Phytochemistry 99 (2014) 127–134

Contents lists available at ScienceDirect

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

Phytochemical analysis of Rosa hybrida cv. ‘Jardin de Granville’ by HPTLC, HPLC-DAD and HPLC-ESI-HRMS: Polyphenolic fingerprints of six plant organs Ludivine Riffault a,b, Emilie Destandau a,⇑, Laure Pasquier b, Patrice André b, Claire Elfakir a a b

Université d’Orléans, Institut de Chimie Organique et Analytique, UMR 7311, rue de Chartres, 45067 Orléans cedex 2, France LVMH recherche, département Innovation Ethnobotanique, 185 avenue de Verdun, 45800 Saint-Jean-de-Braye, France

a r t i c l e

i n f o

Article history: Received 24 April 2013 Received in revised form 20 December 2013 Available online 23 January 2014 Keywords: Roses Polyphenols MAE HPTLC HPLC-DAD HPLC-ESI-HRMS

a b s t r a c t The Rosa hybrida cultivar ‘Jardin de Granville’, a delicate clear pink flower, is here investigated through a progressive analytical strategy using complementary phytochemical screening methods in order to characterize the polyphenol content of several parts of the plant. The microwave hydro-ethanolic extract analysis of six different parts of the plant, carried out by High Performance Thin Layer Chromatography (HPTLC) and High Performance Liquid Chromatography coupled with a Diode Array Detector (HPLC-DAD) enabled initial identification of the polar molecular families present in each organ, namely tannins and flavonoids (quercetin and kaempferol derivatives). The HPLC fingerprints displayed different profiles for each organ, attesting to the original composition and potential valuation of the different plant parts. More detailed analyses of the extracts were then carried out by High Performance Liquid Chromatography coupled with electrospray ionization (ESI) mass spectrometry with a Q-TOF analyzer (ESI-HR-Q-TOF). Around 60 compounds were identified, mainly gallo-tannins, ellagi-tannins, catechin derivatives and glycoside derivatives of quercetin and kaempferol. Some compounds such as hyperoside or ellagic acid appeared to be ubiquitous and were found in abundance in each plant part. Woods were the richest organ in catechin and proanthocyanidin derivatives while kaempferol derivatives were more numerous and abundant in bud and flower parts. Ó 2014 Elsevier Ltd. All rights reserved.

Introduction Roses are principally appreciated for their fragrance, beauty and for ornamental gardening. There are nowadays thousands of horticultural varieties of rose, resulting sometimes from numerous hybridizations. The reference book Modern Roses XI describes over 24,000 cultivars (Cairns et al., 2000). The literature concerning the molecular content of roses focuses mainly on the volatile compounds involved in fragrance and on the anthocyanin molecules involved in plant coloration. Anthocyanins have been extensively studied (Mikanagi et al., 2000; Cai et al., 2005) and contribute to the wide color diversity of the petal and flower stages of development (Schmitzer et al., 2010, 2012). Some flavonols contained in flower organs have also been identified by mass spectrometry (Cai et al., 2005; Kumar et al., 2009). These studies mostly concern the flowers of botanical species, in particular Rosa damascena, one of the most widely used roses in the perfume industry (Panda, 2006). Several studies of Rosa canina, Rosa chinensis or Rosa rugosa ⇑ Corresponding author. Tel.: +33 2 38 41 70 74; fax: +33 2 38 41 72 81. E-mail address: [email protected] (E. Destandau). 0031-9422/$ - see front matter Ó 2014 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.phytochem.2013.12.015

have also been conducted (Mikanagi et al., 2000; Cai et al., 2005; Kumar et al., 2009). However roses possess a huge diversity of secondary metabolites: phenolic compounds (hydrolysable and condensed tannins, flavonoids, phenyl propanoid derivatives), terpenoids (hemi and monoterpenes, carotenoids, triterpenes, steroids), fatty acids, etc. which give them a large range of biological activities (Hashidoko, 1996). Compounds such as rutin, quercetin and kaempferol glycosylated derivatives have also been characterized and contribute to the color of flowers (Hashidoko, 1996; Mikanagi et al., 2000; Kumar et al., 2009). Flavonoids in roses exhibit a variety of biological activities, especially antioxidant properties (Cai et al., 2005; Kumar et al., 2009). Roses are therefore an ingredient of great interest for cosmetic applications, not only for perfume but also for skin care products. While some studies have described part of the molecular content of rose flowers, the compounds contained in the other parts of the plant remain largely unknown. A few papers concerning the total phenolic content of leaves have been published but the identification of all the constituents is incomplete (Nowak and Gawlik-Dziki, 2007; Ghazghazi et al., 2012).

128

L. Riffault et al. / Phytochemistry 99 (2014) 127–134

The present study concerns ‘Jardin de Granville’, a new clear pink variety of rose that was created for Christian Dior Perfumes by the company ‘Les Roses Anciennes, André EVE’, located in Pithiviers-le-Vieil (Loiret, France). Its breeders describe the plant as very vigorous and highly resistant to common rose diseases, especially to oidium (Sphaerotheca pannosa var. rosae) and black spot disease (Marssonina rosae). The biological activities of extracts of this rose variety tested by the Christian Dior Perfumes biological department on skin cell models showed attractive results concerning anti-inflammatory properties, stimulation of dermal extracellular matrix constituents, and skin cell maturation and cohesion. As a result, ‘Jardin de Granville’ extracts have been incorporated in anti-aging luxury creams for skin care. The goal of this study is to characterize the polyphenol composition of the original cultivar ‘Jardin de Granville’, in order to gain better insight into the molecular content of the plant that could contribute to its anti-aging properties. As the phytochemical investigation covered all the parts of the plant, it could provide guidelines for the future valuation of organs that are currently not valorized. Microwave Assisted Extractions (MAE) and analysis conditions were optimized to determine the main polyphenols. Various complementary analytical methods were used such as HPTLC, HPLCDAD and HPLC-ESI-HRMS, to obtain the organ fingerprints and to characterize the major polyphenols. HPTLC is a rapid and economic solvent method, which gives fast access to the molecular classes present in extracts thanks to the use of specific reagents (Bhandari et al., 2007). Reversed phase HPLC coupled with UV detection is suitable for the separation of phenolic compounds containing a chromophore group, while DAD enables the characteristic absorbance spectra of each molecule to be recorded. Thanks to its higher resolution than HPTLC, HPLC analysis allows a more refined identification of the polyphenol compounds within the family. Lastly, HRMS analysis with an electrospray ionization source that is well-adapted to phenolic ionization made it possible to accurately

A

B

E

determine the mass and to propose a molecular formula, leading to a tentative identification of the different compounds separated. This metabolite investigation thus compares for the first time the polyphenol content of several organs of this original rose variety. Results and discussion Extraction of phytochemicals To obtain the most representative extract for the polyphenol content study, different extraction conditions (solvent and method) were investigated. Five extraction methods were compared: maceration, Ultra-Sound (US), Accelerated Solvent Extraction (ASE), Soxhlet and Microwave Assisted Extraction (MAE). No significant differences were observed between the chromatographic fingerprints with these five methods. We therefore selected Microwave Assisted Extraction (MAE) to manage all the further extracts due to its rapidity, facility of use, and efficiency in extraction yields. Thanks to a carrousel which can carry up to 12 reactors, making simultaneous extractions possible, a good time saving was achieved. The extraction solvent consisting of a mixture of EtOH/H2O: 90/10 (v/v) was also optimized for the recovery of the main polyphenols, which are relatively polar compounds that are easily soluble in this solvent system. Moreover, this choice of a safe solvent makes it possible to transpose the extraction system to the cosmetic industry. Six different organs of ‘Jardin de Granville’, as shown in Fig. 1, were collected and stored before microwave extraction. Some of the organs were air dried (woods), other were lyophilized when possible (flowers and leaves). Shoots and buds, which are more fragile, were kept frozen and were powdered in liquid nitrogen just before performing the extraction. For each plant part, yellow– brown extracts were obtained. The extraction yields were calculated as the ratio of the weight of dried extract on the initial plant

C

D

F

Fig. 1. Pictures of the six different organs of ‘Jardin de Granville’. (A) wood (w), (B) shoots (sh), (C) early buds (eb), (D) buds before flowering (bbf), (E) flowers (fl), (F) leaves (le).

129

L. Riffault et al. / Phytochemistry 99 (2014) 127–134

weight. For woods, flowers and leaves, the initial plant was weighed as a dried material whereas shoots and buds were considered as fresh plant material and their weight takes into account the natural water inside the plant organ. This could explain the lower extraction yields observed in Table 1 for these non-dried organs. To obtain comparable conditions, all the extract solutions analyzed by the different chromatographic methods were prepared at the same concentration of 10 mg mL 1 from the dried extracts, thus avoiding any bias due to the differences in the extraction yield. HPTLC fingerprints HPTLC is a rapid and economic solvent method that enables several samples to be analyzed simultaneously. An initial rapid screening of the molecular families present in the different organ extracts was done using specific reagents. Firstly, an exploratory analysis of the polar families present in the hydro-alcoholic extracts was performed. As expected, the first HPTLC investigation revealed the presence of sugars (mainly mono and disaccharides), amino acids in low amounts, and phenolic constituents in the extracts. Fig. 2 shows the system used to reveal polyphenols. Three reference standards were analyzed, corresponding to two polyphenol classes: tannins, (track F: epicatechin gallate), and flavonoids (track G: kaempferol-3-O-rutinoside and track H: quercetin-3-Oglucoside). The presence of these compounds in rose extracts has already been reported in the literature (Kumar et al., 2009; Hashidoko, 1996). Different HPTLC profiles were observed for the six ‘Jardin de Granville’ organ extracts. The presence of tannins (blue fluorescence) was detected in the six extracts (Rf between 0.5 and 0.9), and a more intense spot at Rf 0.9 was observed in the extracts of shoots (track 2), early buds (track 3) and buds before flowering (track 4). In woods (track 1), the concentration of flavonoids seems to be low, as only two tiny yellow bands are visible (Rf 0.5 and 0.52). In shoots (track 2), the same yellow strips are present and they are more intense. They correspond to quercetin derivatives (yellow fluorescence). In early buds (track 3), this yellow fluorescence decreases but a green one, characteristic of kaempferol derivatives, appears (Rf 0.41 and 0.49). This kind of compound has already been observed in some rose extracts (Bhandari et al., 2007). In buds before flowering (track 4) and in flowers (track 5) the green fluorescence increases, and a supplementary band at Rf 0.45 can be observed. It seems that kaempferol derivatives are more abundant at the two bud stages and in flowers. In leaves (track 6), the green fluorescence is weaker than in buds before flowering or flowers, but blue spots in the tannin zone are enhanced. On all the organ profiles, no pink-red spots corresponding to the presence of anthocyanins were observed. HPLC-DAD analysis A system adapted to an exhaustive separation of polyphenolic compounds was developed by HPLC-DAD from six polyphenolic standards (Fig. 3). Based on their separation, each chromatogram obtained for the six rose extracts (Fig. 4) can be divided into three parts: the first one before 31 min, corresponding to tannin elution

Fig. 2. HPTLC polyphenol analysis, revelation system Neu + PEG reagent, k = 366 nm. W RP18 HPTLC plates 10  20 cm. Eluent: ACN/H2O/HCOOH (50/50/ 5). Track: 1: woods, 2: shoots, 3: early buds, 4: buds before flowering, 5: flower, 6: leaves, F: epicatechin gallate, G: kaempferol-3-O-rutinoside, H: quercetin-3-Oglucoside.

and including gallic acid (1), catechin (2) and ellagic acid (3); the second intermediate zone between 27 and 31 min corresponding to the elution of a tannin and flavonoid mix; and the last one, after 27 min, corresponding to flavonoid elution including quercetin-3O-glucoside (4), kaempferol-3-O-glucoside (5) and tiliroside (6) (kaempferol-3-O-coumaroyl glucoside). All the six different organ fingerprint profiles (Fig. 4) show chromatographic peaks eluted over the same wide range of polarity. The wood extract displays the poorest chromatographic content and the peak intensities are slightly more intense in the first part of the chromatogram, which corresponds to the tannin part. This correlates well with the HPTLC blue fluorescence observable for this extract. Shoots and early buds also exhibit relatively low molecular amounts and have similar peak intensities between the tannin and the flavonoid zone. Buds before flowering show a chromatographic profile close to that of the early buds but peak intensities in the tannin zone are higher, especially for compound A. The UV spectrum of this most intense peak is plotted in Fig. 4 and shows a maximum of absorbance at 278 nm. This could correspond to a gallic tannin, given that gallic acid has an absorbance maximum at 271 nm. The flower and leaf extract chromatograms seem to be closer with a higher intensity in the flavonoid zone while the presence of tannins is still observed, which is in good agreement with the previous observations conducted by HPTLC. Moreover, they exhibit the same intense flavonoid peak (compound B). The UV spectrum of this compound B is shown in Fig. 4 (black trace). This constituent shows two absorbance maxima at 264 and 347 nm. The same absorbance maxima and retention time were recorded for the standard of kaempferol-3-Oglucoside (blue trace). From the six chromatographic fingerprints, about 40 chromatographic peaks were separated, showing the richness of the extracts. Flavonoids and more specifically quercetin and kaempferol derivatives seem to be present in view of the characteristic maximum of absorbance shown by some compounds. This is in accordance with the literature (Grossi et al., 1998; Cai et al., 2005; Hanhineva et al., 2008; Kumar et al., 2009). A tentative detection of anthocyanins in

Table 1 Extraction yields of the 6 organ extracts of ‘Jardin de Granville’. Extraction solvent: EtOH/H2O: 90/10. Organs

w

sh*

eb*

bbf*



le

Extraction yields

6.4%

4.5%

3.8%

4.3%

31.4%

23.9%

w: woods, sh: shoots, eb: early buds, bbf: buds before flowering, fl: flowers, le: leaves. * indicates the non-dried organs.

130

L. Riffault et al. / Phytochemistry 99 (2014) 127–134

Flavonoid zone

OH OH

O

Tannin zone

1000

OH

OH

HO

900

O

O

O

OH

O

800

HO

OH

O

OH

OH

OH

HO

700

OH

1

mAU

600

2

500

3 OH

OH

OH

OH

HO

400 HO

HO

O

O

O O

300 OH

O

200 OH

OH

O

OH

OH

O

OH

O

O

OH

O

O

O O

OH

100 OH

OH

O

HO

OH

0 0

5

10

15

20

25

30 35 Minutes

40

45

50

55

60

4

OH

OH

OH

5

6

Fig. 3. HPLC chromatogram obtained for six polyphenolic reference standards. Nucleodur sphinx (150  4.6 mm, 5 lm). Mobile phase A: 0.1% formic acid in water, B: 0.1% formic acid in methanol, gradient elution: 0–17 min: 5–36% B, 17–25 min: 36% B, 25–35 min: 36–50% B, 35–45 min, 50–70% B, 45–50 min: 70–90% B, 50–60 min, 90% B. Flow rate: 1 mL min 1, T = 25 °C. Vinj = 20 lL. Detection UV k = 270 nm. 1: gallic acid, 2: catechin, 3: ellagic acid 4: quercetin-3-O-glucoside, 5: kaempfarol-3-O-glucoside, 6: tiliroside (kaempferol-3-O-coumaroyl-glucoside).

Fig. 4. HPLC analyses of the six different organs of ‘Jardin de Granville’ and absorption spectrum of compounds A and B. Nucleodur sphinx (150  4.6 mm, 5 lm). Mobile phase A: 0.1% formic acid in water, B: 0.1% formic acid in methanol, gradient elution: 0–17 min: 5–36% B, 17–25 min: 36% B, 25–35 min: 36–50% B, 35–45 min, 50–70% B, 45– 50 min: 70–90% B, 50–60 min, 90% B. Flow rate: 1 mL min 1, T = 25 °C. Vinj = 20 lL. Detection UV k = 270 nm.

the extracts was also done, by checking the DAD recording between 500–550 nm. Some minor chromatographic peaks were detected but with a weak intensity close to the background noise level. This result is surprising since this molecular family is known to be natural rose pigments. It could be explained by the extraction process selected and a long-step sample preparation, which may have poorly preserved the structural integrity of these compounds. As the aim of this study was to identify the major compounds present in ‘Jardin de Granville’ organ extracts, no further investigation on anthocyanin composition was performed. HPLC-HRMS analysis An HPLC-HRMS analysis was then carried out to obtain some complementary structural information that would allow a tentative identification of the main compounds, based on their retention

time in comparison with standards and their molecular mass and fragmentation pattern. Some mass parameters concerning plant polyphenol analysis have already been optimized and have been reported in the literature (Cai et al., 2005; Kumar et al., 2009; Steinmann and Ganzera, 2011). The most widely used ionization source for this type of molecule is electrospray. As some studies detail analyses in positive ionization mode and others in negative mode, extracts were first analyzed with an ESI ion source by direct full scan mass spectrometry in negative ionization mode. Compared to the positive ionization mode, the negative mode allowed a more sensitive detection of target solutes and more refined mass spectra were obtained. It also facilitated the detection of less abundant target compounds in ‘Jardin de Granville’ and made the confirmation of molecular ions for the identification of each peak detected easier. Mass analyses showed the presence of more than one compound under some

131

L. Riffault et al. / Phytochemistry 99 (2014) 127–134

chromatographic peaks, demonstrating that the HPLC coelution of some molecules can be resolved thanks to mass spectrometry. Table 2 shows the mass data of 60 compounds detected by mass spectrometry under the 41 major chromatographic peaks detected on HPLC analyses. For each peak, the different ions present under the same peak are displayed. A compound number has been attributed to each ion. The negative ionization mode gives the deprotonated molecular ion [M H] . The [M H] accurate mass measured is indicated with the corresponding molecular formula proposed by the software. For each formula, the precision errors (Dmass (ppm)) with respect to the theoretical accurate mass are given. All error values are below ±2 ppm, indicating that the proposed formulae are reliable. A second analysis was carried out with the AutoMSMS acquisition mode, in which the precursor ions are isolated then fragmented, leading to the specific fragmentation patterns for each solute. The fragmentation mechanisms of polyphenols are now well-known (March and Brodbelt, 2008). Comparing these experimental fragmentation patterns to literature information, a tentative identification of most of the compounds detected (Table 2) in the different organ extracts can be proposed. The compound identification was confirmed by comparing their HPLC retention time and mass spectrum with those of standards. In the tannin zone, gallic, ellagic and quinic acid derivatives, described in the literature as hydrolysable tannins (Cai et al., 2005; Khanbabaee and Van Ree, 2001), were mainly identified. Different characteristic fragments observed in the analyses confirmed the identification of these compounds. A mass difference of 169 u (C7H5O5) or 152 u (C7H4O4) was observed when a galloyl unit was lost from the [M H] ion (8, 14, 15, 18, 20, 23, 31, 34. . .). Tannins with a quinic acid unit were characterized by a specific ion

fragment at m/z 191 (1, 3, 8, 14). Lastly, tannins with the hexahydroxydiphenoyl (HHDP) unit showed an ion fragment at m/z 301 (12, 15, 20, 21, 22, 25, 30, 31, 33). Condensed tannins, identified as proanthocyanidin polymers and catechin derivatives, showed a fragmentation pattern with successive loss of 289 u corresponding to the loss of catechin units (11, 17, 23, 24, 29, 34). The flavonoids identified in the extracts appeared to belong to two flavonol based genins. Some of the molecules were kaempferol derivatives and showed the characteristic fragment ion m/z 285 or 284 (43, 47, 49, 50, 51, 53, 54, 55, 56, 57, 58, 59). Quercetin based flavonoids were also detected and exhibited the characteristic genin fragment at m/z 301 or 300 (41, 44, 45, 52). Some of them seemed to be glycosylated. Molecules which contained a hexose substitute presented a mass loss of 162 u (42, 43, 44, 47, 49, 50, 53, 54, 58), and those with a deoxyhexose such as a rhamnose lost a fragment of 146 u (44, 52). Other molecules with a pentose substitute had a mass loss of 132 u (41, 56). Another case detected was the presence of a hexuronic acid substitute with the mass loss of 176 u (45, 51). Some compounds such as galloyl tannins or quercetin and kaempferol based flavonoids have already been characterized by mass spectrometry in some rose and strawberry extracts (Rosaceae family) (Hashidoko, 1996; Cai et al., 2005; Hanhineva et al., 2008; Kumar et al., 2008, 2009), which is in accordance with the identifications proposed in Table 2. These identifications confirmed that the first part of the chromatogram (0–31 min) is representative of tannins while the end of the chromatogram (23–42 min) corresponds to flavonoids, with a small intermediate mixed zone. This correlates well with the HPLC-DAD observations and HPTLC color patterns. For each molecule detected in one organ, the extracted ion chromatogram was displayed in the other parts of the plant to

Table 2 Mass analysis of six organ extracts of ‘Jardin de Granville’. The tentative identification of compounds followed by * has been confirmed by comparison with a reference standard. peak

1

retention compound accurate mass time (min) number measured [M - H] -

2.0

molecular formula

mass (ppm)

auto MS-MS fragment ions

tentative identification

0.08

191.05595: quinic acid unit

quinic acid tannin derivative quinic acid tannin derivative

1

875.28847

C 31 H 55 O 28

2

533.17260

C 19 H 33 O 17

-0.51

191.05640: quinic acid unit

3

191.05614

C 7 H 11 O 6

-0.15

127.04060: quinic acid fragmentation

quinic acid

4

383.11961

C 14 H 23 O 12

-0.28

191.05649: quinic acid unit

quinic acid tannin derivative

5

371.11913

C 13 H 23 O 12

1.00

191.05632: quinic acid unit

quinic acid tannin derivative

2

4.5

6

331.06698

C 13 H 15 O 10

0.28

169.01350: galloyl unit [M-H - 162(hexose)]

galloyl hexoside

3

5.2

7

169.01432

C7H5O5

-0.43

125.02334: [M-H - COOH]

gallic acid *

4

6.4

8

343.06756

C 14 H 15 O 10

-1.43

191.05556: quinic acid unit [M-H -152(galloyl unit)]

galloyl quinic acid

5

7.1

9

783.06863

C 34 H 23 O 22

0.02

301.00029: HHDP unit

bis-HHDP hexoside coumaroyl quinic acid (chlorogenic acid)

6

9.6

10

353.08739

C 16 H 17 O 9

1.18

191.05580: quinic acid unit [M-H - 162(coumaroyl unit)]; 161.02405: coumaroyl unit

7

10.9

11

577.13421

C 30 H 25 O 12

-0.20

289.07090: catechin unit; 245.08152: catechin rearrangement

proanthocyanidin dimer (isomer)

8

11.6

12

785.08457

C 34 H 25 O 22

-1.64

633.06924: [M-H - 152(galloyl unit)]; 615.06291: [M-H - 170(galloyl unit)]; 483.08111: [M-H - 302(HHDP unit)]; 300.99891: HHDP unit; 169.01237: galloyl unit

HHDP di-galloyl hexoside (isomer)

9

12.7

13

289.07154

C 15 H 13 O 6

0.75

245.08100: catechin rearrangement

catechin *

14

495.07814

C 21 H 19 O 14

-0.22

191.05481: quinic acid unit; 169.01283: galloyl unit

di-galloyl quinic acid

15

633.07397

C 27 H 21 O 18

-1.00

463.05221: [M-H - 170(galloyl unit)]; 300.99880: HHDP unit

galloyl HHDP hexoside

16

952.10967

C 41 H 30 N O 26

-0.38

300.99988: HHDP unit; 169.01381: galloyl unit

galloyl HHDP derivative

17

577.13526

C 30 H 25 O 12

-0.20

289.07182: catechin unit; 245.08055: catechin rearrangement

proanthocyanidin dimer (isomer)

18

635.08897

C 27 H 23 O 18

-1.27

465.06639: [M-H - 170(galloyl unit)]; 313.05532: galloyl hexose - OH; 169.01363: galloyl unit

tri-galloyl hexoside

amount w

sh

le

nd

nd

nd nd

eb bbf

nd

nd

nd nd

nd

nd

10

13.7 19

952.10652

C 41 H 30 N O 26

-0.38

300.99900: HHDP unit; 169.01299: galloyl unit

galloyl HHDP derivative

11

14.5

20

785.08393

C 34 H 25 O 22

-1.64

633.06585: [M-H - 152(galloyl unit)]. 615.05961: [M-H - 170(galloyl unit)]. 483.07944: [M-H - 302(HHDP unit)]. 300.99813: HHDP unit; 169.01147: galloyl unit

HHDP di-galloyl hexoside (isomer)

12

14.5

21

953.09080

C 41 H 29 O 27

-0.66

617.08534: [M-H - 336(galloyl gallate unit)]; 300.99863: HHDP unit; 169.01390: galloyl unit

HHDP galloyl hexoside galloyl gallate

13

15.2

22

785.08492

C 34 H 25 O 22

-1.64

633.07496: [M-H - 152(galloyl unit)]. 615.06570: [M-H - 170(galloyl unit)]. 483.07395: [M-H - 302(HHDP unit)]. 300.99936: HHDP unit; 169.01202: galloyl unit

HHDP di-galloyl hexoside (isomer)

14

15.7

23

729.14443

C 37 H 29 O 16

0.26

577.13510: catechin dimer [M-H - 152(galloyl unit)]; 289.07122: catechin unit

galloyl proanthocyanidin dimer (isomer)

nd

nd

15

16.5

24

729.14592

C 37 H 29 O 16

0.26

577.13825: catechin dimer [M-H - 152(galloyl unit)]; 289.07054: catechin unit

galloyl proanthocyanidin dimer (isomer)

nd

nd

nd

fl

nd

nd

nd nd

(continued on next page)

132

L. Riffault et al. / Phytochemistry 99 (2014) 127–134

Table 2 (continued)

16

16,7

25

935.07814

C 41 H 27 O 26

1.57

300,99762: HHDP unit; 169,01286: galloyl unit

bis-HHDP galloyl hexoside (isomer)

26

635.08979

C 27 H 23 O 18

-1.27

465,06796: [M-H - 170(galloyl unit)]; 313,05604: galloyl hexose - OH; 169,01352: galloyl unit

tri-galloyl hexoside

27

341.08833

C 15 H 17 O 9

-1.54

161,02429: caffeoyl unit -2H [M-H -180(hexose)]

caffeoyl hexoside

17

17,1

28

909.10120

C 40 H 29 O 25

-0.94

300,99508: HHDP unit; 169,01538: galloyl unit

galloyl HHDP derivative

18

17,5

29

981.12191

C 43 H 33 O 27

-0.45

785.08120: [M-H - 196(dimethyl gallate)]; 633,08363: [M-H - 196 - 152(galloyl unit)]; 300,99846: HHDP unit; 195,03021: dimethyl gallate; 169,01337: galloyl unit

galloyl HHDP derivative

nd

19

17,5

30

935.07936

C 41 H 27 O 26

1.57

300,99849: HHDP unit; 169,01440: galloyl unit

bis-HHDP galloyl hexoside (isomer)

20

18,4

31

937.09449

C 41 H 29 O 26

0.82

301,00054: HHDP unit; 169,01359: galloyl unit

HHDP tri-galloyl hexoside

21

19,2

32

787.10047

C 34 H 27 O 22

-0.67

617,07232: [M-H - 170(galloyl unit)]; 313,05431: galloyl hexose; 169,01358: galloyl unit

tetra-galloyl hexoside

22

19,3

33

1105.09979

C 48 H 33 O 31

1.21

301,00023: HHDP unit; 169,01452: galloyl unit

bis-HHDP di-galloyl hexoside

23

19,9

34

441.08275

C 22 H 17 O 10

-0.06

289,07283: catechin [M-H - 152(galloyl unit)]; 245,08280: catechin rearangement; 169,01463: galloyl unit

galloyl catechin

24

20

35

1061.11029

C 47 H 33 O 29

0.95

300,99937: HHDP unit; 169,01416: galloyl unit

galloyl HHDP derivative HHDP tri-galloyl hexoside methyl galloyl gallate

25

21,3

36

1119.11601

C 49 H 35 O 31

0.69

26

20,5

37

463.05095

C 20 H 15 O 13

1.86

300,99868: HHDP unit [M-H - 162(hexose)]

HHDP hexoside

38

939.11081

C 41 H 31 O 26

0.10

769,09375: [M-H - 170(galloyl unit)]; 617,08375: [M-H - 170 -152(galloyl unit)]; 169,01377: galloyl unit

penta-galloyl hexoside

39

615.09956

C 28 H 23 O 16

-0.65

463,09037: [M-H - 152(galloyl unit)]; 313,05560: galloyl hexose unit [M-H -302(quercetin)]; 300,02682: quercetin - H

quercetin galloyl hexoside HHDP tri-galloyl hexoside dimethyl galloyl gallate

27

23,2

40

1133.13160

C 50 H 37 O 31

0.73

937,09864: [M-H -196(dimethyl gallate)], 767,07506: [M-H - 196 -170(galloyl unit)]; 300,99919: HHDP unit; 169,01467: galloyl unit

28

26,3

41

433.07746

C 20 H 17 O 11

0.40

300,02887: quercetin - H [M-H - 133(pentose)]; 271,02402, 255,03017, 243,02959 151,00436: quercetin rearangements

quercetin pentoside

29

27,0

42

463.08876

C 21 H 19 O 12

-1.21

300,02889: quercetin - H [M-H - 163(hexose), 271,02524, 255,02984, 243,02825, 151,00369: quercetin rearangements

hyperoside *

kaempferol galloyl hexoside

30

27,9

43

599.10401

C 28 H 23 O 15

-1.14

447,09929: , kaempferol-hexose [M-H - 152(galloyl unit)]; 313,05741: galloyl hexose; 285,04085: keampferol; 284,03358: kaempferol - H; 271,04309, 255,03081, 241,03616, 151,00386: kaempferol rearangements; 169,01445: galloyl unit

44

609.14624

C 27 H 29 O 16

-0.22

301,03400: quercetin; 300,02908: quercetin - H

rutin *

45

477.06748

C 21 H 17 O 13

-0.03

301,03671: quercetin [M-H - 176(hexuronic acid)]; 273,04258, 255,03039, 245,04549, 151,00410: quercetin rearangements

quercetin hexuronique acid

31

29,2

46

300.99922

C 14 H 5 O 8

-0.77

283,99590, 257,00948, 245,00918, 229,01342: ellagic acid rearrangements

ellagic acid *

32

30,3

47

599.10465

C 28 H 23 O 15

-1.14

447,09475: , kaempferol-hexose [M-H - 152(galloyl unit)]; 313,05597: galloyl hexose; 285,03906: keampferol; 284,03228: kaempferol - H; 255,02559, 241,03325, 151,00274: kaempferol rearangements; 169,01376: galloyl unit

kaempferol hexoside derivative

33

30,9

48

629.11505

C 29 H 25 O 16

-0.39

465,06842: [M-H - 164(coumaroyl unit)]; 169,01388: galloyl unit

di-galloyl coumaroyl hexoside

34

31,6

49

447.09319

C 21 H 19 O 11

-0.25

284,03179: kaempferol - H [M-H - 163(hexose)]; 255,02947, 227,03475: kaempferol rearangements

kaempferol hexoside

50

593.15164

C 27 H 29 O 15

-0.76

285,03870: kaempferol [M-H - 308(rutinose)]; 284,03199: kaempferol - H; 255,02975, 229,05075: kaempferol rearangements

kaempferol rutinoside

51

461.07264

C 21 H 17 O 12

-0.19

285,03871: kaempferol [M-H - 176(hexuronic acid)]; 257,04596, 229,05124: kaempferol rearangements

kaempferol glucuronide

35

36

32,8 52

447.09272

C 21 H 19 O 11

-0.25

300,02669: quercetin - H [M-H - 147(rhamnose)]; 271,02381, 255,02516: quercetin rearangements

quercetin rhamnoside

53

447.09305

C 21 H 19 O 11

-0.25

284,03213: kaempferol - H [M-H - 162(glucose)]; 255,02961, 151,00297, 227,03440: kaempferol rearangements

kaempferol glucoside *

54

599.10269

C 28 H 23 O 15

-1.14

313,05587: galloyl hexose; 285,04047: keampferol; 257,04644, 229,05177, 151,00354: kaempferol rearangements; 169,01179: galloyl unit

kaempferol derivative

55

745.16258

C 34 H 33 O 19

-0.57

459,11490: galloyl rutinose [M-H - 286(kaempferol)]; 285,04030: kaempferol

kaempferol galloyl rutinoside kaempferol pentoside

34,1

nd

nd

37

34,6

56

417.08253

C 20 H 17 O 10

0.45

284,03347: kaempferol - H [M-H - 133(pentose)]; 255,03017, 227,03497: kaempferol rearangements

38

37

57

431.09839

C 21 H 19 O 10

-0.05

284,03235: kaempferol - H [M-H - 146(rhamnose)]

afzelin *

39

40,7

58

593.12981

C 30 H 25 O 13

0.43

285,04083: kaempferol; 284,03209: kaempferol - H; 255,02911, 227,03497: kaempferol rearangements

tiliroside *

40

41,2

59

285.04042

C 15 H 9 O 6

0.16

255,01181, 227,03432: kaempferol rearrangements

kaempferol *

nd

nd

41

41,9

60

582.26132

C 34 H 36 N 3 O 6

-0.61

462,20543: [M-H - 120]; 342,14759: [M-H - 120 - 120]

tri-p coumaroyl spermidine *

nd

nd

very high amount (mass signal intensity > 1000 S/N); sity < 100 S/N); nd: not detected.

intermediate amount (mass signal intensity between 100 and 1000 S/N);

confirm its presence or absence. The amount of each molecule, based on the peak height compared to the background noise level (S/N) is reported in Table 2 for the six organs. Some of the compound proportions in the different organs appear to be ubiquitous; it means that they are abundant in all parts of the plant. This is the case for compounds 1, 5 and 8 corresponding to tannins derived from quinic acid, compound 39, hyperoside (42), quercetin hexuronic acid (45) and ellagic acid (46). The main compounds detected in woods correspond to catechin (13) and quercetin rhamnoside (52). Galloyl proanthocyanidin dimers (23) and proanthocyanidin dimers (11, 17) are more abundant in this organ in particular. Galloyl catechin (34) and the galloyl proanthocyanidin dimer (24) are also present in high amounts. Wood extracts are also characterized by the presence of quinic acid (1, 2, 5, 8, 14) and quercetin (39, 41, 42, 45) derivatives. However, the quantities of kaempferol derivatives are very low.

nd

nd

nd

937,09491: [M-H -182(methyl gallate)]; 767,06908: [M-H - 182 - 170(galloyl unit), 300,99830: HHDP unit; 169,01429: galloyl unit

nd

nd

nd

nd

nd

nd

small amount (mass signal inten-

The major compounds identified in shoots are one quinic acid derivative (2) and quercetin rhamnoside (52), also identified as a main compound of woods. Contrary to woods, no proanthoantocyanidin related compounds were observed but numerous galloyl tannins were detected (6, 12, 20, 26, 31, 32, 38). Some HHDP digalloyl hexosides have already been described as tellimagrandine I in some rose extracts (Hashidoko, 1996) but no study concerning their presence in shoots has been reported. Quinic acid derivatives (1, 3, 5, 8, 14) were also present as well as quercetin derivatives (39, 41, 42, 44, 45). Only two kaempferol derivatives were detected in high amounts (51, 57). Quercetin rhamnoside (52) was identified as a major polyphenol constituent in leaves. This compound is present in the three vegetative plant organs. Quinic acid related molecules are found in large amounts while galloyl tannins are less concentrated except for compounds 12 and 20. Leaves are rich in catechin derivatives

L. Riffault et al. / Phytochemistry 99 (2014) 127–134

(13, 24, 34) and other proanthocyanidins are detected in lesser quantities (17, 23). Quercetin derivatives can be found in high amounts (39, 41, 42, 44, 45) and only two kaempferol derivatives are highly represented (51, 57), as in the shoots. The early buds are characterized by a huge amount of compound 3, an intermediate amount of quinic acid derivatives and numerous galloyl tannins (12, 20, 25, 26, 31, 32, 38). The presence of quercetin based flavonoids was also detected (39, 42, 44, 45, 52) and kaempferol derivatives (50, 52, 53, 57, 58) were observed in this organ, in a higher proportion than the previously described ones. The composition of buds before flowering is relatively similar to the early stage but another major constituent is present: kaempferol glucoside (53). An increase in the abundance of kaempferol derivative was observed. A non-phenolic constituent, tri-p-coumaroyl-spermidine (60), seems to appear during this bud stage. This compound has been previously described as a representative molecule of pollen and it has been surmised that it plays a protective role against environmental stress (Walters, 2003). It has already been detected in different Rosaceae species (Strack et al., 1990). In flowers, the major compound identified was kaempferol glucoside (53) and the second most abundant constituent was also a kaempferol derivative: afzelin (57). The flower extract appears to be the richest in kaempferol based flavonoids, which is in good agreement with the HPTLC observations. The pollen specific compound tri-p-coumaroyl-spermidin (60) was also detected in significant proportions in this organ. Conclusion An initial rapid HPTLC analysis of the hydro-alcoholic rose extracts highlighted various molecular families. Two groups of polyphenols were detected and identified: tannins and flavonoids. An appropriate HPLC method was then developed to investigate this family in greater depth. 41 major chromatographic peaks were distinguished and thanks to high resolution mass spectrometry, a tentative identification of more than 60 compounds present in the different organs can be proposed. The hydrolysable tannins detected are mainly composed of gallic and ellagic acid derivatives, whereas flavonoids are mostly quercetin and kaempferol glycoside derivatives. Some compounds seem to be present in all the organs while others are more specific since they were detected in high amounts only in certain organs. This highlights the specificities of each plant sample, guiding their valuation as potential new cosmetic ingredients. Analyses show that condensed tannins and catechin derivatives are more abundant in woods and leaves, whereas kaempferol derivatives are more abundant in buds and flowers. This study has characterized for the first time numerous polyphenolic compounds of an original rose cultivar: ‘Jardin de Granville’. It would be of a great interest to enhance our knowledge of the phytochemistry of this variety by conducting further analyses focused on less polar families of compounds, which remain poorly investigated. Experimental Plant material The rose cultivar ‘Jardin de Granville’ was grown at Pithiviersle-Vieil (Loiret, France), by the company ‘Les Roses Anciennes, André EVE’. After a meticulous selection process, during which the new varieties are exposed to environmental stress and naturally occurring diseases, only the most resistant cultivars are conserved for potential future commercialization. ‘Jardin de Granville’ passed these tests and is considered as a very vigorous, floribundant and disease resistant cultivar by its breeders. Different parts of the

133

plant, shown in Fig. 1, were collected by hand during the year 2012: woods (w) in February, shoots at the stage of 3 young leaves (sh) in April, early buds (eb) (without stem) measuring less than 2 cm and buds just before flowering (bbf) (without stem) measuring at least 3 cm in May, flowers not fully open (fl) in June and leaves collected halfway up the stems (le) in June. All the plants were grown on the same soil, in a 400 m2 field containing 400 plants which is also used as a plant source for the industrial production of Christian Dior Perfumes. A random sampling on the entire field was carried out and a pool of the different plant samples was collected to accumulate sufficient quantities for the extractions. Concerning transport and storage, woods did not require any specific precautions and were air dried in the laboratory. They were then reduced to shavings by a Retsch (Eragny sur Oise, France) cutting mill, followed by crushing with a basic electric grinder to obtain powder. Buds, flowers and leaves were brought back to the laboratory in boxes containing ice to preserve their integrity whereas shoots were transported in boxes with dry ice ( 80 °C) because of their fragility. The raw material was then stored at 20 °C until extraction. Flowers and leaves were lyophilized before being powdered by the basic grinder. Buds and shoots, which do not withstand defrosting, were ground with the addition of liquid nitrogen in a mortar. The powder was then immediately stored at 20 °C and used for extraction within a few days. Standards and reagents All solvents: ethanol (EtOH), methanol (MeOH) and acetonitrile (ACN) were of analytical grade and were provided by SDS Carlo Erba (Val de Reuil, France). Water was purified (resistance < 18 MX) by an Elgastat UHQ II system (Elga, Antony, France). Formic acid (HCOOH) was provided by Sigma–Aldrich (Saint Quentin Fallavier, France). Reference compounds (gallic acid, ellagic acid, epicatechin gallate, catechin, kaempferol, kaempferol-3-O-glucoside, kaempferol-3-O-rutinoside, afzelin (kaempferol-3-O-pentoside), quercetin, quercetin-3-O-glucoside, hyperoside (quercetin-3-O-galactoside) and rutin) were purchased from Extrasynthese (Genay, France). Tiliroside (kaempferol-3-O-coumaroyl glucoside) was provided by LVMH collaborators and was purified from ‘Jardin de Granville’ flower extract. Microwave extraction The device used was a Milestone MicroSYNTH microwave oven (Sorisole, Italy) monitored with the ‘easyCONTROL’ software. A carrousel suited for microwave extraction, containing twelve reactors of 100 mL was used to manage the extractions. For extractions, 1.5 g of raw material was introduced in each reactor with 60 mL of EtOH/H2O (90/10, v/v). Three cycles of 30 s each were performed at an irradiation power of 1000 W. To limit temperature increase and molecular degradation, a period of cooling, by putting the reactors in ice, was necessary between each cycle. The extracts were then centrifuged and the supernatants were evaporated at 45 °C using a rotary evaporator (Buchi, Labortechnik AG, Switzerland). After notation of the extraction yields, calculated as the ratio between the weight of dried extract and the initial plant weight, solutions at 10 mg mL 1 in methanol were prepared from the dried crude extracts and filtered at 0.45 lm before chromatographic analyses. Apparatus and chromatographic conditions HPTLC analysis The methanol solutions were applied on 10  20 cm HPTLC plates (Merck, Germany) by a CAMAG (Muttenz, Switzerland)

134

L. Riffault et al. / Phytochemistry 99 (2014) 127–134

ATS4 Automatic TLC Sampler controlled by Win-CATS software and were observed thanks to a Reprostar 3 illumination unit. 10 lL of extract solutions at 1 mg mL 1 and 4 lL of standard solutions at 1 mg mL 1 were laid down. Plate elution was performed in a horizontal developing chamber. After development, the plate was dried and observed under visible and UV light (254 and 366 nm). The plate was then sprayed with different specific or non-specific reagents. For the polyphenol system, 10  20 cm W RP18 HPTLC plate was used and the elution mixture was composed of ACN/H2O/ HCOOH (50/50/5). Neu reagent (1 g of diphenyl boric acid ethylamino ether in 100 mL of MeOH) was sprayed, the plate was dried, and immediately afterwards PEG (polyethylene glycol (PEG) 4000 at 5% in EtOH) was applied on the plate. After drying, visualization of the spots was done at 366 nm. HPLC-DAD analysis Molecular content was analyzed using a LaChrom HPLC-DAD instrument (VWR, Fontenay-sous-Bois, France) controlled by EZChrom Elite workstation software. The DAD was set from 200 to 600 nm to record absorbance spectra. Chromatograms were visualized at 270 nm. The acquisition system used was the Ezchrom software, version 3-2-1. The column used to separate the constituents was a C18 Nucleodur sphinx (Macherey-Nagel, Hoerdt, France), 150  4.6 mm, with a particle size of 5 lm fitted with a C18 security guard cartridge system. The mobile phase was made up of 0.1% formic acid in water (phase A), and 0.1% formic acid in methanol (phase B). A solvent gradient was applied as follows: 0–17 min: 5–36% B, 17–25 min: 36% B, 25–35 min: 36–50% B, 35–45 min, 50–70% B, 45–50 min: 70–90% B, 50–60 min, 90% B and finally 60–60.1 min 5% B, maintained during 10 min before each new injection. The column was introduced in an oven Jetstream and heated at 25 °C. The injection volume was 20 lL. HPLC-ESI-Q-TOF-HRMS analysis Separations were performed using an UltiMate 3000 RSLC system equipped with a binary pump, an autosampler and a thermostated column compartment (Dionex, Germering, Germany). The same chromatographic conditions as HPLC-DAD were used. MS experiments were carried out on a maXis UHR-Q-TOF mass spectrometer (Bruker, Bremen, Germany) in negative electrospray ionization mode. Capillary voltage was set at 4.5 kV. The flows of nebulizing and drying gas (nitrogen) were respectively set at 1.2 bar and 8.5 L/min and drying gas was heated at 200 °C. Mass spectra were recorded at 1 Hz in the range of 50–2500 m/z. Chemical formulae were generated using accurate mass measurements and the SmartFormula algorithm from DataAnalysis 4.0 software (Bruker). MS/MS experiments were conducted using the AutoMSMS acquisition mode with two CID collision energies for each compound: 35 and 60 eV for monocharged ions; 17.5 and 30 eV

for dicharged ions. 25 eV of In Source Collision Induced Dissociation (ISCID) were applied to limit the formation of [2M H] ions. Acknowledgments The authors thank ‘Les Roses Anciennes André Eve’ for their advice during collection of the plants and Cyril Colas for his help concerning the use of the HPLC-HRMS device. References Bhandari, P., Kumar, N., Gupta, A.P., Singh, B., Kaul, V.K., 2007. A rapid RP-HPTLC densitometry method for simultaneous determination of major flavonoids in important medicinal plants. J. Sep. Sci. 30 (13), 2092–2096. Cai, Y.-Z., Xing, J., Sun, M., Zhan, Z.-Q., Corke, H., 2005. Phenolic antioxidants (hydrolyzable tannins, flavonols, and anthocyanins) identified by LC–ESI-MS and MALDI-QIT-TOF MS from Rosa chinensis flowers. J. Agric. Food Chem. 53 (26), 9940–9948. Cairns, T., Young, M., Adams, J., Edberg, B., 2000. Modern Roses XI: The World Encyclopedia of Roses. Academic Press, pp. 11–12. Ghazghazi, H., Miguel, M.G., Hasnaoui, B., Sebei, H., Figueiredo, A.C., Pedro, L.G., Barroso, J.G., 2012. Leaf essential oil, leaf methanolic extract and rose hips carotenoids from Rosa sempervirens L. growing in North of Tunisia and their antioxidant activities. J. Med. Plants Res. 6 (4), 574–579. Grossi, C., Raymond, O., Jay, M., 1998. Flavonoid and enzyme polymorphisms and taxonomic organisation of Rosa sections: Carolinae, Cinnamomeae, Pimpinellifoliae and Synstylae. Biochem. Syst. Ecol. 26 (8), 857–871. Hanhineva, K., Rogachev, I., Kokko, H., Mintz-Oron, S., Venger, I., Kärenlampi, S., Aharoni, A., 2008. Non-targeted analysis of spatial metabolite composition in strawberry (Fragaria x ananassa) flowers. Phytochemistry 69 (13), 2463–2481. Hashidoko, Y., 1996. The phytochemistry of Rosa rugosa. Phytochemistry 43 (3), 535–549. Khanbabaee, K., Van Ree, T., 2001. Tannins: classification and definition. Nat. Prod. Rep. 18 (6), 641–649. Kumar, N., Bhandari, P., Singh, B., Gupta, A.P., Kaul, V.K., 2008. Reversed phase-HPLC for rapid determination of polyphenols in flowers of rose species. J. Sep. Sci. 31 (2), 262–267. Kumar, N., Bhandari, P., Singh, B., Bari, S.S., 2009. Antioxidant activity and ultraperformance LC–electrospray ionization-quadrupole time-of-flight mass spectrometry for phenolics-based fingerprinting of Rose species: Rosa damascena, Rosa bourboniana and Rosa brunonii. Food Chem. Toxicol. 47 (2), 361–367. March, R., Brodbelt, J., 2008. Analysis of flavonoids: tandem mass spectrometry, computational methods, and NMR. J. Mass Spectrom. 43 (12), 1581–1617. Mikanagi, Y., Saito, N., Yokoi, M., Tatsuzawa, F., 2000. Anthocyanins in flowers of genus Rosa, sections Cinnamomeae (=Rosa), Chinenses, Gallicanae and some modern garden roses. Biochem. Syst. Ecol. 28 (9), 887–902. Nowak, R., Gawlik-Dziki, U., 2007. Polyphenols of Rosa L. leaves extracts and their radical scavenging activity. Z. Naturforsch. C 6, 32–38. Panda, H., 2006. Cultivation and Utilization of Aromatic Plants. Asia Pacific Business Press, pp. 14–14. Schmitzer, V., Veberic, R., Osterc, G., Stampar, F., 2010. Color and phenolic content changes during flower development in groundcover rose. J. Am. Soc. Hort. Sci. 135 (3), 195–202. Schmitzer, V., Veberic, R., Stampar, F., 2012. Prohexadione-Ca application modifies flavonoid composition and color characteristics of rose (Rosa hybrida L.) flowers. Sci. Hort. 146, 14–20. Strack, D., Eilert, U., Wray, V., Wolff, J., Jaggy, H., 1990. Tricoumaroylspermidine in flowers of rosaceae. Phytochemistry 29 (9), 2893–2896. Steinmann, D., Ganzera, M., 2011. Recent advances on HPLC/MS in medicinal plant analysis. J. Pharm. Biomed. Anal. 55 (4), 744–757. Walters, D.R., 2003. Polyamines and plant disease. Phytochemistry 64 (1), 97–107.

Phytochemical analysis of Rosa hybrida cv. 'Jardin de Granville' by HPTLC, HPLC-DAD and HPLC-ESI-HRMS: polyphenolic fingerprints of six plant organs.

The Rosa hybrida cultivar 'Jardin de Granville', a delicate clear pink flower, is here investigated through a progressive analytical strategy using co...
1MB Sizes 0 Downloads 0 Views