9
Metabolic responses of Thellungiella halophila/salsuginea to biotic and abiotic stresses: Metabolite profiles and quantitative analyses M. Soledade C. Pedras * , Qing-An Zheng Department of Chemistry, University of Saskatchewan, 110 Science Place, Saskatoon, SK, Canada S7N 5C9 article info Article history: Received 20 October 2009 Received in revised form 9 December 2009 Available online 1 February 2010 Keywords: Albugo candida Brassicaceae Crucifer Glucosalsuginin Leptosphaeria maculans Phytoalexin Salt cress UV stress Thellungiella halophila T. salsuginea Thiane Wasalexins abstract The metabolite profiles of the model crucifer Thellungiella salsuginea (salt cress) ecotype Shandong sub- jected to various biotic and abiotic stresses were analyzed using HPLC-DAD-ESI-MS. Two different crucif- erous microbial pathogens, Albugo candida, a biotrophic oomycete, and Leptosphaeria maculans,a necrotrophic fungus, elicited formation of the phytoalexins wasalexins A and B without causing visual damage on inoculated leaves. Analyses of non-polar and polar metabolites led to elucidation of the chem- ical structures of five metabolites: 4 0 -O-(E)-sinapoyl-7-methoxyisovitexin-2 00 -O-b-D-glucopyranoside, 4 0 -O-(E)-sinapoylisovitexin-2 00 -O-b-D-glucopyranoside, 4-O-b-D-glucopyranosyl-7-hydroxymatairesinol, 5 0 -O-b-D-glucopyranosyldihydroneoascorbigen and 3-O-b-D-glucopyranosylthiane. 3-O-b-D-glucopyrano- sylthiane, an unique metabolite for which we suggest the name glucosalsuginin, is proposed to derive from the glucosinolate glucoberteroin. In addition, the identification of a broad range of polar metabolites identical to those of other crucifers was carried out. Quantification of several metabolites over a period of eight days showed that concentrations of the polar phytoanticipin 4-methoxyglucobrassicin increased substantially in leaves irradiated with UV light (k max 254 nm) relative to control leaves, but not in leaves subjected to other stresses. Ó 2009 Elsevier Ltd. All rights reserved. 1. Introduction Cruciferous extremophiles of the genus Thellungiella (Brassica- ceae family) appear to show resistance to stress caused by salinity, cold and draught. T. halophila (C.A. Meyer) O.E. Schultz, T. salsugi- nea (Pallas) O.E. Schulz and T. parvula (Schrenk) Al-Shehbaz and O’Kane are among the species of Thellungiella described so far. The Shandong (from China) and Yukon (from Canada) ecotypes have been cited as T. halophila (salt cress), although both ecotypes appear to belong to the species T. salsuginea (Amtmann, 2009; hence, these ecotypes will be hereon referred as T. salsuginea). The Shandong (Inan et al., 2004) and Yukon ecotypes are annual crucifers and important model systems due to their small genomes (about twice that of Arabidopsis thaliana) and high resistance to salinity. T. salsuginea (Shandong and Yukon ecotypes) abiotically stressed with CuCl 2 produced the phytoalexins wasalexins A (1) and B (2), 1-methoxybrassenin B (3) and rapalexin A (4)(Pedras and Adio, 2008). The Shandong ecotype exposed to UV-radiation (k max 254 nm) produced the largest quantities of wasalexins A (1) and B (2), together with the unique wasalexin photoaddition prod- ucts biswasalexins A1 (5) and A2 (6)(Pedras et al., 2009). By con- trast, irrigation of T. salsuginea Shandong with a NaCl solution induced substantially smaller amounts of phytoalexins. The pro- duction of biswasalexins A1 (5) and A2 (6) in leaves of UV stressed Shandong appeared to result from a photochemical reaction that might protect plants from UV-radiation (Fig. 1). The phytoantici- pins indolyl-3-acetonitrile (7), arvelexin (8), caulilexin C (9), neo- ascorbigen (10) and methylsulphanylpropylisothiocyanate (11) were also isolated from both elicited and control leaves. To date, polar metabolites from Thellungiella species subjected to any of those abiotic stresses or even in naturally healthy plants remain to be reported. To continue with the evaluation of stress responses of crucifer- ous species, we have analyzed the metabolite profiles of T. salsugi- nea Shandong inoculated with two different cruciferous microbial pathogens, Albugo candida (Pers. ex Chev.) Kuntze, a biotrophic oomycete, and Leptosphaeria maculans (Desm.) Ces. et de Not. (asex- ual stage Phoma lingam (Tode ex Fr.) Desm.), a necrotrophic fungus. Races of A. candida show specific pathogenicity to different Brassica species and cultivars (Rimmer et al., 2000). Similarly, the fungus L. maculans comprises various subgroups virulent to B. napus and B. rapa or B. juncea. In this study, (i) A. candida races 7 V (virulent to B. rapa cv. Reward, rapeseed) and 2 V (virulent to B. juncea cv. Cutlass, brown mustard) and (ii) L. maculans isolates virulent to 0031-9422/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.phytochem.2009.12.008 * Corresponding author. Tel.: +1 306 966 4772; fax: +1 306 966 4730. E-mail address: [email protected] (M.S.C. Pedras). Phytochemistry 71 (2010) 581–589 Contents lists available at ScienceDirect Phytochemistry journal homepage: www.elsevier.com/locate/phytochem

Metabolic responses of Thellungiella halophila/salsuginea to biotic and abiotic stresses: Metabolite profiles and quantitative analyses

Embed Size (px)

Citation preview

Page 1: Metabolic responses of Thellungiella halophila/salsuginea to biotic and abiotic stresses: Metabolite profiles and quantitative analyses

Phytochemistry 71 (2010) 581–589

Contents lists available at ScienceDirect

Phytochemistry

journal homepage: www.elsevier .com/locate /phytochem

Metabolic responses of Thellungiella halophila/salsuginea to biotic and abioticstresses: Metabolite profiles and quantitative analyses

M. Soledade C. Pedras *, Qing-An ZhengDepartment of Chemistry, University of Saskatchewan, 110 Science Place, Saskatoon, SK, Canada S7N 5C9

a r t i c l e i n f o

Article history:Received 20 October 2009Received in revised form 9 December 2009Available online 1 February 2010

Keywords:Albugo candidaBrassicaceaeCruciferGlucosalsugininLeptosphaeria maculansPhytoalexinSalt cressUV stressThellungiella halophilaT. salsugineaThianeWasalexins

0031-9422/$ - see front matter � 2009 Elsevier Ltd. Adoi:10.1016/j.phytochem.2009.12.008

* Corresponding author. Tel.: +1 306 966 4772; faxE-mail address: [email protected] (M.S.C. Pedras).

a b s t r a c t

The metabolite profiles of the model crucifer Thellungiella salsuginea (salt cress) ecotype Shandong sub-jected to various biotic and abiotic stresses were analyzed using HPLC-DAD-ESI-MS. Two different crucif-erous microbial pathogens, Albugo candida, a biotrophic oomycete, and Leptosphaeria maculans, anecrotrophic fungus, elicited formation of the phytoalexins wasalexins A and B without causing visualdamage on inoculated leaves. Analyses of non-polar and polar metabolites led to elucidation of the chem-ical structures of five metabolites: 40-O-(E)-sinapoyl-7-methoxyisovitexin-200-O-b-D-glucopyranoside,40-O-(E)-sinapoylisovitexin-200-O-b-D-glucopyranoside, 4-O-b-D-glucopyranosyl-7-hydroxymatairesinol,50-O-b-D-glucopyranosyldihydroneoascorbigen and 3-O-b-D-glucopyranosylthiane. 3-O-b-D-glucopyrano-sylthiane, an unique metabolite for which we suggest the name glucosalsuginin, is proposed to derivefrom the glucosinolate glucoberteroin. In addition, the identification of a broad range of polar metabolitesidentical to those of other crucifers was carried out. Quantification of several metabolites over a period ofeight days showed that concentrations of the polar phytoanticipin 4-methoxyglucobrassicin increasedsubstantially in leaves irradiated with UV light (kmax 254 nm) relative to control leaves, but not in leavessubjected to other stresses.

� 2009 Elsevier Ltd. All rights reserved.

1. Introduction

Cruciferous extremophiles of the genus Thellungiella (Brassica-ceae family) appear to show resistance to stress caused by salinity,cold and draught. T. halophila (C.A. Meyer) O.E. Schultz, T. salsugi-nea (Pallas) O.E. Schulz and T. parvula (Schrenk) Al-Shehbaz andO’Kane are among the species of Thellungiella described so far.The Shandong (from China) and Yukon (from Canada) ecotypeshave been cited as T. halophila (salt cress), although both ecotypesappear to belong to the species T. salsuginea (Amtmann, 2009;hence, these ecotypes will be hereon referred as T. salsuginea).The Shandong (Inan et al., 2004) and Yukon ecotypes are annualcrucifers and important model systems due to their small genomes(about twice that of Arabidopsis thaliana) and high resistance tosalinity. T. salsuginea (Shandong and Yukon ecotypes) abioticallystressed with CuCl2 produced the phytoalexins wasalexins A (1)and B (2), 1-methoxybrassenin B (3) and rapalexin A (4) (Pedrasand Adio, 2008). The Shandong ecotype exposed to UV-radiation(kmax 254 nm) produced the largest quantities of wasalexins A (1)and B (2), together with the unique wasalexin photoaddition prod-

ll rights reserved.

: +1 306 966 4730.

ucts biswasalexins A1 (5) and A2 (6) (Pedras et al., 2009). By con-trast, irrigation of T. salsuginea Shandong with a NaCl solutioninduced substantially smaller amounts of phytoalexins. The pro-duction of biswasalexins A1 (5) and A2 (6) in leaves of UV stressedShandong appeared to result from a photochemical reaction thatmight protect plants from UV-radiation (Fig. 1). The phytoantici-pins indolyl-3-acetonitrile (7), arvelexin (8), caulilexin C (9), neo-ascorbigen (10) and methylsulphanylpropylisothiocyanate (11)were also isolated from both elicited and control leaves. To date,polar metabolites from Thellungiella species subjected to any ofthose abiotic stresses or even in naturally healthy plants remainto be reported.

To continue with the evaluation of stress responses of crucifer-ous species, we have analyzed the metabolite profiles of T. salsugi-nea Shandong inoculated with two different cruciferous microbialpathogens, Albugo candida (Pers. ex Chev.) Kuntze, a biotrophicoomycete, and Leptosphaeria maculans (Desm.) Ces. et de Not. (asex-ual stage Phoma lingam (Tode ex Fr.) Desm.), a necrotrophic fungus.Races of A. candida show specific pathogenicity to different Brassicaspecies and cultivars (Rimmer et al., 2000). Similarly, the fungus L.maculans comprises various subgroups virulent to B. napus andB. rapa or B. juncea. In this study, (i) A. candida races 7 V (virulentto B. rapa cv. Reward, rapeseed) and 2 V (virulent to B. juncea cv.Cutlass, brown mustard) and (ii) L. maculans isolates virulent to

Page 2: Metabolic responses of Thellungiella halophila/salsuginea to biotic and abiotic stresses: Metabolite profiles and quantitative analyses

N

N

O

OCH3

H3CSSCH3

NO

OCH3

NH3CS

H3CS

N

N

OCH3

H3CSSCH3O

NH

OCH3 NCS

1 2 3 4

NOCH3

OO

OHO

OH

OHN

R1

R2

CN

S NCS

7 R1=R2=H8 R1=OCH3; R2=H9 R1=H; R2=OCH3

11

10

N

N N

S

S

N

S

S

OO

OO

5

N

NN

NO

O

O

O

SS

S

S

6

Fig. 1. Phytoalexins 1–6 and phytoanticipins 7–11 produced in leaves of Thellungiella salsuginea Shandong ecotype irradiated with UV light (Pedras et al., 2009).

582 M.S.C. Pedras, Q.-A. Zheng / Phytochemistry 71 (2010) 581–589

B. napus (cv. Westar, canola) and B. juncea (cv. Cutlass) were used toinoculate T. salsuginea Shandong. To the best of our knowledge, noevaluation of potential interactions between T. salsuginea and anyof these cruciferous pathogens has been reported to date. In addi-tion, the profiles of polar metabolites of T. salsuginea Shandong sub-jected to stress caused by UV-radiation, NaCl irrigation and CuCl2

spray were analyzed.In this work, while no visual damage on leaves of T. salsuginea

Shandong was caused by any of the pathogens, formation of thephytoalexins wasalexins A (1) and B (2) were induced in everyplant–microbe interaction. In addition, elucidation of the chemicalstructures of five new glucosylated metabolites and identificationof a very broad range of polar metabolites identical to those ofother cruciferous species was achieved. Among the new metabo-lites, 3-O-b-D-glucopyranosylthiane (21) is a particularly interest-ing structure, as no naturally occurring thianes appear to havebeen reported to date.

Fig. 2. Production of wasalexins A (1) and B (2) in leaves of Thellungiella salsugineaShandong ecotype incubated with microbial pathogens Albugo candida races 2 Vand 7 V, Leptosphaeria maculans isolates BJ-125 and Laird-2 up to eight days post-elicitation (dpe).

2. Results and discussion

2.1. Microbial elicitation and analysis of non-polar metabolites

Leaves of T. salsuginea were inoculated with A. candida races 2 Vand 7 V and L. maculans isolates BJ-125 and Laird-2, as described inthe Experimental. Plants were harvested 2, 4, 6 and 8 days afterelicitation, the aerial parts were frozen in liquid nitrogen, ground,and extracted with MeOH. Control plants were extracted similarly.After concentration to dryness, the residues were rinsed withCH2Cl2 to yield non-polar fractions (soluble in CH2Cl2), and polarfractions (remaining residues soluble in MeOH–H2O). Compoundsdetected in each extract were identified either by direct compari-son with authentic samples available in our metabolite libraries(HPLC-DAD and HPLC-ESI-MS), or by isolation and structure deter-mination using spectroscopic data previously reported (Pedraset al., 2006, 2007, 2008b). Four known phytoalexins, wasalexin A

(1), wasalexin B (2), 1-methoxybrassenin B (3), and rapalexin A(4), and two phytoanticipins, caulilexin C (9, 1-methoxyindolyl-3-acetonitrile) and 3-methylsulfanylpropylisothiocyanate (11)were detected in extracts of leaves inoculated with the four

Page 3: Metabolic responses of Thellungiella halophila/salsuginea to biotic and abiotic stresses: Metabolite profiles and quantitative analyses

M.S.C. Pedras, Q.-A. Zheng / Phytochemistry 71 (2010) 581–589 583

pathogens. Among the phytoalexins, only wasalexins 1 and 2 wereproduced in quantities sufficient for quantification (Fig. 1).

Phytoalexin quantification by HPLC-DAD (using calibrationcurves built with synthetic compounds) showed that A. candidarace 7 V elicited accumulation of the largest quantities of wasalex-ins A (1) and B (2) two days after elicitation (ca. 50 and 30 nmoles/gfresh weight, respectively, Fig. 2). Leaves inoculated with L. macu-lans isolates BJ-125 and Laird-2 produced similar amounts of was-alexins A (1) and B (2) (ca. 15 and 30 nmoles/g fresh weight,Fig. 2), whereas inoculation with A. candida race 2 V elicited gener-ation of the lowest amounts (Fig. 2). The amounts of wasalexins A(1) and B (2) elicited by these microbes were substantially lowerthan those previously determined to be elicited by UV-radiation(kmax 254 nm); however these quantities were comparable to those

HOHO

HO

O

OH

R2O

O

OR1

OHO

HOHO

OH

OO

HOHO O

OO

HO

HOHO

RO

O

OCH3

OCH3

OCH3O

OH

OCH3

OCH3

H3

H

HOH

17 R = H18 R = sinapoyl

OOHO

HOHO HO

S

N

S

NOSO3

-

R1

R2

26 R1=R2=H27 R1=OCH3; R2=H28 R1=H; R2=OCH3

OOH

OHOH

HO S

NOSO3

-

HOS

S

NOSO3

-

O

OHO

29

31

R

NH2

COOH

O

OCH3

R

33 R = H34 R = OCH3

35 R = H36 R = OH

OHO

HOHO

HO

21 OH

O

O

OH

OH22

HOHO

HO

12 R1=R2=H13 R1=CH3; R2=H14 R1=H; R2=Glc

OH

OHHO2C3

Fig. 3. Polar metabolites isolated from leaves of Thellungiella salsuginea Shangd

elicited by NaCl irrigation (Supplementary Data, Fig. S1). In addi-tion, the quantities of caulilexin C (9) in leaves of T. salsuginea in-fected with any of the four pathogens were similar to those ofcontrol leaves but in general lower than in leaves sprayed withCuCl2 or UV irradiated (kmax 254 nm). By contrast, the quantitiesof neoascorbigen (10) two days after inoculation of leaves with L.maculans were substantially higher than in leaves subjected toother stresses (Supplementary Data, Fig. S2).

2.2. Analysis and structure elucidation of polar metabolites

The polar fractions accumulated in leaves of T. salsuginea Shan-dong elicited with pathogens (A. candida races 2 V and 7 V, L. mac-ulans isolates BJ-125 and Laird-2), CuCl2 spray, UV-radiation, and

O

OH

RO

O

O

O

O

15 R=CH3

16 R=H

O

O

HO

CO

OOCH3

HO

OO

OHO 19

N

OCH3

OHO

OHHO

HOOHO

HOHO

HOO

20

OR1

OCH3

R2OO

OCH3

HO

OOH

OHOHHO

H3CO

O

25

OOH

OHOH

S

NOSO3

-

OOH

OHOH

HOS

OH

OHH

30

32

S

NOSO3

-

OOH

OHOH

HO

N

N

N

N

NH2

O

OHHO

HO

38

OH

NH

COOH

NH2

37

O

OH

OH

OCH3

OCH3

O

23 R1=Glc; R2=H24 R1=H; R2=Glc

O

2

35

81'

4'

1''

1'''

ong. Metabolites 15, 16, 19, 20 and 21 are reported here for the first time.

Page 4: Metabolic responses of Thellungiella halophila/salsuginea to biotic and abiotic stresses: Metabolite profiles and quantitative analyses

Fig. 4. Production of 4-methoxyglucobrassicin (27) and neoglucobrassicin (28) inleaves of Thellungiella salsuginea Shangdong under different stress conditions and incontrol leaves up to eight days post-elicitation (dpe).

584 M.S.C. Pedras, Q.-A. Zheng / Phytochemistry 71 (2010) 581–589

NaCl irrigation were analyzed by HPLC-DAD and HPLC-ESI-MS.Although the chromatograms were rather complex (SupplementaryData, Fig. S3), many metabolites could be identified by comparisonof their retention times, UV and mass spectral data with those ofcompounds available in our libraries (Pedras et al., 2008a,b).Twenty-seven metabolites (12–38) were identified in the polar ex-tracts of leaves of Shangdong elicited with UV-radiation (kmax

254 nm). Major metabolites included flavonoids 12–16, glucosyl es-ters 17 and 18, phenylpropanoids 19, 22, 33, 34, phenolic glucosides

Table 11H NMR chemical shifts (ppm), multiplicity (s, d, t, brs), coupling constants (J, Hz) for flav

15 (CD3OD)

A B

3 6.05, s 6.43, s8 6.48, s 6.49, s20 , 60 7.47, d, 8.0 7.55, d, 8.030 , 50 6.72, d, 8.0 6.66, d, 8.05-OH – ––OMe 3.85, s 3.81, s

Glc-100 4.86, d, 7.5 4.82, d, 7.5200 4.32, t, 8.5 4.57, t, 8.5300

400

500

600

Glc-1000 4.39, d, 10.0 4.44, d, 10.02000

3000

4000

5000

6000

Sinapoyl2,6 6.33, s 6.51, sa 5.86, d, 16.0 5.94, d, 16.0b 6.95, d, 16.0 7.03, d, 16.03,5-OMe 3.75, s 3.71, s

23, 24, 25, glucosinolates 26–32, together with a few primarymetabolites (35–38). Five metabolites not available in our librarieswere further isolated and their structures were determined byspectroscopic analyses (described below in Section 2.2.1) andestablished to be unknown compounds 15, 16, 19, 20, 21 (Fig. 3).

Glucosinolates were the major group of metabolites identifiedin leaves of Shandong. The aliphatic glucosinolates glucoibervirin(29), glucoiberin (30), and sinigrin (31) were present in all sam-ples. Their identification was carried out by HPLC-DAD-ESI-MS,which showed typical negative ions [M�1]� at 406, 422, and 358respectively. Partial purification followed by 1H NMR spectro-scopic analyses confirmed the presence of aliphatic side-chains(–CH2)3–) and thiomethyl groups in 29 and 30, and an allyl group(=CH) at d 5.9 ppm in 31 (spectral data in Supporting Data). Glu-coibervirin (29) was present in very high amount in both elicitedand control leaves. Indolyl glucosinolates 26–28 and glu-conasturtiin (32) were present in control leaves as well. Theiridentification using HPLC-DAD-ESI-MS was confirmed after purifi-cation and 1H NMR spectroscopic analyses (Pedras et al., 2008a,b).Quantitative analyses of 27 and 28 were carried out by HPLC-DAD(Fig. 4) using calibration curves built with pure compounds. AfterUV elicitation (kmax 254 nm) of leaves, the amounts of glucosino-late 27 increased 3- to 6-fold relative to those of control leaves,with the highest amount detected in leaves collected six days afterUV elicitation (1200 nmol/g fresh wt.). Interestingly, no other elic-itation methods affected substantially the amounts of 27 relativeto those of control samples. Glucosinolate 28 was detected inleaves inoculated with L. maculans in lower concentrations (ca.6-fold) than those of leaves inoculated with A. candida. Other stres-ses did not affect substantially the amounts of 27 or 28 relative tothose of control samples. Recently, seven glucosinolates (10-meth-ylsulfinyldecylglucosinolate, glucobrassicin (26), 1-methoxygluco-brassicin (28), glucoibervirin (29), glucoiberin (30), sinigrin (31),and gluconasturtiin (32)) were identified in various tissues ofShandong throughout its life cycle (Pang et al., 2009). Althoughthe glucosinolate profiles varied significantly among different tis-sues, indolyl glucosinolates 26 and 28 were produced in muchsmaller amounts than aliphatic, and roots contained the highestconcentrations of total, aromatic and indole glucosinolates.

onoids 15 and 16.

15 (DMSO-d6) 16 (D2O)

A B A

6.05, s 6.43, s 5.66, s6.48, s 6.49, s 6.06, s7.82, d, 8.0 7.83, d, 8.0 6.92, d, 8.06.90, d, 8.0 6.86, d, 8.0 6.39, d, 8.013.65, brs 13.53, brs –3.90, s 3.87, s

4.70, d, 7.5 4.69, d, 7.5 4.47, d, 7.54.25, t, 8.5 4.54, t, 8.5 4.40, t, 8.5

4.29, d, 10.0 4.29, d, 10.0 4.58, d, 10.0

6.33, s 6.51, s 5.90, s5.86, d, 16.0 5.94, d, 16.0 5.43, d, 16.06.95, d, 16.0 7.03, d, 16.0 6.62, d, 16.03.75, s 3.71, s 3.45, s

Page 5: Metabolic responses of Thellungiella halophila/salsuginea to biotic and abiotic stresses: Metabolite profiles and quantitative analyses

Table 213C NMR chemical shifts (ppm) for new flavonoids 15 and 16.

15 (CD3OD) 15 (DMSO-d6) 16 (D2O)

2 164.6 164.4 164.93 103.4 103.7 102.54 182.3 182.7 182.45 159.9 161.7 157.26 109.1 109.1 107.97 165.7 164.1 164.58 91.1 90.4 93.99 157.7 157.3 157.310 104.4 105.7 104.510 121.7 121.6 121.720 , 60 129.1 128.9 128.130 , 50 116.3 116.2 115.340 161.7 161.6 161.1–OMe 56.6 56.8 –Glc-100 71.1 71.4 71.9200 76.7 80.8 79.1300 79.3 79.1 79.1400 70.7 70.8 70.0500 82.5 82.2 81.2600 60.7 61.9 62.2Glc-1000 106.2 104.9 103.62000 72.7 72.7 73.43000 73.5 73.8 74.14000 68.9 69.3 68.95000 74.8 74.8 76.76000 62.5 62.7 61.5

Sinapoyl1 124.6 124.8 124.82,6 106.2 106.5 104.73,5 148.3 148.3 147.74 138.6 138.7 137.9a 114.4 114.9 113.6b 145.5 145.4 145.4C@O 166.8 166.7 167.53, 5-OMe 56.4 56.5 55.3

M.S.C. Pedras, Q.-A. Zheng / Phytochemistry 71 (2010) 581–589 585

Flavonoids were the second major type of metabolites isolatedfrom leaves of Shandong: isovitexin (12), 4-methoxyisovitexin(13), isosaponarin (14), 40-O-(E)-sinapoyl-7-methoxyisovitexin-200-O-b-D-glucopyranoside (15) and 40-O-(E)-sinapoyl-isovitexin-200-O-b-D-glucopyranoside (16). Among them, 15 and 16 do notappear to have been reported to date. Their structures were iden-tified by spectroscopic methods, especially 2D NMR spectra, as de-scribed below. The chemical structures of metabolites 12, 13, and14 first assigned using HPLC-DAD-ESI-MS data were confirmedafter isolation, analyses of 1H NMR spectroscopic data and directcomparison with authentic samples available in our libraries. Iso-vitexin (12) was reported from Thlaspi arvense and Wasabia japon-ica (Pedras et al., 2003; Takahiro et al., 2005). Methylatedisovitexin (13) was reported from several other plant species asfor example, Piper umbellatum (Tabopda et al., 2008) and Aquilegiavulgaris (Bylka et al., 2002), but is reported here for the first timefrom a cruciferous species. Isosaponarin (14) was reported fromSpirodela oligorrhiza and Wasabia japonica (Jurd et al., 1957; Taka-hiro et al., 2005).

Phenylpropanoids were another significant group of metabo-lites present in leaves of T. salsuginea; a compound yet unknown(19), chlorogenic acid (22), coniferin (33), and syringin (34) wereisolated and characterized. Metabolite 19 was identified as de-scribed below in Section 2.2.1. The chemical structures of metabo-lites 33 and 34 were determined by direct comparison of theirNMR spectroscopic data with those of authentic samples. The 1HNMR of 22 showed the characteristic signals of a caffeoyl moiety,esterified with a polyol-acid. This polyol-acid moiety was deter-mined to be quinic acid from analyses of the 1H coupling constants(3JHH), which suggested that the structure of 22 was chlorogenicacid, a well-known metabolite reported from several plant species(Cheminat et al., 1988; Tanaka et al., 1993).

A few primary metabolites, including phenylalanine (35), tyro-sine (36), tryptophan (37) and adenosine (38) were readily de-tected and identified by HPLC-DAD analysis. Phenylalanine (35)and tyrosine (36) are precursors of aromatic glucosinolates whiletryptophan (37) is a precursor of several indole derivatives,including indolyl glucosinolates, and phytoalexins of Brassicaceae.Quantitative analysis of tryptophan (37) indicated no significantchanges in leaves inoculated with pathogens relative to those ofcontrol leaves. Interestingly, the amounts of tryptophan in sam-ples elicited with NaCl were 2- to 8-fold higher than in controlsamples.

2.2.1. Isolation, characterization and structure determination of newmetabolites 15, 16, 19, 20 and 21

Compound 15 (6 mg) was obtained as a yellow amorphouspowder whose molecular formula was determined to beC39H42O19 by HR-ESI-MS. Both the 1H- and 13C NMR spectra werecomplex, showing two sets of peaks for each signal (Table 1). InCD3OD, the peak ratio was about 1:1, while in DMSO-d6, the ratiowas about 2:1. The 1H NMR in CD3OD showed two sets of signalsfor each of: eight aromatic protons (d 6.05/6.43, 6.48/6.49,6.72 � 2/6.66 � 2, 7.47 � 2/7.55 � 2, 6.33 � 2/6.51 � 2), two ole-finic protons (d 6.95/7.03, 5.86/5.94; each 1 H, J = 16.0 Hz respec-tively), two anomeric protons (d 4.39/4.44, 4.86/4.82; each 1 H,J = 10.0 and 7.5 Hz respectively), carbohydrate protons (d 2.95–4.59), and three methoxyl groups (d (3.75/3.71) � 2, 3.85/3.81).The 1H NMR in DMSO showed one additional D2O exchangeableproton (d 13.53/13.65) (Table 1). The 13C NMR in CD3OD showedsignals of a conjugated carbonyl carbon (d 182.4/182.5), an estercarbonyl carbon (d 167.5/167.2), 19 aromatic or olefinic carbons,12 carbons attached to oxygen, and three methoxyl carbons (d(55.4/55.3) � 2, 55.4/55.5) (Table 2). 2D NMR spectroscopic analy-sis indicated the presence of a C-6-substituted apigenin-type fla-vone bound to a sinapoyl moiety and to two glucosyl residues, as

follows. For example, HMBC data revealed a correlation betweenan anomeric proton (d 4.39/4.44) with aromatic carbons (d 159.8,108.9, 165.7/160.8, 108.8, 164.2) indicating a bond between aflavone carbon and an anomeric carbon, which was consistent withthe anomeric carbon chemical shift (d 71.7/71.2). Another anomer-ic proton (d 4.86/4.82) showed a correlation with a carbohydratecarbon (d 80.4/78.9), which was proven to be C-2 of the C-glucosylgroup from analysis of HMQC and COSY spectroscopic data. Theb-configurations for both glucosyl residues were determined fromthe 3JH1,H2 coupling constants (H100–H200 = 10.0 Hz and H1000–H2000 = 7.5 Hz). The 7-methoxyl group (d 3.85/3.81) was deducedfrom HMBC data showing its correlation with an aromatic carbon(d 165.7/164.2). The additional exchangeable proton (d 13.53/13.65) observed in 1H NMR in DMSO-d6 was due to the 5-OH group(hydrogen bonding with 4-carbonyl oxygen). Accordingly, thestructure of 15 was assigned to be 40-O-(E)-sinapoyl-7-methoxy-isovitexin-200-O-b-D-glucopyranoside. The double sets of peaks foreach NMR signal are two rotamers likely due to the hindered rota-tion about the C-6–C-100 bond caused by the flavone substituents atC-5 and C-7 (Fig. 5). Similar observations were previously reportedin other glucosylated apigenin-type flavones (Takahiro et al.,2005).

Compound 16 (16 mg) was also obtained as a yellow amor-phous powder with a molecular formula of C38H40O19 (HR-ESI-MS), suggesting a CH2 less than 15. Similar to compound 15, the1H- and 13C NMR spectra were complex, showing all characteristicsignals of a C-6-substituted apigenin-type flavone (absence ofMe(O) group), a sinapoyl group, and two glucosyl residues (Tables1 and 2). Hence, after analysis of the spectroscopic data, as de-scribed above for 15, 16 was identified as 40-O-(E)-sinapoylisovit-exin-200-O-b-D-glucopyranoside.

Page 6: Metabolic responses of Thellungiella halophila/salsuginea to biotic and abiotic stresses: Metabolite profiles and quantitative analyses

15

O

OH

H3CO

O

O

OHO

HOHO

OOHO

HOHO HO

O

OCH3

OH

OCH3

H

H

H

H

2

345

6

78

9

10

1'

4'

1''

2'

3'

5'6'

1'''2'''

2''

4''6''

6'''4'''

1

23

4

56

α

β

sinapoyl

Fig. 5. Selected HMBC correlations of new metabolite 15.

O

O

HO

H3CO

OOCH3

HO

OHO

HOHO

HO19

H H

H

4'3'

3

1''

1

7'

Fig. 6. Selected HMBC correlations of new metabolite 19.

586 M.S.C. Pedras, Q.-A. Zheng / Phytochemistry 71 (2010) 581–589

Compound 19 (4 mg) was obtained as an amorphous powderwith a likely molecular formula of C26H32O12 (HR-ESI-MS). Its 1HNMR spectrum showed signals for: two ABX spin systems on thearomatic region, indicating two phenyl rings with C-1, C-3, andC-4 substitutions, a carbohydrate moiety, seven protons attachedto sp3 carbons and two methoxy groups. The 13C NMR showed26 carbon signals, including one carbonyl at d 183.3 ppm, 12 sp2

carbons, two methoxy, and 11 sp3 carbons. HMBC spectroscopicdata showed that the carbohydrate moiety was attached to C-4of a phenyl group and two methoxy groups were attached to C-3and C-30 of each phenyl ring. The remaining part of the structure,including the glucosyl residue was determined from 1H–1H COSYand HMBC spectroscopic analyses (Fig. 6). Hence, based on thesecorrelations the aglycone portion, including the absolute configu-ration, is proposed to be 70-hydroxymatairesinol, earlier isolatedfrom pine (Freudenberg and Knof, 1957) and later on shown tobe a precursor of 5-methoxypodophyllotoxin (Xia et al., 2000).Lignans of the a,b-dibenzyl-c-butyrolactone series, carrying a hy-

OOHO

HOHO

HOS

21

S

N-O3SO

OOH

OHOH

HO

S D-Glc

D-Glc

39

CH3S–(CH2)5–

Fig. 7. Proposed metabolic pathway for the formation of metabolite 21 (gluco

droxyl group a to the carbonyl group, are widely distributed inplants (Ward, 1997). However, to the best of our knowledge its40-glucoside has not been reported to date. Based on the aboveinformation, compound 19 was identified as 40-O-b-D-glucopyran-osyl-70-hydroxymatairesinol.

Two additional new metabolites were isolated from polar frac-tions and their structures assigned based on spectroscopic analy-ses. A new indolyl derivative (20) (4 mg) of molecular formulaC22H29NO12 displaying 1H NMR signals typical of a 1,3-substitutedindole, a carbohydrate moiety and signals attributable to the dihy-droneoascorbigen moiety, previously reported (Pedras et al., 2008).HMBC spectroscopic data showed a correlation between theanomeric proton of glucose and C-50 of dihydroneoascorbigen.Therefore, compound 20 was identified as 50-O-b-D-glucopyra-nosyldihydroneoascorbigen.

Compound 21 (77 mg) was obtained as a white amorphous pow-der of molecular formula C11H20O6S. The 1H NMR spectrum showedsignals of a glucopyranosyl moiety (C6H11O6) and eight additionalprotons. HMQC spectroscopic data showed, in addition to the gluco-syl residue, four methylene carbons and one methyne attached tooxygen. Considering that the fragment C5H9S remained to be as-signed, only a 6-membered ring would fit the data, hence a thiane(tetrahydrothiopyran) was assigned but the absolute configurationat C-3 remains unknown at this time. Finally, metabolite 21 was as-signed as 3-O-b-D-glucopyranosylthiane. The biosynthetic origin ofthis metabolite is quite intriguing, as no similar natural product hasbeen reported to date. 3-O-b-D-glucopyranosylthiane (21) could re-sult from enzymatic degradation of a more complex structure suchas glucosinolate 39, which upon elimination of sulfate and cycliza-tion could yield ion pair 41. Next, formation of methylthiocyanateand thiane (42) followed by oxidation of 42–43 and glucosydationcould yield metabolite 21 (Fig. 7). This hypothesis is based on pre-vious work by Benn and co-workers, suggesting that glucosinolateswith R = CH3S–(CH2)n in which n = 5–8 (e.g. 39, n = 5), could yield

SH

NOSO3

-

S SO42Š

SSCN

CH3SCN

SS

HO

4041

4243

[O]

salsuginin) from glucosinolate 39 (structures in brackets are postulated).

Page 7: Metabolic responses of Thellungiella halophila/salsuginea to biotic and abiotic stresses: Metabolite profiles and quantitative analyses

M.S.C. Pedras, Q.-A. Zheng / Phytochemistry 71 (2010) 581–589 587

cyclic sulfides via cyclic S-methyl sulfonium salts (Benn and Singh,1986). In fact, glucosinolate 39 (glucoberteroin) is a natural productfirst reported from Berteroa incana L. (Kjaer et al., 1955), and morerecently from Dithyrea wislizenii Engelm. (Brassicaceae) (Montautet al., 2009). Glucoberteroin (39) was not detected in this work, per-haps due to lower concentrations as it might be converted to 21immediately after its formation, via the metabolic pathway pro-posed in Fig. 7. Hence, previous work on the reactions of NCS�withS-methyl-(1,n)-epithionium ions (Benn and Singh, 1986) suggestedthat thiane 42 could be a natural product, although it has not beenreported to date. Taking these considerations together with ourdata, here we suggest that the enzymatic transformation of 42 tothe glucoside 21 is a metabolic process occurring in T. salsuginea;however, additional work needs to be carried out before thesehypotheses are confirmed and the ecological role of 21 is under-stood. Here we propose the trivial name glucosalsuginin for 3-O-b-D-glucopyranosylthiane (21).

3. Conclusion

Analyses of the metabolic responses of T. salsuginea ecotypeShandong using HPLC-DAD-ESI-MS showed that phytoalexins andphytoanticipins are produced in response to biotic and abioticstresses. Biotic elicitation using different cruciferous pathogensdid not cause visual leaf damage but induced phytoalexin produc-tion, suggesting that T. salsuginea is either resistant or tolerant ofthese microbes. Interestingly, phytoalexin production in responseto UV elicitation was significantly different from responses to otherstresses, that is, substantially larger quantities of wasalexins A (1)and B (2) were produced in plants exposed to UV-radiation (kmax

254 nm). Considering the significance of the cruciferous phyto-alexin pathway, and that to date no genes/enzymes involved inbrassinin (and derivatives or similar metabolites) biosynthesishave been cloned, T. salsuginea appears to be a good model plantfor such studies. Furthermore, the profiles of polar metabolites ofleaves showed that T. salsuginea produced a number of metabolitesidentical or similar to those of other crucifers, except for the un-ique metabolite 21, for which the name glucosalsuginin is pro-posed. Future work will have to address new questions arisingfrom this discovery.

4. Experimental

4.1. Chemicals and instrumentation

All chemicals were purchased from Sigma–Aldrich Canada Ltd.,Oakville, ON; solvents were HPLC grade and used as such. Organicextracts were dried with Na2SO4 and solvents removed under re-duced pressure in a rotary evaporator. Flash column chromatogra-phy (FCC) was carried out using silica gel grade 60, mesh size 230–400 Å. Preparative thin layer chromatography (prep TLC) was car-ried out on silica gel plates, Kieselgel 60 F254 and compounds werevisualized under UV light.

Nuclear magnetic resonance (NMR) spectra (1H, 13C, HMQC –heteronuclear multiple quantum coherence, HMBC – heteronu-clear multiple bond coherence) were recorded on Bruker Avance500 MHz spectrometers. High resolution (HR) electron impact(EI) mass spectra (MS) were obtained on a VG 70 SE mass spec-trometer, employing a solids probe.

HPLC analysis was carried out with Agilent high performanceliquid chromatographs equipped with quaternary pump, auto-matic injector, and diode array detector (DAD, wavelength range190–600 nm), degasser, and an Eclipse XDB C-18 column (5 lmparticle size silica, 4.6 i.d. � 200 mm), having an in-line filter.Method A: (for non-polar metabolites): 75:25 H2O–CH3CN to

100% CH3CN, for 35 min plus 10 min at 100% CH3CN, linear gradi-ent, and at a flow rate 1.0 ml/min; method B (for polar metabo-lites, both H2O and CH3CN containing 0.1% TFA): 100% H2O to89:11 H2O–CH3CN linear gradient for 15 min, 89:11 H2O–CH3CNisocratic for 10 min, to 70:30 H2O–CH3CN linear gradient for20 min, to 50:50 H2O–CH3CN linear gradient for 5 min, at a flowrate of 0.9 ml/min.

HPLC-DAD-ESI-MS analysis was carried out with an Agilent1100 series HPLC system equipped with an autosampler, binarypump, degasser, and a diode array detector connected directly toa mass detector (Agilent G2440A MSD-Trap-XCT ion trap massspectrometer) with an electrospray ionization (ESI) source. Chro-matographic separation was carried out at room temperature usingan Eclipse XDB C-18 column (5 lm particle size silica,150 � 4.6 mm I.D.). The mobile phase consisted of a linear gradientof 0.2% HCO2H acid in H2O and 0.2% HCO2H acid in CH3CN: 75:25 to25:75 H2O–CH3CN in 35 min, to 0:100 H2O–CH3CN in 5 min andthen 100 for 5 min, and a flow rate of 1.0 ml/min. Data acquisitionwas carried out in positive and negative polarity modes in a singleLC run. Data processing was carried out by Agilent ChemstationSoftware. HPLC-HR-ESI-MS was performed on an Agilent HPLC1100 series directly connected to a QSTAR XL Systems Mass Spec-trometer (Hybrid Quadrupole-TOF LC/MS/MS) with turbo spray ESIsource. Chromatographic separation was carried out at room tem-perature using a Hypersil ODS C-18 column (5 lm particle size sil-ica, 200 � 2.1 mm I.D.) or a Hypersil ODS C-18 column (5 lmparticle size silica, 100 � 2.1 mm I.D.). The mobile phase consistedof a linear gradient of 0.1% HCO2H acid in H2O and 0.1% HCO2H acidin CH3CN (75:25 to 25:75 in 35 min, to 0:100 in 5 min) and a flowrate of 0.25 ml/min. Data processing was carried out by Analyst QSSoftware.

4.2. Plant material and elicitation

Thellungiella salsuginea ecotype Shandong was obtained fromTAIR (envelope labeled as Thellungiella halophila). Seeds were sownin a perlite and nutrient free LG-3 soil (Sun Gro Horticulture Can-ada) in a Petri dish and stratified at 4 �C in the dark for 10 days.After then the seeds were kept in a growth chamber at 16 h oflight/8 h of dark, 25 �C day/16 �C night, light intensity of150 lEm�2 s�1, and with ambient humidity. After seven days theseedlings were transferred into soil in small pots (50 pots/tray),and kept under the same conditions. After four weeks of growth,potted plants were elicited as described below.

4.2.1. Abiotic elicitation4.2.1.1. UV elicitation. Four weeks old plants were irradiated withtwo lamps of 13 W (UV kmax 368 nm, 315–400 nm), or one lampof 30 W (UV kmax 254 nm), in either case for 60 min. After irradia-tion, plants were kept in the growth chamber under conditions de-scribed above and plants were harvested 2, 4, 6, and 8 days afterirradiation.

4.2.1.2. CuCl2. elicitation. Four weeks old plants were sprayed with aCuCl2 solution (0.5 mM) using a hand sprayer. After irradiation, theplants were kept in the growth chamber under conditions de-scribed above and plants were harvested 2, 4, 6, and 8 days afterspraying.

4.2.1.3. NaCl elicitation. Four weeks old plants in pots not wateredfor two days (soil allowed to dry) were placed in trays containinga NaCl solution (0.5 M) until plants were harvested (after 2, 4, 6,and 8 days) and kept in the growth chamber under conditions de-scribed above. Control plants were treated similarly but placed intrays containing only water.

Page 8: Metabolic responses of Thellungiella halophila/salsuginea to biotic and abiotic stresses: Metabolite profiles and quantitative analyses

588 M.S.C. Pedras, Q.-A. Zheng / Phytochemistry 71 (2010) 581–589

4.2.2. Biotic elicitation4.2.2.1. Albugo candida races 2V and 7V. Sporangia of A. candidawere multiplied as described previously (Pedras et al., 2008). Forpreparation of a sporangia suspension, 10 g of frozen stored spo-rangia was added into distilled H2O (50 ml). The suspension waskept at 16 �C in the dark for 3 h before inoculation. For inoculation,the spore solution was sprayed evenly on leaves of four-week-oldplants using a sprayer. The trays were covered with a plastic dometo keep the moisture in, and kept in a growth chamber at 16 �C un-der dark with 95% humidity. After 24 h, the conditions were chan-ged back to normal conditions as described above.

4.2.2.2. Leptosphaeria maculans isolates BJ-125 and Laird-2. Sporesof L. maculans were added into distilled H2O (50 ml) and the solu-tion was adjusted to a concentration 106 spores/ml. For inocula-tion, solution was sprayed evenly on leaves of four-week-oldplants using a sprayer. Then the trays were covered with a plasticdome to keep the moisture in, and kept in a growth chamber undernormal conditions, as described above.

4.3. Extractions and time-course analysis of plant metabolites

Four to six elicited plants were combined to make one analyti-cal sample and triplicate samples were prepared for analysis. Theplant tissues (aerial parts) were frozen in liquid N2, ground, and ex-tracted with MeOH (10 ml � 2) with sonication, each time 20 min.Control plants were collected and processed similar to elicitedplants. After filtration, the combined MeOH extracts were concen-trated under reduced pressure and the residue was extracted withCH2Cl2 (2 � 3 ml). The CH2Cl2 fraction was concentrated under re-duced pressure, and the residue was dissolved in CH3CN (100 ll) toyield the non-polar fraction for HPLC and LC-MS analysis. Theinsoluble residue was dissolved in MeOH–H2O (1:1, 1.0 ml) andwas filtered to yield the polar fraction for HPLC and LC-MS analysis.

4.4. Isolation of new compounds 15, 16, 19, 20 and 21

Leaves of T. salsuginea Shangdong were irradiated under a UVlamp (kmax 254 nm) as described above. After two days, leaveswere collected (120 g) were first frozen in liquid N2, ground andthen extracted with MeOH (2 � 1 liter). The extract was concen-trated under reduced pressure and the residue was partitioned be-tween CH2Cl2 and H2O (1:1, 100 ml). The aqueous solution wasconcentrated and the residue was fractionated on a Diaion HP-20column eluted with H2O/MeOH (100% to 50%, 100 ml fractions)to yield 13 fractions. Frs. 7 and 8 (60:40, 200 ml, 940 mg) werecombined and fractionated repeatedly on reversed phase silicagel and Sephadex LH-20 to yield 15 (6 mg), 16 (16 mg) and 21(77 mg). Fr. 9 was fractionated on silica gel to yield eight fractions,of which Fr. 6 was fractionated on Sephadex LH-20 followed byPTLC to yield 19 (4.0 mg) and 20 (4.4 mg).

4.4.1. 40-O-(E)-sinapoyl-7-methoxyisovitexin-200

-O-�-D-glucopyranoside (15)HPLC tR = 41.6 min (method B). HR-ESI-MS (negative mode):

calc. for C39H41O19 m/z 813.2247, found 813.2252. ESI-MS (posi-tive): m/z 815.3. ESI-MS (negative): m/z 813.3. In 1H and 13CNMR (CD3OD and DMSO-d6), see Tables 1 and 2. tmax (KBr)/cm�1:3352, 2938, 1696, 1652, 1606, 1513, 1490, 1450, 1348, 1175,1079, 1022, 838. kmax (MeOH)/nm: 272 (log e, 4.35), 330 (log e,4.45). [a]D � 52 (c 0.15, MeOH).

4.4.2. 40-O-(E)-sinapoylisovitexin-2-O-b-D-glucopyranoside (16)HPLC tR = 39.2 min (method B). HR-ESI-MS (negative mode):

calc. for C38H39O19 m/z 799.2091, found 799.2102. ESI-MS (posi-tive): m/z 801.3. ESI-MS (negative): m/z 799.3. For 1H NMR and

13C NMR (D2O), see Tables 1 and 2. tmax (KBr)/cm�1: 3352, 2917,1695, 1651, 1608, 1514, 1489, 1456, 1350, 1177, 1077, 1017,836. kmax (MeOH)/nm: 272 (log e, 4.30), 331 (log e, 4.50).

4.4.3. 4-O-b-D-glucopyranosyl-7-hydroxymatairesinol (19)HPLC tR = 34.8 (method B). HR-ESI-MS (negative mode): calc. for

C26H31O12 m/z 535.1821, found 535.1820. ESI-MS (positive): m/z554.2. ESI-MS (negative): m/z 535.2, 373.1. 1H NMR (D2O): d 6.91(d, J = 8.0 Hz, 1H), 6.68 (d, J = 8.0 Hz, 1H), 6.53 (d, J = 8.0 Hz, 1H),6.39 (s, 2H), 6.33 (d, J = 8.0 Hz, 1H), 5.03 (d, J = 6.0 Hz, 1H), 4.78 (s,1H), 4.53 (t, J = 8.0 Hz, 1H), 4.39 (d, J = 7.5 Hz, 1H), 3.63 (m, 2H),3.79 (m, 2H), 3.59 (s, 3H), 3.56 (s, 3H), 3.42 (m, 1H), 2.79 (d,J = 11.0 Hz, 1H), 2.65 (d, J = 10.0 Hz, 1H), 2.58 (s, 1H), 2.46 (t,J = 11.0 Hz, 1H). 13C NMR (D2O): d 183.3, 148.1, 147.0, 144.3,143.1, 136.1, 130.1, 121.6, 117.4, 114.9, 114.7, 112.4, 109.3, 100.1,76.0, 75.5, 73.4, 72.9, 71.7, 69.3, 60.5, 55.5, 55.2, 44.9, 42.2, 35.1.tmax (KBr)/cm�1: 3352, 2945, 1754, 1599, 1451, 1270, 1124, 1071,1028. kmax (MeOH)/nm: 278 (log e, 4.14). [a]D � 108 (c 0.04, MeOH).

4.4.4. 50-O-b-D-glucopyranosyldihydroneoascorbigen (20)HPLC tR = 41.0 min (method B). HR-ESI-MS (negative mode):

calc. for C22H28NO12 m/z 498.1616, found 498.1609. ESI-MS (posi-tive): m/z 500.2, 338.2, 307.1, 160.1, 130.1. ESI-MS (negative): m/z 498.2. 1H NMR (D2O): d 7.55 (d, J = 8.0 Hz, 1H), 7.44 (d,J = 8.0 Hz, 1H), 7.36 (s, 1H), 7.22 (t, J = 8.0 Hz, 1H), 7.08 (t,J = 8.0 Hz, 1H), 4.60 (d, J = 8.5 Hz, 1H), 4.48 (d, J = 8.0 Hz, 1H),4.16 (dd, J = 3.0, 8.5 Hz, 1H), 3.99 (s, 3H), 3.91 (m, 1H), 3.78 (m,2H), 3.60 (m, 2H), 3.39 (m, 1H), 3.36 (m, 4H), 3.19 (m, 2H). 13CNMR (D2O): d 177.9, 131.7, 124.3, 123.9, 122.7, 120.0, 119.2,108.5, 103.3, 103.2, 79.9, 78.3, 78.3, 75.8, 75.6, 74.2, 73.2, 69.4,65.9, 60.7, 60.5, 27.1. tmax (KBr)/cm�1: 3352, 2933, 1777, 1511,1451, 1121, 1073, 1029, 743. kmax (MeOH)/nm: 276 (log e, 3.92).[a]D � 12 (c 0.05, MeOH).

4.4.5. 3-O-b-D-glucopyranosylthiane (21)HPLC tR = 16.2 min (method B). HR-ESI-MS (negative mode):

calc. for C11H19O6S m/z 279.0907, found 279.0908. ESI-MS (posi-tive): m/z 281.0, 119.1. ESI-MS (negative): m/z 279.0. 1H NMR(D2O): d 1.36 (m, 1H), 1.63 (m, 1H), 1.93 (m, 1H), 2.03 (m, 1H),2.39 (m, 2H), 2.44 (m, 1H), 2.74 (m, 1H), 3.12 (t, J = 8.0 Hz, 1H),3.28 (m, 1H), 3.32 (m, 1H), 3.36 (dd, J = 9.0, 9.0 Hz, 1H), 3.59 (dd,J = 12.5, 5.5 Hz, 1H), 3.78 (dd, J = 12.5, 2.0 Hz, 1H), 3.87 (m, 1H),4.49 (d, J = 8.0 Hz, 1H). 13C NMR (D2O): d 100.3, 76.2, 75.9, 75.6,72.9, 69.6, 60.7, 31.7, 31.2, 27.1, 27.0. tmax (KBr)/cm�1: 3409,2932, 2072, 1634, 1156, 1075, 1017. [a]D + 32 (c 0.3, MeOH).

Acknowledgements

Financial support for the authors’ work was obtained from theNatural Sciences and Engineering Research Council of Canada (Dis-covery and AGENO Grants to M.S.C.P.), the Canada Research ChairsProgram, Canada Foundation for Innovation, the SaskatchewanGovernment, and the University of Saskatchewan. We acknowl-edge the technical assistance of K. Brown (NMR) and K. Thoms(MS), from the Department of Chemistry.

Appendix A. Supplementary data

Supplementary data associated with this article can be found, inthe online version, at doi:10.1016/j.phytochem.2009.12.008.

References

Amtmann, A., 2009. Learning from evolution: Thellungiella generates newknowledge on essential and critical components of abiotic stress tolerance inplants. Molecular Plant 2, 3–12.

Page 9: Metabolic responses of Thellungiella halophila/salsuginea to biotic and abiotic stresses: Metabolite profiles and quantitative analyses

M.S.C. Pedras, Q.-A. Zheng / Phytochemistry 71 (2010) 581–589 589

Benn, M.H., Singh, V.K., 1986. A simple, biogenetically modeled synthesis of 4-(methylthio)butyl thiocyanate: the reaction of thiocyanate anion with S-methyl-(1,n)-epithionium ions. Canadian Journal of Chemistry 64, 940–942.

Bylka, W., Franski, R., Stobiecki, M., 2002. Differentiation between isomeric acacetin-6-C-(600-O-malonyl)glucoside and acacetin-8-C-(600-O-malonyl)glucoside byusing low-energy CID mass spectra. Journal of Mass Spectrometry 37, 648–650.

Cheminat, A., Zawatzky, R., Becker, H., Brouillard, R., 1988. Caffeoyl conjugates fromEchinacea species: structures and biological activity. Phytochemistry 27, 2787–2794.

Freudenberg, K., Knof, L., 1957. The lignans of fir wood. Chemische Berichte 90,2957–2969.

Inan, G., Zhang, Q., Li, P., Wang, Z., Cao, Z., Zhang, H., Zhang, C., Quist, T.M., Goodwin,S.M., Zhu, J., Shi, H., Damsz, B., Charbaji, T., Gong, Q., Ma, S., Fredricksen, M.,Galbraith, D.W., Jenks, M.A., Rhodes, D., Hasegawa, P.M., Bohnert, H.J., Joly, R.J.,Bressan, R.A., Zhu, J.-K., 2004. Salt cress: a halophyte and cryophyte Arabidopsisrelative model system and its applicability to molecular genetic analyses ofgrowth and development of extremophiles. Plant Physiology 135, 1718–1737.

Jurd, L., Geissman, T.A., Seikel, M.K., 1957. Flavonoid constituents of Spirodelaoligorrhiza. II. Flavone constituents. Archives Biochemistry Biophysics 67, 284–297.

Kjaer, A., Larsen, I., Gmelin, R., 1955. Isothiocyanates. XIV. 5-Methylthiopentylisothiocyanate, a new mustard oil present in nature as a glucoside(glucoberteroin). Acta Chemica Scandinavica 9, 1311–1316.

Montaut, S., Grandbois, J., Righetti, L., Barillari, J., Iori, R., Rollin, P., 2009. Updatedglucosinolate profile of Dithyrea wislizenii. Journal of Natural Products 72, 889–893.

Pang, Q., Chen, S., Li, L., Yan, X., 2009. Characterization of glucosinolate-myrosinasesystem in developing salt cress Thellungiella halophila. Physiologia Plantarum136, 1–9.

Pedras, M.S.C., Adio, A.M., 2008. Phytoalexins and phytoanticipins from the wildcrucifers Thellungiella halophila and Arabidopsis thaliana: rapalexin A, wasalexinsand camalexin. Phytochemistry 69, 889–893.

Pedras, M.S.C., Chumala, P.B., Suchy, M., 2003. Phytoalexins from Thlaspi arvense, awild crucifer resistant to virulent Leptosphaeria maculans: structures, synthesesand antifungal activity. Phytochemistry 64, 949–956.

Pedras, M.S.C., Adio, A.M., Suchy, M., Okinyo, D.P.O., Zheng, Q.A., Jha, M., Sarwar,M.G., 2006. Detection, characterization and identification of cruciferphytoalexins using high performance liquid chromatography with diode arraydetection and electrospray ionization mass spectrometry. Journal ofChromatography A 1133, 391–411.

Pedras, M.S.C., Zheng, Q.A., Sarma-Mamillapalle, V.K., 2007. The phytoalexins fromBrassicaceae: structure, biological activity, synthesis and biosynthesis. NaturalProduct Communications 2, 319–330.

Pedras, M.S.C., Zheng, Q.A., Gadagi, R.S., Rimmer, S.R., 2008a. Phytoalexins and polarmetabolites from the oilseeds canola and rapeseed: differential metabolicresponses to the biotroph Albugo candida and to abiotic stress. Phytochemistry69, 894–910.

Pedras, M.S.C., Zheng, Q.A., Strelkov, S., 2008b. Metabolic changes in roots of theoilseed canola infected with the biotroph Plasmodiophora brassicae:phytoalexins and phytoanticipins. Journal of Agriculture and Food Chemistry56, 9949–9961.

Pedras, M.S.C., Zheng, Q.A., Shatte, G., Adio, A.M., 2009. Photochemical dimerizationof wasalexins in UV-irradiated Thellungiella halophila and in vitro generatesunique cruciferous phytoalexins. Phytochemistry 70, 2010–2016.

Rimmer, S.R., Mathur, S., Wu, C.R., 2000. Virulence of isolates of Albugo candida fromWestern Canada to Brassica species. Canadian Journal of Plant Pathology 22,229–235.

Tabopda, T.K., Ngoupayo, J., Liu, J., Mitaine-Offer, A.C., Tanoli, S.A., Khan, S.N., Ali,M.S., Ngadjui, B.T., Tsamo, E., Lacaille-Dubois, M.A., Luu, B., 2008. Bioactivearistolactams from Piper umbellatum. Phytochemistry 69, 1726–1731.

Takahiro, H., Yun, Y.S., Kunugi, A., 2005. Five novel flavonoids from Wasabiajaponica. Tetrahedron 61, 7037–7044.

Tanaka, N., Yuhara, H., Wada, H., Murakami, T., Cambie, R.C., Braggins, J.E., 1993.Chemical and chemotaxonomical studies of ferns. Part 82. Phenolic constituentsof Pteridium esculentum. Phytochemistry 32, 1037–1039.

Ward, R.S., 1997. Lignans, neolignans and related compounds. Natural ProductReports 14, 43–74.

Xia, Z.Q., Costa, M.A., Proctor, J., Davin, L.B., Lewis, N.G., 2000. Dirigent-mediatedpodophyllotoxin biosynthesis in Linum flavum and Podophyllum peltatum.Phytochemistry 55, 537–549.