h of illumination/inoculation.

1177




conditions (data not shown). On the other hand, SbDFR3 expression was not detectable in roots which were treated with both MeJA and SA ( Fig. 4C ), consistent with the SA suppression of MeJA-induced 3-deoxyanthocyanidin accumu-lation ( Fig. 4A ).

Complementation of Arabidopsis fl avonoid mutants by the sorghum fl avonoid genes The Arabidopsis tt mutants are defi cient in different fl avonoid enzyme activities ( Fig. 1 ), resulting in the lack of fl avonoid accumulation in different tissues. The in planta biochemical function of SbCHS2 was previously established by complemen-tation analysis in transgenic tt4 mutants ( Yu et al. 2005 ). Due to their high degree of sequence identity ( > 97.5 % ), all the other SbCHS genes are also expected to encode functional CHS enzymes. In this study, the coding regions of the other fl avonoid structural genes which showed expression following illumina-tion or fungal inoculation ( Fig. 2 ) were placed under the control of the 35S promoter and transformed into the appro-priate Arabidopsis tt mutants. Transgene expression was con-fi rmed by RT–PCR in 14-day-old T 1 seedlings (data not shown) and three independent lines for each transgene were selected for phenotypic investigations. Low nitrogen is a stress condition known to induce the anthocyanin biosynthesis pathway in Arabidopsis wild-type seedlings ( Dong et al. 2001 ), but the different tt mutants fail to produce any pigments under such conditions. As shown in Fig. 5 , the anthocyanin pigmentation in tt5 , tt6 , tt3 and tt18 mutants was restored by the overexpres-sion of SbCHI , SbF3H1 or SbF3H2 , SbDFR1 or SbDFR3 , and SbANS , respectively. Hence, our transgenic complementation analysis

demonstrated that these sorghum genes all encode functional fl avonoid enzymes in planta.

During the phenotypic investigations of the transgenic tt3 plants, more intense anthocyanin pigmentation was observed in the lines overexpressing SbDFR1 than in those overexpressing SbDFR3 ( Fig. 5 ). Methanol extracts were then prepared from 2-week-old seedlings of three independent transformant lines for each sorghum transgene. Acid hydrolysis was performed to release the anthocyanidin (cyanidin) for quantifi cation. As shown in Fig. 6 , the tt3 + SbDFR1 plants accumulated approxi-mately three times the amount of anthocyanin detected in the tt3 + SbDFR3 plants, suggesting that SbDFR1 is a more effective DFR enzyme for the biosynthesis of anthocyanin pigments in planta. On the other hand, no 3-deoxyanthocyanidins were detected in all the transgenic Arabidopsis extracts.

Enzyme activities of SbDFR1 and SbDFR3 From the gene expression analyses ( Figs. 2–4 ), it is apparent that there is a specifi c DFR (i.e. SbDFR3) for the production of 3-deoxyanthocyanidin in sorghum seedlings following fungal inoculation and MeJA treatment. On the other hand, a different DFR (i.e. SbDFR1) is likely to be involved in the light-dependent anthocyanin pigmentation. We then investigated the enzyme activities of the two SbDFR proteins which were overexpressed in bacteria. The coding regions of SbDFR1 and SbDFR3 were subcloned into the pET28a vector and transformed into the Escherichia coli strain BL21. Soluble crude protein extracts were subsequently prepared from isopropyl- β - D -thiogalactopyranoside (IPTG)-induced cell cultures and the (His 6 )-tag SbDFR proteins purifi ed for enzyme activity assays.

100 140 180 220 260 300m/z, amu

4.0e5

8.0e5

1.2e6

1.6e6

Intensity, cps

157.1171.1

255.2128.1

115.0 152.2143.3

121.2 181.2 227.0

apigeninidin

H2O SA SA +MeJA

MeJA

1 cm

12 24 12 121224 24 24

H2O SA SA+MeJA MeJA

Actin

SbDFR3

A B

C

Fig. 4 MeJA induction of 3-deoxyanthocyanidin accumulation in sorghum roots. (A) Sorghum cultivar DK18 seedlings were treated with the indicated chemicals. Note the orange pigmentation in the MeJA-treated root and the repression of pigmentation in the JA/MeJA-treated root. (B) Metabolite analysis of MeJA-treated root extracts confi rmed the accumulation of apigeninidin. The MS/MS spectrum shown is identical to that obtained for an authentic standard (not shown). (C) RT–PCR (28 cycles) analysis of SbDFR3 expression in sorghum roots following the different chemical treatments. Actin expression was used as the internal control.

1178

H. Liu et al.



The recombinant proteins were tested on several common dihydrofl avonol and fl avanone substrates in the presence of NADPH. The enzyme activities were assayed at their optimum pH values and temperatures. The direct reaction products, fl avan-3,4-diols (from dihydrofl avonols) and fl avan-4-ols

0

5

10

15

20

25

30

Cya

nid

in (

mg/g

FW

)

NT 1 2 3

tt3 + SbDFR3

tt3 + SbDFR3

NT 1 2 3

tt3 + SbDFR1

tt3 + SbDFR1

A

B

Fig. 6 (A) Quantitative analysis of anthocyanidin in acid-hydrolyzed extracts of transgenic Arabidopsis tt3 plants. Cyanidin was the major anthocyanidin detected in the plant samples. Data are expressed as average values of three transformant lines for each transgene, and error bars represent standard deviation. FW, fresh weight. (B) Semi-quantitative RT–PCR (27 PCR cycles) expression analyses of the SbDFR1 and SbDFR3 transgenes in the transformant lines used above.

tt5 tt5 + SbCHI tt6 tt6 + SbF3H2tt6 + SbF3H1

1 mm

tt3 tt3 + SbDFR3tt3 + SbDFR1 tt18 tt18 + SbANS

Fig. 5 Complementation analysis of sorghum fl avonoid structural genes in transgenic Arabidopsis tt mutants. Anthocyanin pigmentation was restored in the transgenic tt5 , tt6 , tt3 and tt18 mutants overexpressing SbCHI , SbF3H1 or SbF3H2 , SbDFR1 or SbDFR3 , and SbANS , respectively. Seedlings were grown in low nitrogen medium to induce the anthocyanin biosynthesis pathway.

(from fl avanones), were converted by acid treatments to anthocyanidins and 3-deoxyanthocyanidins, respectively ( Shimada et al. 2005 ), for liquid chromatography–mass spec-trometry (LC-MS) detection and quantifi cation. In general, both SbDFR1 and SbDFR3 were found to show substantially stronger activities toward dihydrofl avonols (dihydrokaemp-ferol and dihydroquercetin) than fl avanones (naringenin and eriodictyol) ( Fig. 7 ). Nevertheless, the presence of 3-deoxyan-thocyanidins in the acid-treated fl avanone reactions was positively confi rmed by their retention times and MS/MS spectra. In all cases, anthocyanidins or 3-deoxyanthocyanidins were not detectable by LC-MS when no enzymes or boiled enzymes were included in the reactions.

Sequence analysis of DFR proteins The four sorghum DFR proteins share at least 82 % sequence identity. Phylogenetic analysis showed that DFR sequences from different plant species are separated according to their taxonomic groups ( Fig. 8A ). Thus, the sorghum sequences are closely related to the maize and rice DFRs. In addition, sequence alignment revealed the highly conserved N-terminal NADPH-binding region in all of the sorghum DFRs ( Supple-mentary Fig. S1 ). The third residue (asparagine) within a 26 amino region proposed to be important for DFR substrate determination ( Beld et al. 1989 ) was also identifi ed in SbDFR1, SbDFR2 and SbDFR4 ( Fig. 8B ). DFR proteins having an aspartate residue in this position have been found to reduce dihydro kaempferol less effectively than the asparagine-type

1179




DFRs (Xie et al. 2004, Shimada et al. 2005 ). SbDFR3 contains a serine residue in this position ( Fig. 8B ) and was also shown to have slightly lower activities toward dihydrokaempferol when compared with SbDFR1 ( Fig. 7 ).

Discussion

Sorghum is a valuable crop best known for its C4 metabolism, drought tolerance and potential as a biofuel source ( Paterson 2008 ). The plant is also a rich source of distinct natural prod-ucts, including dhurrin, sorgoleone and 3-deoxyanthocyanidins. The 3-deoxyanthocyanidins are rare pigmented fl avonoids accumulated in fl oral tissues of sinningia, maize and sorghum. However, pathogen-inducible biosynthesis of 3-deoxyanthocy-anidin phytoalexins occurs only in sorghum. Recent completion of the sorghum genome sequencing project has allowed us to identify specifi c fl avonoid structural gene members which are responsive to inoculation with C. heterostrophus , a fungus known to elicit rapid and intense 3-deoxyanthocyanidin production in sorghum seedlings.

Analysis of the sorghum genome revealed the presence of nine CHS genes which are all located on chromosome 5, and three of them ( SbCHS2a–SbCHS2c ) are identical tandem repeats. In contrast to the large CHS gene family in sorghum, only two functional CHS genes ( C2 and Whp ) were reported in maize ( Franken et al. 1991 ). A common feature of gene families is that individual members are differentially regulated in response to environmental or developmental signals. Using the sorghum mesocotyl inoculation system, we demonstrated that the expressions of most gene members, i.e. SbCHS1–

SbCHS4 , SbCHS6 and SbCHS7 , were only inducible by pathogen inoculation under our experimental conditions. On the other hand, SbCHS5 was up-regulated following either light exposure or fungal inoculation. CHS is a plant-specifi c polyketide synthase that catalyzes the committed reaction step leading to the formation of different fl avonoid metabolites ( Dixon and Paiva 1995 ). Multiplications of the pathogen-inducible CHS genes in sorghum may refl ect an evolutionary adaptation for the de novo synthesis of defense-related fl avonoid metabolites.

O

OH

HO

OH

O

OH

R

O

OH

HO

OH

OH

R

OH

O+

OH

HO

OH

OH

R

Dihydroflavonols(Dihydrokaempferol, R=H; Dihydroquercetin, R=OH)

Flavan-3,4-diols

DFR Acidtreatment

Flavan-4-ols

Anthocyanidins(Pelargonidin, R=H;Cyanidin, R=OH)

O

OH

HO

OH

O

R

O+

OH

HO

OH

R

O

OH

HO

OH

R

OH

3-Deoxyanthocyanidins(Apigeninidin, R=H;Luteolinidin, R=OH)

DFR Acidtreatment

Flavanones(Naringenin, R=H; Eriodictyol, R=OH)

Dihydrokaempferol

Dihydroquercetin

Naringenin

Eriodictyol

SbDFR1 SbDFR3

66.4 ± 0.61

31.1 ± 1.9

20.0 ± 1.6 (×10-3)

94.1 ± 22.3 (×10-3)

50.9 ± 2.5

14.9 ± 1.5

21.6 ± 9.1 (×10-3)

102.3 ± 5.7 (×10-3)

Substrate Enzyme activity (nmol mg-1min-1)

Fig. 7 Enzyme activities of purifi ed recombinant SbDFR1 and SbDFR3 proteins. After acid treatment, reaction products from dihydrofl avonols and fl avanones were converted to anthocyanidins and 3-deoxyanthocyanidins, respectively, for quantifi cation in enzyme assays. Both proteins showed substantially higher activities toward dihydrofl avonols than fl avanones. No anthocyanidins or 3-deoxyanthocyanidins were detected in reactions containing no enzymes or boiled enzymes.

1180

H. Liu et al.



The biosynthesis of anthocyanin pigments is one of the best characterized secondary metabolic pathways in plants. Starting with fl avanones as substrates, the production of anthocyanidins takes place through sequential reactions cata-lyzed by F3H, DFR and ANS. Similarly, light-induced expression of SbF3H1 , SbF3H2 , SbDFR1 and SbANS was detected in sor-ghum DK46 mesocotyls during the accumulation of anthocya-nin pigments ( Fig. 2 ). All these genes were demonstrated to encode functional fl avonoid enzymes by transgenic analysis in the corresponding Arabidopsis tt mutants ( Fig. 5 ). On the other hand, reduction of F3H activity has been suggested to be a key factor for directing the fl ow of carbon to 3-deoxyfl avonoid biosynthesis ( Halbwirth et al. 2003 , Winefi eld et al 2005 ). Consistently, expression of the functional F3H genes, SbF3H1 and SbF3H2 , was not detected in sorghum seedlings during the light-independent fungal- or MeJA-induced

3-deoxyanthocyanidin biosynthesis ( Figs, 2 , 4 ). In addition, these light-inducible genes were transcriptionally down-regulated when DK46 seedlings were exposed to both stimuli of light and fungal inoculation ( Fig. 3 ). Under the above circumstances, the levels of dihydrofl avonols, which are con-verted from fl avanones by F3H, are expected to be effectively reduced. This is particularly important since the sorghum DFR proteins were found to have substantially higher in vitro activities for dihydrofl avonols than fl avanones ( Fig. 7 ).

Recombinant SbDFR1 and SbDFR3 showed low but detect-able NADPH-dependent activities toward naringenin and eriodictyol ( Fig. 7 ). Hence, biochemically, both sorghum enzymes are able to participate in 3-deoxyanthocyanidin biosynthesis through the reduction of fl avanones, although they may not be the most preferred substrates under normal physiological conditions. On the other hand, it is clearly

MtDFR1

Lotus DFR2

Lotus DFR1

MtDFR2

Lotus DFR5

Malus DFR

Arabidopsis TT3

Sinningia DFR

OsDFR1

Maize A1

SbDFR3

SbDFR1

SbDFR2

SbDFR4

99

48

99

100

73

52

100

37

36

30

96

0.05

B

A

Fig. 8 Sequence analysis of DFR proteins from sorghum and other plants. (A) An unrooted phylogenetic tree of different DFR proteins. The scale bar represents 0.05 substitutions per site. Mt, M. trunculata ; Os, rice. (B) Partial alignment of the sorghum DFR proteins and selected sequences from the above tree. The 26 amino acid region proposed to determine substrate specifi city is boxed. In the M. trunculata and lotus proteins, the indicated residue (N or D) is important for the utilization of DHK. The N-type DFRs were reported to reduce dihydroquercetin more effectively than the D-type. Most of the sorghum sequences are N-type DFRs, except for SbDFR3 which contains a serine residue in this position.

1181




evident that SbDFR3 is specifi cally responsive to conditions that stimulate 3-deoxyanthocyanidin biosynthesis in sorghum, such as fungal inoculation ( Figs. 2 , 3 ) or MeJA treatment ( Fig. 4 ). Evidence for the existence of a multienzyme assembly in fl avonoid biosynthesis has been provided for Arabidopsis ( Winkel 2004 ). Hence, the sorghum pathogen/stress-inducible fl avonoid enzymes, without the involvement of F3H, may form a macromolecular complex that facilitates the conversion of fl avanones to fl avan-4-ols by SbDFR3, thereby allowing more effi cient metabolite channeling for 3-deoxyanthocyandin phy-toalexin production. The remaining step for the conversion of fl avan-4-ols to 3-deoxyanthocyanidins is still uncharacterized. The corresponding reaction for anthocyanidin formation is catalyzed by ANS, which is a member of the family of 2-oxoglutarate-dependent oxygenases (DOXs) ( Winkel-Shirley 2001 ). However, since the single-copy ANS gene ( SbANS ) is not expressed during fungal or MeJA-induced 3-deoxyantho-cyanidin accumulation, an independent pathogen- or stress-responsive DOX gene with low homology to SbANS may instead be involved in the fi nal step of the 3-deoxyanthocyaindin pathway.

There have been very few reports on the regulation of fl avonoid biosynthesis in sorghum. In maize, P1 is a MYB regulatory protein controlling the accumulation of 3-deoxyfl a-vonoids in fl oral tissues. Thus, P1 activates a subset of fl avonoid structural genes encoding CHS, CHI and ANS, but not F3H ( Grotewold et al. 1994 , Grotewold et al. 1998 ). Recently, Y1 was reported to represent a sorghum P1 homolog which deter-mines pericarp pigmentation resulting from phlobaphene and 3-deoxyanthocyanidin accumulation ( Boddu et al. 2006 ). The cultivar BTx623 harbors a deletion Y1 allele and produces seeds without these 3-deoxyfl avonoid pigments ( Boddu et al. 2005 ). However, since inoculated BTx623 plants are able to synthesize apigeninidin and its methyl ether ( Lo et al. 1999 ), a different regulatory mechanism is likely to be responsible for the activation of 3-deoxyanthocyanidin phytoalexin pro-duction. In this study, we also demonstrated that the patho-gen-specifi c 3-deoxyanthocyanidin accumulation and SbDFR3 expression could be induced in roots by MeJA treatment but its effects were antagonized by SA ( Fig. 4 ). Similarly, SA and MeJA co-treatment was previously shown to reduce the fold induction of 18 MeJA-induced genes by > 25 % in sorghum seed-lings within 27 h ( Salzman et al. 2005 ). The antagonistic effect of SA on jasmonate (JA)-inducible genes is well documented in Arabidopsis. For example, expression of the JA-responsive marker genes PDF1.2 , LOX2 and VSP2 was strongly suppressed by biological or chemical induction of the SA pathway ( Koornneef et al. 2008 ). Cross-talk between SA and JA signaling pathways is believed to play a key role in the fi ne-tuning of inducible defenses in plants ( Reymond and Farmer 1998 ). Whether JA and SA are also involved as a signal for regulating the pathogen-inducible 3-deoxyanthocyanidin biosynthesis in sorghum remains to be elucidated. Further dissection of the remaining enzymatic step and the molecular regulatory network in 3-deoxyanthocyanidin biosynthesis will be facilitated

by our identifi cation of the pathogen-inducible SbDFR3 gene and the recent publication of the complete sorghum genome sequence.

Materials and Methods

Sorghum growth, fungal inoculation and chemical treatment conditions Seeds of the sorghum cultivars (BTx623, DK46 and DK18) used in this study were former collections of the late Professor Ralph L. Nicholson of Purdue University (West Lafayette, IN, USA). Sorghum seeds were planted in rolls of germination paper and kept in the dark for 4 d as described previously ( Lo et al. 1996 ). Etiolated seedlings were then placed under constant light at room temperature. For inoculation experiments, plants were sprayed with conidial suspensions of C . heterostrophus at 5.5 × 10 4 conidia ml − 1 with Tween-20 as a wetting agent (100 µl 100 ml − 1 ). The inoculated plants were incubated at 100 % relative humidity (24 h) and kept in the dark at room temperature. For chemical treatments, the rolls of sorghum seedlings (4 d old) were placed in distilled water supplemented with SA, MeJA or SA + MeJA. Stock solutions of SA and MeJA (Sigma) were added to fi nal concentrations of 1 mM and 100 µM, respectively, as described previously for treatments of sorghum seedlings at similar stages ( Salzman et al. 2005 ).

RT–PCR experiments Total RNA was extracted from sorghum mesocotyl tissues using the Trizol method (Invitrogen, CA, USA). DNase I-treated RNA samples (4 µg) were reversed transcribed by M-MLV reverse transcriptase (Promega, WI, USA). Gene-specifi c primers were designed from the 3 ′ -untranslated regions of the different sorghum fl avonoid structural genes. Amplifi cation of target cDNA for semi-quantitative analysis was performed with the GoTaq Flexi DNA Taq polymerase (Promega, WI, USA) using the following program: 94 ° C (10 min); 28 cycles of 94 ° C (30 s), 55 ° C (30 s), 72 ° C (1 min); and 72 ° C (7 min). To obtain full-length coding regions of SbCHI , SbF3H1 , SbF3H2 , SbDFR1 , SbDFR3 and SbANS for the complementation analyses described below, Pfx DNA polymerase (Invitrogen) was used for high-fi delity amplifi cations using the same program.

For quantitative gene expression analysis, real-time PCR was performed with the DNA Masterplus SYBR Green I kit (Roche Applied Science) using the LightCycler Instrument (Roche). The PCR program was as follows: pre-incubation (95 ° C for 10 min), amplifi cation (95 ° C for 10 s, 60 ° C for 9 s, 72 ° C for 4 s with a single fl uorescence measurement, 45 cycles), melting curve analysis (56–95 ° C with a heating rate of 0.1 ° C s − 1 and a continuous fl uorescence measurement) and cooling (40 ° C for 30 s). PCR products were confi rmed by agarose gel (2 % ) electrophoresis and their identities verifi ed by sequenc-ing. Experiments were performed in triplicate for each sample. Relative expression ratios of the target genes were calculated essentially as described previously ( Pfaffl et al. 2002 ).

1182

H. Liu et al.



A list of primers used in the different RT–PCR experiments described above is given in Supplementary Table S1 .

Arabidopsis complementation analysis The Arabidopsis tt3 , tt5 , tt6 and tt18 mutants (Arabidopsis Biological Resources Center, Columbus, OH, USA) defective in genes encoding DFR, CHI, F3H and ANS, respectively, were used for transgenic complementation analyses. Coding regions of the selected sorghum fl avonoid genes were cloned into an overexpression vector ( Yu et al. 2005 ) containing the cauli-fl ower mosaic virus (CaMV) 35S promoter and the nopaline synthase 3 ′ terminator. The resulting plasmids were introduced into the binary vector pCAMBIA 1300 (CAMBIA, Australia). Agrobacterium -mediated transformation of the appropriate Arabidopsis mutants was performed by the fl oral dip method ( Clough and Bent 1998 ). Harvested seeds were surface-sterilized and germinated on Murashige and Skoog (MS) (Sigma) agar plates containing 3 % (v/v) sucrose and 25 µg ml − 1 hygromycin (Sigma). Resistant seedlings were transplanted and placed in a growth chamber (22 ° C; 16 h light, 8 h dark). At least three independent lines with strong transgene expression for each construct were selected for phenotypic analysis. To induce anthocyanin accumulation, seeds were germinated on MS plates without nitrogen sources. For analysis of anthocyanidins by HPLC (see below), plant tissues (0.5–1.0 g) were collected from 7-day-old T 1 seedlings, extracted in 100 % methanol, followed by acid hydrolysis in 1 % hydrochloric acid at 95 ° C for 30 min.

Heterologous expression of recombinant SbDFR proteins in bacteria Full-length coding regions of SbDFR1 and SbDFR3 amplifi ed from sorghum cDNA prepared as described above were subcloned in-frame with the (His) 6 -tag in the pET28a vector (Novagen). The fi nal constructs were transformed into E. coli BL21 competent cells. SbDFR protein expression was induced by the addition of 1 mM IPTG to the bacterial cultures for 2 h at 25 ° C. Escherichia coli cells were harvested by centrifugation at 9,400 × g for 3 min, followed by washing in 0.1 M Tris–HCl (pH 7.5) containing 10 mM EDTA and 50 mM sucrose. The cell pellet was then suspended in 0.1 M potassium phosphate buffer (pH 7.0) containing 10 % (w/v) sucrose, sonicated in ice, and centrifuged to collect the supernatant for purifi cation through an affi nity column of Ni-NTA agarose (Qiagen) according to the manufacturer’s instructions. After elution, the (His) 6 -tag DFR fusion proteins were concentrated by centrifugation in Microcon YM-3 (3,000 Da molecular weight cut-off) spin columns (Millipore). Protein purity was examined by SDS–gel electrophoresis.

In vitro enzyme assay of recombinant SbDFR1 and SbDFR3 For enzyme assays, the reaction mixtures contained 0.1 M Tris–HCl buffer (pH 7.0), 2 mM NADPH, 100 µM dihydrofl a-vonol (dihydrokaempferol and dihydroquercetin, Sigma)

or fl avanone (naringenin and eriodictyol, Sigma) substrates in fi nal volumes of 500 µl. Reactions were initiated by the addi-tion of 10 µg of purifi ed proteins and terminated by extraction with 500 µl of ethyl acetate (twice) after incubation at 25 ° C for 1 h. The ethyl acetate extracts were evaporated in vacuo at room temperature. The residues were dissolved in 100 µl of 5 % hydrochloric acid in n -butanol and boiled for 1 h. Filtered samples (20 µl) were injected onto a HP 1100 series HPLC system (Agilent Technologies) connected to a Nucleosil 100-5 C18 column (5 µm, 150 × 2 mm, Agilent Technologies). Separation was performed using a solvent system of 0.5 % formic acid (v/v) (A) and acetonitrile (B) with a gradient of 15–60 % B over 25 min at a fl ow rate of 0.2 ml min − 1 . The LC elution was analyzed by an on-line API2000-QTrap mass spectrometer (Applied Biosystems) operating in multiple-reaction monitoring (MRM) mode (positive ionization). Transition reactions of pelargonidin (271.2 → 121.1), cyanidin (287.2 → 137.1), apigeninidin (255.2 → 171.2) and luteolinidin (271.2 → 115.1) were monitored and the LC-MRM peak areas used for concentration determination. Authentic standards of anthocyanidins and 3-deoxyanthocyanidins (Extrasynthese) were used for compound identifi cation and quantifi cation.

Sequence analysis Sequences of different DFR proteins retrieved from GenBank were aligned by the ClustalW method ( www.ebi.ac.uk/clustalw ). The alignment was used to construct a phylogenetic tree by the Neighbor–Joining method in the MEGA4 program using the default parameters from the results of 1,000 bootstrap replicates ( Kumar et al. 2004 ). Gene accession numbers for the amino acid sequences of DFRs are: Lotus japonicus DFR1, BAE19948; L. japonicus DFR2, BAE19949; L. japonicus DFR5, BAE19953; Arabidopsis TT4, BAA85261; Sinningia cardinalis DFR, AAR01565; rice OsDFR1, BAA36183; Medicago trunculata MtDFR1, AAR27014; M. trunculata MtDFR2, AAR27015; maize A1, P51108; Malus domestica DFR, AAO39816.

Supplementary data

Supplementary data are available at PCP online.

Funding

This work was supported by the Research Grants Council of the Hong Kong Special Administrative Region, China [grant Nos. HKU7527/06M and HKU 3/07C]; the HKU Seed Funding Program [200911159094].

Acknowledgments

The authors thank Dr J. S. T. Tsang (Biological Sciences, The University of Hong Kong) for access to their Roche LightCycler machine.

1183




References

Awika , J.M. , Rooney , L.W. and Waniska , R.D. ( 2004 ) Properties of 3-deoxyanthocyanins from sorghum . J. Agric. Food. Chem. 52 : 4388 – 4394 .

Beld , M. , Martin , C. , Hutis , H. , Stuitje , A.R. and Gerats , A.G. ( 1989 ) Flavonoid synthesis in Petunia hydrida : partial characterization of dihydrofl avonol 4-reductase genes . Plant Mol. Biol. 13 : 491 – 502 .

Boddu , J. , Jiang , C. , Sangar , V. , Olson , T. , Peterson , T. and Chopra , S. ( 2006 ) Comparative structural and functional characterization of sorghum and maize duplications containing orthologous myb transcription regulators of 3-deoxyfl avonoid biosynthesis . Plant

Mol. Biol. 60 : 185 – 199 . Boddu , J. , Svabek , C. , Ibraheem , F. , Jones , A.D. and Chopra , S. ( 2005 )

Characterization of a deletion allele of a sorghum Myb gene yellow seed1 showing loss of 3-deoxyfl avonoids . Plant Sci. 169 : 542 – 552 .

Chen , M. , SanMiguel , P. and Bennetzen , J.L. ( 1998 ) Sequence organization and conservation in sh2/a1 -homologous regions of sorghum and rice . Genetics 148 : 435 – 443 .

Chen , M. , SanMiguel , P. , de Oliveira , A.C. , Woo , S.S. , Zhang , H. , Wing , R.A. , et al . ( 1997 ) Microcolinearity in sh2-homologous regions of the maize, rice, and sorghum genomes . Proc. Natl Acad. Sci. USA 947 : 3431 – 3435 .

Clough , S.J. and Bent , A.F. ( 1998 ) Floral dip: a simplifi ed method for Agrobacterium- mediated transformation of Arabidopsis thaliana . Plant J. 16 : 735 – 743 .

Dong , X. , Braun , E.L. and Grotewold , E ( 2001 ) Functional conservation of plant secondary metabolic enzymes revealed by complementation of Arabidopsis fl avonoid mutants with maize genes . Plant Physiol. 127 : 46 – 57 .

Dixon , R.A. and Paiva , N.L. ( 1995 ) Stress-induced phenylpropanoid metabolism . Plant Cell 7 : 1085 – 1097 .

Franken , P. , Niesbach-Klösgen , U. , Weydemann , U. , Maréchal-Drouard , L. , Saedler , H. and Wienand , U. ( 1991 ) The duplicated chalcone synthase genes C2 and Whp (white pollen) of Zea mays are independently regulated; evidence for translational control of Whp expression by the anthocyanin intensifying gene in . EMBO J. 10 : 2605 – 2612 .

Gao , J. , Kashfi , K. , Liy , X. and Rigas , B. ( 2006 ) NO-donating aspirin induces phase II enzymes in vitro and in vivo . Carcinogenesis 27 : 803 – 810 .

Grotewold , E. , Chamberlin , M. , Snook , M. , Siame , B. , Butler , L. , Swenson , J. , et al . ( 1998 ) Engineering secondary metabolism in maize cells by ectopic expression of transcription factors . Plant Cell 10 : 721 – 740 .

Grotewold , E. , Drummond , B.J. , Bowen , B. and Peterson , T. ( 1994 ) The myb-homologous P gene controls phlobaphene pigmentation in maize fl oral organs by directly activating a fl avonoid biosynthetic gene subset . Cell 76 : 543 – 553 .

Halbwirth , H. , Martens , S. , Wienand , U. , Forkman , G. and Stich , K. ( 2003 ) Biochemical formation of anthocyanins in silk tissue of Zea mays . Plant Sci. 164 : 489 .

Hipskind , J.D. , Goldsbrough , P.B. , Urmeev , F. and Nicholson , R.L. ( 1996 ) Synthesis of 3-deoxyanthocyanidin phytoalexins in sorghum does not occur via the same pathway as 3-hydroxylated anthocyanidins and phlobaphenes . Maydica 41 : 155 – 166 .

Koornneef , A. , Leon-Reyes , A. , Ritsema , T. , Verhage , A. , Den Otter , F.C. , Van Loon , L.C. , et al . ( 2008 ) Kinetics of salicylate-mediated suppression of jasmonate signaling reveals a role for redox modulation . Plant Physiol. 147 : 1358 – 1368 .

Kumar , S. , Tamura , K. and Nei , M. ( 2004 ) MEGA3: integrated software for molecular evolutionary genetics analysis and sequence alignment . Brief Bioinform. 5 : 150 – 163 .

Lo , C. , Coolbaugh , R.C. and Nicholson , R.L. ( 2002 ) Molecular characterization and in silico expression analysis of a chalcone synthase gene family in sorghum . Physiol. Mol. Plant Pathol. 61 : 179 – 188 .

Lo , S.C.C. , de Verdier , K. and Nicholson , R.L. ( 1999 ) Accumulation of 3-deoxyanthocyanidin phytoalexins and resistance to Colletotrichum sublineolum in sorghum . Physiol. Mol. Plant Pathol. 55 : 263 – s273 .

Lo , S.C.C. and Nicholson , R.L. ( 1998 ) Reduction of light-induced anthocyanin accumulation in inoculated sorghum mesocotyls: implications for a compensatory role in the defense response . Plant Physiol. 116 : 979 – 989 .

Lo , S.C. , Weiergang , I. , Bonham , C. , Hipskind , J. , Wood , K. and Nicholson , R.L. ( 1996 ) Phytoalexin accumulation in sorghum: identifi cation of a methyl ether of luteolinidin . Physiol. Mol. Plant

Pathol. 49 : 21 – 31 . Paterson , A.H. ( 2008 ) Genomics of sorghum . Int. J. Plant Genomics

Article ID 362451 . Pfaffl , M.W. , Horgan , G.W. and Dempfl e , L. ( 2002 ) Relative expression

software tool (REST) for group-wide comparison and statistical analysis of relative expression results in real-time PCR . Nucleic Acids

Res. 30 : c36 . Reymond , P. and Farmer , E.E. ( 1998 ) Jasmonate and salicylate as

global signals for defense gene expression . Curr. Opin. Plant Biol. 1 : 404 – 411 .

Salzman , R.A. , Brady , J.A. , Finlayson , S.A. , Buchanan , C.D. , Summer , E.J. , Sun , F. , et al . ( 2005 ) Transcriptional profi ling of sorghum induced by methyl jasmonate, salicylic acid, and aminocyclopropane carboxylic acid reveals cooperative regulation and novel gene responses . Plant Physiol. 138 : 352 – 368 .

Shih , C.H. , Siu , S.O. , Ng , R. , Wong , E. , Chiu , L.C.M. , Chu , I.K. , et al . ( 2007 ) Quantitative analysis of anticancer 3-deoxyanthocyanidins in infected sorghum seedlings . J. Agric. Food Chem. 55 : 254 – 259 .

Shimada , N. , Sasaki , R. , Sato , S. , Kaneko , T. , Tabata , S. , Aoki , T. , et al . ( 2005 ) A comprehensive analysis of six dihydrofl avonol 4-reductases encoded by a gene cluster of the Lotus japonicus genome . J. Exp. Bot. 56 : 2573 – 2585 .

Snyder , B.A. and Nicholson , R.L. ( 1990 ) Synthesis of phytoalexins in sorghum as a site-specifi c response to fungal ingress . Science 248 : 1637 – 1639 .

Weiergang , I. , Hipskind , J.D. and Nicholson , R.L. ( 1996 ) Synthesis of 3-deoxyanthocyanidin phytoalexins in sorghum occurs independent of light . Physiol. Mol. Plant Pathol. 49 : 377 – 388 .

Wharton , P. and Nicholson , R.L. ( 2000 ) Temporal synthesis and radiolabelling of the sorghum 3-deoxyanthocyanidin phytoalexins and the anthocyanin, cyanidin 3-dimalonyl glucoside . New Phytol. 145 : 457 – 469 .

Winefi eld , C.S. , Lewis , D.H. , Swinny , E.E. , Zhang , H. , Arathoon , H.S. , Fischer , T.C. , et al . ( 2005 ) Investigation of the biosynthesis of 3-deoxyanthocyanins in Sinningia cardinalis . Physiol. Plant. 124 : 419 – 430 .

Winkel , B.S.J. ( 2004 ) Metabolic channeling in plants . Annu. Rev. Plant

Biol. 55 : 85 – 107 . Winkel-Shirley , B. ( 2001 ) It takes a garden. How work on diverse plant

species has contributed to an understanding of fl avonoid metabolism . Plant Physiol. 127 : 1399 – 1404 .

1184

H. Liu et al.



Yang , L. , Browning , J.D. and Awika , J.M. ( 2009 ) Sorghum 3-deoxyantho-cyanins possess strong phase II enzyme inducer activity and cancer growth inhibition properties . J. Agric. Food Chem. 57 : 1797 – 1801 .

Yi , W. , Fischer , J. , Krewer , G. and Akoh , C.C. ( 2005 ) Phenolic compounds from blueberries can inhibit colon cancer cell proliferation and induce apoptosis . J. Agric. Food Chem. 53 : 7320 – 7329 .

Yu , C.K.Y. , Springob , K. , Schmidt , J. , Nicholson , R.L. , Chu , I.K. , Yip , W.K. , et al . ( 2005 ) A stilbene synthase gene ( SbSTS1 ) is involved in host and nonhost defense responses in sorghum . Plant Physiol. 138 : 393 – 401 .

Zhang , Y. , Vareed , K. and Nair , M.G. ( 2005 ) Human tumor cell growth inhibition by nontoxic anthocyanidins, the pigments in fruits and vegetables . Life Sci. 76 : 1465 – 1472 .

1185




1 running title: sorghum 3-deoxyanthocyanidin biosynthesis

Documents