Upload
others
View
10
Download
0
Embed Size (px)
Citation preview
Phytochemical Characterization of
Stevia rebaudiana by
Hande Karaköse
A thesis submitted in partial fulfillment
of the requirements for the degree of
Doctor of Philosophy in Chemistry
Approved Dissertation Committee
Prof. Dr. Nikolai Kuhnert (supervisor)
Professor of Organic and Analytical Chemistry, Jacobs University Bremen
Prof. Dr. Gerd-Volker Röschenthaler
Professor of Chemistry, Jacobs University Bremen
Dr. Adam Le Gresley
Doctor of Organic Chemistry, Kingston University London
Date of Defense: December 21, 2012
School of Engineering and Science
Declaration of Authorship
I, Hande Karaköse, hereby declare that the thesis I am submitting is entirely my own
original work unless where clearly indicated otherwise. I have used only the sources, the
data and the support that I have clearly mentioned. This PhD thesis has not been
submitted for conferral of degree elsewhere.
Bremen, November 30, 2012
Signature
Table of Contents
Acknowledgments.......................................................................................................... i
Abbreviations ................................................................................................................ ii
List of Figures .............................................................................................................. iv
List of Tables .............................................................................................................. vii
Abstract ...................................................................................................................... viii
Chapter Page
I. INTRODUCTION ......................................................................................................1
II. REVIEW OF LITERATURE....................................................................................4
2.1. Steviol Glycosides in Stevia rebaudiana ..........................................................4
2.2. Pharmacology, Toxicology and Regulations ....................................................5
2.3. Pharmacokinetics of Stevioside: Absorption, Metabolism, Excretion .............9
2.4. Biosynthesis of the Steviol Glycosides ...........................................................10
2.5. Analysis of Steviol Glycosides of Stevia rebaudiana.....................................14
2.6. Phenolic Acids ................................................................................................16
2.7. Proteomics of Stevia rebaudiana ....................................................................21
2.8. Lipid Analysis .................................................................................................25
III. RESEARCH OBJECTIVE ....................................................................................30
Chapter Page
IV. Steviol Glycosides Analysis by LC-MS ................................................................31
4.1. Overview .........................................................................................................31
4.2. Materials & Methods ......................................................................................31
4.2.1. Extraction method ................................................................................31
4.2.2. LC-MS analysis of steviol glycosides ...................................................31
4.2.3. HPLC conditions ...................................................................................32
4.2.4. Calibration curve of steviol glycoside standards ..................................33
4.2.5. Method Validation ................................................................................33
4.2.6. Solid phase extraction (SPE) of steviol glycosides ...............................33
4.3. Results & Discussion ......................................................................................34
4.3.1. Identification of steviol glycosides .......................................................36
4.3.2. Method Validation ................................................................................39
4.3.3. Comparison to SPE sample clean up ....................................................40
4.3.4. Quantification of steviol glycosides ......................................................42
4.4. Conclusion ..................................................................................................... 44
V. Polyphenols in Stevia rebaudiana .........................................................................45
5.1. Overview .........................................................................................................45
5.2. Materials & Methods ......................................................................................45
5.2.1. Sample preparation ...............................................................................45
5.2.2. LC-MS analysis of polyphenols ............................................................45
5.2.3. Calibration curve of standard compounds ............................................46
5.2.4. Hydrolysis of flavonoid glycosides ......................................................46
5.2.5. Statistical analysis .................................................................................46
5.3. Results & Discussion ......................................................................................47
5.3.1. Characterization of chlorogenic acids ...................................................50
5.3.2. Characterization of flavonoid glycosides ..............................................54
5.3.3. Quantification of chlorogenic acids and flavonoid glycosides .............56
3.3.1. Sample variation ........................................................................57
3.3.2. Flavonoid quantification ............................................................65
3.3.3. Principal component analysis (PCA) .........................................67
5.3.4. Statistical evaluation of quantification data of polyphenols in stevia ...69
3.4.1. Statistical spread of data .............................................................69
3.4.2. Correlations .................................................................................70
3.4.3. Analysis of variance (ANOVA) ..................................................76
5.4. Conclusion. .....................................................................................................80
Chapter Page
VI. Lipid Analysis of Stevia ........................................................................................81
6.1. Overview .........................................................................................................81
6.2. Materials & Methods ......................................................................................81
6.2.1. Extraction method ................................................................................81
6.2.2. Methyl ester formation ..........................................................................81
6.2.3. GC-FID conditions................................................................................81
6.2.4. GC-MS conditions ................................................................................82
6.2.5. Calibration curve of FAME ..................................................................82
6.2.6. MALDI-TOF MS ..................................................................................82
6.3. Results & Discussion ......................................................................................83
6.4. Conclusion ......................................................................................................93
VII. Proteomics of Stevia .............................................................................................94
7.1. Overview .........................................................................................................94
7.2. Materials & Methods ......................................................................................94
7.2.1. Extraction of proteins ............................................................................94
7.2.2. Protein analysis .....................................................................................95
7.2.3. MALDI-TOF MS conditions ................................................................97
7.3. Results & Discussion ......................................................................................98
7.3.1. SDS results ............................................................................................98
7.3.2. 2D-SDS .................................................................................................99
7.3.3. MALDI-TOF MS results ....................................................................100
7.4. Conclusion ....................................................................................................105
VIII. Summary ...........................................................................................................106
IX. References ............................................................................................................107
APPENDIX ................................................................................................................113
Publications ...............................................................................................................149
Curriculum Vitae
i
ACKNOWLEDGMENTS
This research has been completed with the support of a large number of people. I would like to
express my gratitude to them.
First of all, sincere thanks to my supervisor Prof.Nikolai Kuhnert for his guidance and expert
advices during my study.
I wish to acknowledge the support of European Union (Project DIVAS) and Jacobs University
Bremen for the full scholarship and funding of the research. I am grateful also to Dr. Kienle at
the University of Hohenheim for the valuable discussions.
A special thanks to my friends, Agnieszka Golon and Rohan Shah for their assistance during the
experimental work of the study. Also, thanks to my colleagues at Jacobs University for providing
a pleasant working atmosphere and Anja Müller for the technical assistance.
Finally, my warmest thanks belong to my parents Dilek and Nejdet and my brother Çağatay, for
their confidence in me and for being always supportive and interested in my work.
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse ii
Jacobs University Bremen
ABBREVIATIONS
HPLC high performance liquid chromatography
MS mass spectrometry
LC-MS liquid chromatography coupled with mass spectrometry
GC-MS gas chromatography coupled with mass spectrometry
GC-FID gas chromatography coupled with flame ionization detector
EI-MS electron impact ionization mass spectrometry
TIC total ion chromatogram
EIC extracted ion chromatogram
BPC base peak chromatogram
MS2/MS
3 tandem mass
ESI-MS electrospray ionization mass spectrometry
m/z mass-charge ratio
UV ultra-violet
MALDI matrix assisted laser ionization
TOF-MS time of flight mass spectrometry
HR-MS high resolution mass spectrometry
HILIC hydrophilic interaction chromatography
CGAs chlorogenic acids
CQA caffeoylquinic acid
FQA feruloylquinic acid
MeOH methanol
ACN acetonitrile
DXS deoxyxylose-5-phosphate synthase
CDPS copalyl diphosphate synthase
KS kaurene synthase
KO kaurene oxidase
KAH kaurenoic acid hydroxylase
UGTs UDP-glycosyltransferases
RT Retention time
RebA rebaudioside A
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse iii
Jacobs University Bremen
RebC rebaudioside C
PCA principal component analysis
ANOVA analysis of variance
TCA trichloroacetic acid
IEF isoelectric focusing
HCCA α-cyano-4-hydroxycinnamic acid
FAME fatty acid methyl esters
FTICR Fourier transform ion cyclotron resonance
APCI atomic pressure chemical ionization
SPE solid phase extraction
S/N signal to noise ratio
LOD limit of detection
LOQ limit of quantification
RSD % relative standard deviation %
K7g kaempferol-7-O-glycoside
Q3g quercetin-3-O-glycoside
KS kolmogorov-smirnov test
DTT dithiothreitol
TFA trifluoroacetic acid
MVA mevalonic acid pathway
MEP 2-C-methyl-D-erythritol 4-phosphate/1-deoxy-D-xylulose 5-phosphate pathway
APS ammonium persulfate
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse iv
Jacobs University Bremen
LIST OF FIGURES
Figure 1.General structure of steviol glycosides and representative structure of rebaudioside A
Figure 2.Structures of few artificial sweeteners
Figure 3.Structures of steviol metabolites
Figure 4.Hypothetical excretion route of stevioside
Figure 5.Steviol glycoside biosynthesis via the MEP pathway
Figure 6.Alternative MVA pathway
Figure 7.Examples for hydroxybenzoic and hydroxycinnamic acids
Figure 8.General structure of quinic acid and one of chlorogenic acids as an example
Figure 9.Generic structure of major classes of flavonoids
Figure 10.Strategies for MS based protein identification
Figure 11.Peptide fragmentation nomenclature
Figure 12.Examples of lipid categories
Figure 13.Total ion chromatogram in negative ion mode using C18 column of methanolic Stevia
rebaudiana extract showing phenolics (chlorogenic acids, flavonoids) and steviol glycosides
Figure 14.Base peak chromatogram of steviol glycosides obtained using HILIC column
Figure 15.Mechanism of fragmentation in tandem MS spectra of rebaudioside A and
rebaudioside E illustrating how isomeric compounds can be distinguished by tandem MS
Figure 16.Tandem MS spectra of rebaudioside A (above) and rebaudioside E (below) in
negative ion mode
Figure 17.Tandem MS spectra of rebaudioside D in negative ion mode
Figure 18.Total ion chromatograms for comparison of different amounts of material I in SPE
cleanup procedure
Figure 19.Total ion chromatograms for comparison of SPE cleanup of the stevia extract with
materials I and II cartridges
Figure 20.Radar plot of steviol glycoside concentrations varying between seven varieties
(average values taken within +/- 3σ) and in comparison to non-EU samples
Figure 21.Radar plot of steviol glycoside concentrations varying between all origins (average
values taken within +/- 3σ) and in comparison to non-EU samples
Figure 22.Base peak chromatogram in negative ion mode using C18 column of methanolic
Stevia rebaudiana extract showing chlorogonic acids, flavonoids and steviol glycosides
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse v
Jacobs University Bremen
Figure 23.Structures of caffeoylquinic acids and flavonoid glycosides
Figure 24.Extracted ion chromatogram of m/z 353 of three mono-caffeoylquinic acids 3-
caffeoylquinic acid, 5-caffeoylquinic acid, and 4-caffeoylquinic acid (from left to right) in
negative ion mode
Figure 25.Consecutively MS, MS2
and MS3
spectra of 3-caffeoylquinic acid in negative ion
mode
Figure 26.Consecutively MS, MS2
and MS3
spectra of 4-caffeoylquinic acid in negative ion
mode
Figure 27.Consecutively MS, MS2, MS
3 and MS
4spectra of 3,5-dicaffeoylquinic acid in negative
ion mode
Figure 28.Consecutively MS, MS2, MS
3 and MS
4spectra of 4,5-dicaffeoylquinic acid in negative
ion mode
Figure 29.Chemical structure of four flavonoid aglycones identified in Stevia rebaudiana leaves
Figure 30.Extracted ion chromatogram of m/z 447.0 in negative ion mode
Figure 31.An example of tandem MS spectra for compound 1, revealing its identity as
kaempferol glucopyranoside
Figure 32.Fragmentation illustration on luteolin-7-glucoside
Figure 33.Radar plot of individual chlorogenic acid concentrations varying between seven
varieties (average values taken within +/- 3σ) and in comparison to non-EU samples
Figure 34.Radar plot of mono- and di-acyl quinic acids concentrations varying between seven
varieties (average values taken within +/- 3σ) and in comparison to non-EU samples
Figure 35.Bar plot of total mono- and di-acyl quinic acids concentrations varying between seven
varieties (average values taken within +/- 3σ) and in comparison to non-EU samples
Figure 36.Map showing the origins of stevia cultivation within the project
Figure 37.Radar plot of mono- and di-acyl quinic acids concentrations varying between all
origins (average values taken within +/- 3σ) and in comparison to non-EU samples
Figure 38.Bar plot of total mono- and di-acyl quinic acids concentrations varying between all
origins (average values taken within +/- 3σ) and in comparison to non-EU samples
Figure 39a.PCA analysis of phenol profile of 35 stevia leaf LC-MS datasets
Figure 39b.PCA analysis of phenol profile of 40 stevia leaf LC-MS datasets
Figure 40.Histogram of 5-CQA, showing the normal distribution of the dataset
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse vi
Jacobs University Bremen
Figure 41.Graph showing the correlation between a) 3,5-diCQA/4,5-diCQA b) 3-CQA/5-CQA
c) 5-CQA/4,5-diCQA and d) 4-CQA/3,5-diCQA
Figure 42.Linear dependency of cis-5-CQA with 5-CQA and two isomers of cis-4,5-diCQAs
Figure 43.Amount of 5-CQA in mg/100g dry leaves from three harvests from location A (TCV)
and location B (Amfilikeia) during 2011
Figure 44.log of trans/cis-5-CQA concentrations against the number of sunshine hours in the
month for a total of ten harvests from six locations
Figure 45.GC-MS chromatogram of total lipid extracts from Stevia rebaudiana leaves from
sample (Uconor, Var.4)
Figure 46.GC-MS chromatogram of FAME standard mixture
Figure 47.Representative EI-MS spectra obtained from GC-MS measurement of stevia extract
Figure 48.Structures of fatty acids in Stevia rebaudiana extract
Figure 49.Fatty acid profile of average stevia leaf in % X:Y denominates the number of carbon
atoms in the fatty acid (X) and the number of double bonds in the fatty acid (Y)
Figure 50.MALDI-MS spectrum of total lipid extract in positive ion mode using 2,5-DHB as a
matrix
Figure 51.Chemical structures of terpenes identified in Stevia rebaudiana leaves
Figure 52.GC chromatogram of methylesterified steviol and stevia extract
Figure 53.Extraction procedure of proteins
Figure 54.SDS gel for sample number 8 TCV harvest I, loaded on gel at different concentrations
1 mg/mL and 0.5 mg/mL
Figure 55.2D-SDS separation of stevia total protein extract. 7cm strip of pH 4-7, where spot 1
and 2 are at 55 kDa, and spot 5 at 15kDa
Figure 56.MALDI-TOF MS spectra and mass list of trypsin digested 2D-SDS spot and Mascot
search result showing the sequence information for RuBisCO enzyme with the score of 45%
Figure 57.Mascot search result of MALDI spectra
Figure 58.MS/MS de novo sequencing of the m/z 1230. Series of y and b fragments are labeled
Figure 59.Structure and fragmentation of m/z 1230 based on denovo sequencing
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse vii
Jacobs University Bremen
LIST of TABLES
Table 1.High resolution mass spectrometrical data of stevia terpene glycosides in negative ion
mode from LC-TOF MS analysis
Table 2.Enzymes involved in biosynthesis of steviol glycosides
Table 3.Steviol glycosides values from 166 samples
Table 4.Chromatographic and MS data on flavonoid glycosides and CGAs present in stevia
Table 5.Average values (taken within +/- 3σ) for chlorogenic acids in seven different varieties
Table 6.Average values (taken within +/- 3σ) for chlorogenic acids between origins
Table 7.Average values (taken within +/- 3σ) for chlorogenic acids between harvests
Table 8.Comparison of average values (taken within +/- 3 σ) for chlorogenic acids between three
harvests of same variety and origin
Table 9.Flavonoid glycosides average values for two major flavonoids in samples between
origins determined by LC-MS directly from extracts without hydrolysis
Table 10.Flavonoid glycosides average values between varieties
Table 11.Values for flavonoids quercetin, kaempferol, luteolin and apigenin determined after
hydrolysis of total polyphenol fractions using HCl/MeOH, determined by LC-MS
Table 12.Descriptive statistics of caffeoylquinic acids
Table 13.Correlation coefficients of mono and di-CQAs
Table 14.Correlation coefficients of cis isomers according to Spearman’s rule
Table 15.Results of test of homogeinity of variances
Table 16.ANOVA results for effect of origin on stevia CGA content
Table 17.Test of homogeneity of variances
Table 18.ANOVA results for effect of variety on stevia CGA content
Table 19.Total lipid values in weight % from 46 samples
Table 20.Quantities of polyunsaturated fatty acids
Table 21.Retention time and NIST scores of some terpenes identified in stevia extract
Table 22.Amount and properties of chosen stevia leaves for protein extraction
Table 23.Preparation of separation and stacking gel for 2D SDS PAGE
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse viii
Jacobs University Bremen
ABSTRACT
Stevia rebaudiana (Bertoni) is from the Asteraceae family of plants with significant economic
value due to its high content of natural zero calorie steviol glycoside sweeteners in its leaves.
The leaves contain ent-kaurene glycosides, comprising stevioside, rebaudioside A, B, C, D, E, F
and dulcoside A. Rebaudioside A and stevioside are the most abundant diterpene glycosides
(steviol glycosides) in the leaves.
The phytochemical characterization of stevia leaves is playing an important role in a future EU
consumption of stevia as a novel food. For this purpose, the chemical composition of stevia
(phenols and steviol glycosides detailed, with lipids and proteins in representative cases) was
studied and methods have been developed for quantitative and qualitative analysis.
Stevia leaves cultivated in more than ten locations inside and outside of Europe with seven
different varieties, corresponding to total of 166 stevia samples, were extracted and their
chemical composition was profiled and quantified by LC-MS for steviol glycosides and
polyphenols (chlorogenic acids and flavonoids). Profiling, identification and quantification of
terpenoids and lipids were achieved by using GC, GC-MS and MALDI-TOF techniques. In
addition, protein extraction and analysis was carried out to identify potentially allergenic proteins
in stevia leaves. Protein separation and isolation was achieved with 2-dimensional
electrophoresis (2DE) and MALDI-TOF MS analysis was performed for the identification of
individual proteins.
Furthermore, as stevia may cultivated within various regions of the EU with different soil and
climatic conditions it is important to know whether an EU-common specification will be
achieved and how stevia leaves from regions outside EU can be distinguished on a scientific
basis. For this purpose, principal component analysis (PCA) was performed based on the LC-MS
dataset of stevia phenols. In addition, effect of growth origin and variety on stevia secondary
metabolite profile was analyzed statistically by ANOVA (analysis of variance).
1
1. INTRODUCTION
Stevia rebaudiana (stevia or S.rebaudiana) is native to Paraguay and belongs to the Asteraceae
family of plants. Stevia and its sweet taste was botanically described by M.S. Bertoni in 18991.
The high content of natural, zero-calorie sweeteners in its leaves makes stevia of a significant
economic value in the food industry in many applications as a sweetener. Interest in stevia
products has dramatically increased recently due to its approval by European and US legislating
authorities. Stevia is likely to become a major source of high-potency sweetener for the growing
natural food market in the future.
The majority of the annual stevia production of an estimated 4000 t is produced in China and
South America. The stevia crop has been shown to be highly adaptable to cultivation in many
other parts of the world. S. rebaudiana occurs naturally on acid soils of pH 4 – 5, but will also
grow on soils with pH levels of 6.5 – 7.5 making it an interesting alternative to plants cultivated
on poor soils such as tobacco2.
Stevia contains ent-kaurene glycosides, comprising stevioside, rebaudioside A, B, C, D, E, F and
dulcoside A (Figure 1, Table 1), which give the leaves its characteristic taste of 200-400 times
sweeter than sucrose. Stevioside has a sweetening power 300 times that of sucrose, and
rebaudioside A is 400 times sweeter than sucrose3. Rebaudioside A and stevioside are the most
abundant compounds; steviolbioside and rebaudioside B are believed to be formed by partial
hydrolysis during the extraction process4. The rest of the steviol glycosides (e.g. dulcoside A,
rebaudioside C) are at trace levels. In addition to being a natural sweetener, steviol glycosides
have functional and sensory properties superior to those of many other high-potency sweeteners.
Stevia leaves can be used in their natural state (fresh or dried form), due to its high sweetening
intensity. Only small quantities are needed for comparison with white sugar. It does not increase
the blood sugar level therefore; it can be used by diabetics without adverse glycemic responses.
The human fecal microflora hydrolyzes stevioside and rebaudioside A to their common aglycon
steviol in 10 and 24 h, respectively but steviol is not degraded by the human body 5.
In addition to diterpene glycosides, a number of secondary plant metabolites have been identified
from S. rebaudiana including labdane-type diterpenes, triterpenoids and steroids, phenolic acids
(flavonoid glycosides and chlorogenic acids), and oil components. From S. rebaudiana, ten
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 2
Jacobs University Bremen
labdane-type diterpenoids were identified, including austroinulin,isoaustroinulin6, sterebins (A -
H)7, 8
. A triterpenoid, lupeol 3-palmitate, was also separated from stevia9. As plant sterols, β-
sitosterol, stigmasterol and campesterol were identified from S. rebaudiana10
.
The presence of chlorogenic acids (CGAs) and flavonoid glycosides in stevia leaves gives the
plant additional health benefits, and it could as well affect its organoleptic properties.
CGAs are a large family of esters formed between quinic acid and certain trans -
hydroxycinnamic acids, most commonly caffeic, p-coumaric, and ferulic acid. Similar to
chlorogenic acids the presence of flavonoid compounds adds a health benefit to the usage of
stevia leaves in food products. Flavonoids are a class of secondary metabolites that are produced
ubiquitously in fruits and vegetables. By definition flavonoids are compounds with a C6-C3-C6
structure comprising two aromatic ring, one fused as a benzopyran.
The secondary metabolites of interest in the present study were; steviol glycosides, chlorogenic
acids, flavonoid glycosides, lipids, volatile terpenes and proteins. The main objective of this
project was to provide a scientific basis for a future EU specification for stevia. The steviol
glycoside and polyphenol profile and quantities of stevia samples cultivated in different
European and non-European countries with seven different botanical varieties, harvested at three
different times were obtained by analyzing stevia leaf extracts using a HPLC-TOF MS system.
The identification of the compounds was achieved by analyzing tandem mass spectra and high
resolution mass spectrometry (HR-MS) and for selected unknown phenolic compounds
spectroscopic MS rules previously developed in our laboratory was used to elucidate structures.
Stevia leave proteins were purified separated and sequenced with an aim to identify potentially
allergenic proteins using 2D gel electrophoresis and MALDI-TOF MS technique. Lipids and
volatile terpenes were determined by subjecting non-polar solvent extracts of stevia leaves to
GC-MS and MALDI-TOF MS. Identification of the compounds was achieved using NIST
library and comparison of retention times and GC-MS data of fatty acid methyl ester standard
mixtures.
Statistical analysis (PCA, correlation studies and ANOVA) served for differentiating the non-EU
and EU cultivated stevia samples and for studying the relations between each components and
growth conditions of stevia.
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 3
Jacobs University Bremen
O
OH
HO
HO O
HO
O
CH3
CH3
O
CH2
O
OH
O OH
OO
O
HO
OH OH
OH
OH
OH
HO
COORH3C
CH3
OR1
CH2
General structure ofsteviol glycosides
Rebaudioside A
O O
CH3
CH3
O
CH2
O
OH
O OH
OHO
HO
OH OH
HO
Stevioside
OH
O
OH
HO
HO
HO
Figure 1.General structure of steviol glycoside and representative structure of rebaudioside A.
Table 1.High resolution mass spectrometrical data of stevia terpene glycosides in negative ion
mode from LC-TOF MS analysis
Compound R R1 Molecular
Formula
Experimental
m/z (M-H+)
-
Theoretical
m/z (M-H+)
-
Relative Error
(ppm)
Steviol
Steviolbioside
H
H
H
glc2 - 1glc
C20H30O3
C32H50O13
317.0819
641.3181
317.0717
641.3179
9.0
0.4
Rubusoside Glc glc C32H50O13 641.3166 641.3179 2.0
Stevioside Glc glc2 - 1glc C38H60O18 803.3751 803.3707 5.5
Rebaudioside A Glc glc32 -1glc
1glc
C44H70O23 965.425 965.4235 1.6
Rebaudioside B H glc32 -1glc
1glc
C38H60O18 803.368 803.3707 2.8
Rebaudioside C
(Dulcoside B)
Glc glc32 -1rham
1glc
C44H70O22 949.427 949.4286 1.7
Rebaudioside D glc2-1glc glc32 -1rham
1glc
C50H80O28 1127.4726 1127.4763 3.3
Rebaudioside E glc2-1glc glc2-1glc C44H70O23 965.4199 965.4235 3.7
Rebaudioside F Glc glc32 -1xyl
1glc
C43H68O22 935.4097 935.4129 3.5
Dulcoside A Glc glc2 - 1rham C38H60O17 787.3732 787.3758 3.3
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 4
Jacobs University Bremen
2. REVIEW OF LITERATURE
2.1. Steviol Glycosides in Stevia rebaudiana
The most commercially important compounds from leaves of Stevia rebaudiana Bertoni
(S.rebaudiana/stevia) are the sweet tasting ent-kaurene diterpenoid glycosides. Two main
glycosides are stevioside and rebaudioside A. There are other related compounds including
Rebaudioside B-E, Dulcoside A and C which occur as minor components. Summaries of the
compounds from stevia are shown in Table 1.
Diterpene glycosides from S. rebaudiana contain a common aglycone called steviol (13-
hydroxy-ent-kaur-16-en-19-oic acid), and differ only in the glycosidic constituents attached at
C-13 and/or C-19.
Stevioside is the main sweet tasting glycoside in stevia (5-10 %) and was reported to be 250-300
times sweeter than sucrose. Rebaudioside A (2-4%) is the second most abundant ent-kaurene and
sweetest compound in stevia, its sweetness is 400 times more than sucrose. It was reported to
have a more pleasant taste and it is more water soluble than stevioside. Rebaudioside B, D, and E
may be also present in minor quantities; however, it is suspected that rebaudioside B is a
byproduct of the isolation technique11
. The two main compounds stevioside and rebaudioside,
primarily responsible for the sweet taste of stevia leaves, were first isolated by two French
chemists, Bridel and Lavielle (1931)12
.
The diterpene, steviol (Table 1) is the aglycone of stevia glycosides. Diterpene glycosides form
with the formation of ester bond between glucose molecule and carboxyl group of steviol and
replacing of hydroxyl hydrogen with combinations of glucose, rhamnose and xylose.
Stevioside has two linked glucose molecules at the hydroxyl site, whereas rebaudioside A has
three glucoses, with the central glucose of the saccharate connected to the central steviol
structure. Rebaudioside C and Dulcoside A possess a rhamnose sugar, whereas Rebaudioside F
possesses one xylose unit in its structure.
After sensory panel testing, Rebaudioside A was reported to have the least bitterness of all the
steviol glycosides in the stevia plant. Glycosides are molecules that contain glucose and other
non-sugar substances called aglycones. The taste receptor of tongue reacts to the glucose in the
glycosides, thus steviol glycosides with more glucose (e.g. rebaudioside A) taste sweeter than
those with less glucose (e.g.stevioside)13
. The bitter receptors of the tongue react to the
aglycones, or to polyphenols in the case of stevia leaves usage.
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 5
Jacobs University Bremen
2.2. Pharmacology, Toxicology and Regulations
A sugar substitute is a food additive that enhances the effect of sugar in taste. There are natural
and synthetic sugar substitutes. Those that are not natural are, in general, called artificial
sweeteners. Food additives must be approved by the FDA, which publishes a Generally
Recognized as Safe (GRAS) list of additives. The majority of sugar substitutes approved for use
are artificially-synthesized compounds. Sugar substitutes are used for reasons, including weight
loss, dental care, diabetes and hypoglycemia. Sugar substitutes which are commonly used in
foods are; aspartame, cyclamate, saccharin and sucralose. Starting with aspartame, it was
produced from two amino acids: aspartic acid and phenylalanine. It is about 200 times sweeter
than sucrose. The safety of aspartame has been studied extensively including animal studies,
clinical and epidemiological research14
. Hypotheses of adverse health effects have focused on the
three metabolites of aspartame, which are aspartic acid, methanol and phenylalanine and further
breakdown products including formic acid and formaldehyde15
. Aspartame is rapidly hydrolyzed
in the small intestines. Even with ingestion of very high doses of aspartame (over 200 mg/kg), no
aspartame is found in the blood due to the rapid breakdown16
. Furthermore, people with the
genetic disorder phenylketonuria should avoid aspartame since they have a decreased ability to
metabolize naturally occurring essential amino acid phenylalanine. The acceptable daily intake
(ADI) value for aspartame is determined as 40 mg/kg of body weight 17
.
Sucralose is a chlorinated sugar which is 600 times sweeter than sucrose. FDA approved usage
of sucralose after reviewing 110 studies in humans and animals18
. However, some adverse
effects were observed at doses that significantly exceeded the estimated daily intake which is 1.1
mg/kg/day 19
.
Saccharin was produced first in 1878 by a chemist working on coal tar derivatives. Studies in
laboratory rats during the early 1970s linked saccharin with the development of bladder cancer in
rodents. As a consequence, all food containing saccharin was labeled with a warning20
. However,
in 2000, the warning labels were removed because rodents, unlike humans, have a unique
combination of high pH, high calcium phosphate, and high protein levels in their urine, which
leads to formation of microcrystals that damages the bladder and over-produced cells to repair
the damage leads to tumor formation. Since this does not occur in humans, the conclusion was
there is no cancer risk21, 22
. In the European Union, saccharin is also known by the E number
(additive code) E954.
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 6
Jacobs University Bremen
Beginning in the 1970s with saccharin until the present day, artificial sweeteners have generated
a lot of controversy. The chemical nature of many artificial sweeteners (Figure 2) do not present
a good public image; e.g. saccharin was first produced from a coal-tar derivative23
, aspartame
breaks down into formaldehyde upon digestion15
, it also presents a health hazard to people born
with phenylketonuria, and sucralose is manufactured by the selective chlorination of sucrose24
.
With the scientific evidence on a particular sweetener, which is often inconclusive, and with
many interests at stake, including the food additive approval process, and potential political and
economic consequences, the results of these disagreements have not been entirely consistent or
logical. Aspartame, for example, gained FDA approval over vocal opposition from certain public
health advocates, while stevia extract, a substance which arguably presents health risks, cannot
have FDA approval and avoids a complete ban only by classification as a “dietary supplement”
rather than as a food additive25
.
NH
OH
OCH3
O
O
O
NH2
OCl
HOOH
O
HO
O
OHHO
Cl
Cl
SNH
O
OO
aspartame sucralose saccharin
Figure 2. Structures of few artificial sweeteners.
Stevia was used extensively by the Guarani Indians for more than 1,500 years in Paraguay and
Brazil26
. Stevia was first used as a sweetener in Japan in the 1970s, and it was a natural substance
that had been in use before 1958s with no apparent ill effects. However, the FDA banned stevia
as unsafe food additive in 1991 after receiving an anonymous industry complaint, and restricted
its import27
. The stated reason of FDA was that toxicological information on stevia was
inadequate to demonstrate its safety.
Health controversies about stevia started with the study of Pezzuto in 198528
, which reported that
steviol, a breakdown product of stevioside and rebaudioside A is a mutagen in the presence of a
liver extract of pre-treated rats. But this finding was criticized and stated that it might be worth
exploring the possibility that the mutagenicity of steviol (as in the experiments of Pezzuto et
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 7
Jacobs University Bremen
al.,1986)29
is due to an impurity and the very high dose used in the experiments30
. Metabolism
of steviol in rat liver is complex and some metabolites detected are shown in Figure 3. The major
metabolite is 15-α-hydroxysteviol, which is non-mutagenic both in the presence and absence of a
metabolic activating system. Other metabolites are 7-β-hydroxysteviol, 17-hydroxyisosteviol and
ent-16-oxo-17-hydroxybeyeran-19-oic acid29, 31
. The mutagenic substance was proposed to be
15-oxosteviol. But, this compound was not detected as a metabolite of steviol and it was reported
to be bactericidal and weakly mutagenic30
. Nevertheless, other bacterial mutagenic assays failed
to demonstrate steviol mutagenic activity32
. The nature of mutagenic metabolite thus remained in
doubt.
Stevia remained banned until 1994, when forced under the Dietary Supplement Health and
Education Act, the FDA revised the decision on stevia and permitted it to be used as a dieatary
supplement. Over the following years studies on the toxicology and adverse effects of stevia
showed contradictory results.
CH2
R1
R2
CH3HOOC
H3C
OH
CH2
CH3HOOC
H3C
OH
O
R1=OH R2=H; 15α-hydroxysteviol 15-oxosteviol
R1=H R2=OH; 7β-hydroxysteviol
O
CH3HOOC
H3C
OHCH2
O
CH3HOOC
H3C
CH2OH
Steviol-16,17-oxide 17-hydroxyisosteviol
Figure 3.Structures of steviol metabolites.
Toskulkao et. al reported stevioside and steviol to have very low acute oral toxicity in the mouse,
rat and hamster 33
. Xili et al.34
have performed a combined chronic and carcinogenicity study, in
Wistar rats and in this study however stevioside administration in the diet showed no
carcinogenic effects in the rat. Through the review of many other toxicological studies 35-38
on
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 8
Jacobs University Bremen
stevia and its compounds, the EFSA committee concluded in 1999 that steviol, one metabolite of
stevioside, that is produced by the human microflora is genotoxic and induces developmental
toxicity. Therefore, the European Commission in 1999 banned stevia and its products in foods in
the European Union pending further research.
The European Food Safety Authority reevaluated the safety of steviol glycosides and expressed
its opinion on 10 March 2010. The Authority established an Acceptable Daily Intake (ADI) for
steviol glycosides, expressed as steviol equivalents, of 4 mg/kg (BW/day). The European
Commission allowed the usage of steviol glycosides as a food additive, establishing maximum
content levels for different types of foods and beverages on 11 November 201139
.
Regarding the effect of stevia in diabetes, a 2011 study by Misra et. al. on diabetes induced to
rats by injection of alloxan, have shown that leaf extract of S. rebaudiana (200 and 400 mg/kg)
produced a significant decrease in the blood glucose level, without producing condition of
hypoglycemia after treatment 26
. In addition, a 2009 review indicated that stevioside and related
compounds have anti-hyperglycemic, anti-hypertensive, anti-inflammatory, anti-tumor, anti-
diarrheal, diuretic, and immunomodulatory actions40
. The effect of stevioside and steviol on
glucose absorption was investigated by Toskulkao et al. and it was reported that 1mM steviol
inhibits glucose absorbtion, whereas 5 mM has no inhibitory effect. The inhibition of glucose
absorption by steviol was related to steviol concentration and incubation time 41
. However, the
announced acceptable daily intake of steviol glycosides (4 mg/kg BW/day) would yield a
maximum plasma concentration of steviol of approx. 20 μM if stevioside is completely
converted to steviol. This concentration is far below the reported value to inhibit intestinal
glucose absorption. Therefore, more studies should be conducted using ADI amount to
reevaluate the effect of steviol on glucose absorption. However, it is worth pointing out that
stevioside does not interfere with glucose absorption40
.
Stevioside, other related steviol glycosides, or stevia leaves themselves have been used
commercially in many countries, especially in Asia, as food additives for sweetening a variety of
products without any side effects. Moreover, phytochemicals (especially polyphenolics and
steviol glycosides) of stevia were reported to exhibit significant pharmacological activities.
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 9
Jacobs University Bremen
2.3. Pharmacokinetics of Stevioside: Absorbtion, Metabolism, Excretion
Absorption and metabolism studies on steviol glycosides showed that the uptake of stevioside by
the intestinal tract is extremely low due to its high molecular size and hydrophlicity40, 42, 43
.
Stevioside is not degraded by the enzymes of the intestinal tract, gastric juice or digestive
enzymes from animals and humans 5, 42, 44
. However, bacterial intestinal flora of humans is able
to convert stevioside to steviol and Bacteroides sp. is responsible for this conversion in the lower
gastrointestinal tract of both rat and human 5. Koyoma et. al
44. investigated the metabolism of
stevia by incubating stevioside, rebaudioside A and steviol with pooled human faecal
homogenates obtained from healthy volunteers for 0.8 and 24 h under anaerobic conditions.
Stevioside, rebaudioside A were completely hydrolysed in 24 h, and no degradation of steviol
was observed. The author proposed a metabolic pathway for rebaudioside A, which suggests that
majority of rebaudioside A is hydrolyzed via stevioside to steviol and minority via rebaudioside
B to steviol. Steviol was not further metabolized in human intestinal microflora being
inconsistent with the study of Pezzuto et.al28
reporting the oxidation of steviol to hydroxysteviol,
or to 15-oxo-steviol (Figure 3).
Another study in 10 healthy volunteers showed that after 3 days of consumption of stevioside
(every day 3 times 250 mg capsules with 8 h intervals), steviol glucoronide is the only excretion
product of stevioside in urine. There was no detection of free steviol in urine. Moreover, after
enzymatic hydrolysis of urine by β-glucuronidase/sulfatase, steviol was the only aglycone and
there was no indication of steviol sulfates 42
. The excretion route proposed by Geuns et al. is
presented in Figure 4.
O-Glc-Glc
H
HH3C CO2Glc
H3C
OH
H
HH3C CO2H
H3C
OH
H
HH3C C
H3C
O
O O
H
H
CO2H
H
H
HO OH
bacteria
colon
liver
stevioside steviol steviol glucuronide
Figure 4.Hypothetical excretion route of stevioside42
.
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 10
Jacobs University Bremen
Another study in humans reported that 72 h after oral stevioside ingestion, steviol glucuronide
excretion in urine and free steviol in feces accounting for 62% and 5.2% of the total dose of
stevioside administered respectively40, 45
.
As conclusion from the reviewed literature, steviol glucorunonide is the main metabolite of
stevioside consumption and urinary excretion is responsible for the disposal from the body.
2.4. Biosynthesis of the Steviol Glycosides
Biosynthesis of steviol glycosides are still subject of discussion. There are two main proposed
pathways for steviol glycosides; 2-C-methyl-D-erythritol 4-phosphate/1-deoxy-D-xylulose 5-
phosphate pathway (MEP/DOXP pathway) (Figure 5) and mevalonic acid pathway (MVA)
(Figure 6).
MEP pathway of isoprenoid (terpenoids) biosynthesis is a metabolic pathway which leads to the
formation of isopentenyl pyrophosphate (IPP) (9) and dimethylallyl pyrophosphate (DMAPP)
(10) in the plastids of the plants. MEP pathway is an additional alternative pathway to mevalonic
acid pathway (MVA) for formation of isoprenoids (terpenoids). MVA reactions take place in
cytosol whereas MEP reactions occur in plastids. Pyruvate (1) and glyceraldehyde-3-phosphate
(2) are converted by DOXP synthase to 1-deoxy-D-xylulose-5-phosphate (3) and by DOXP
reductase to 2-C-methyl-D-erythritol 4-phosphate (4) (MEP). After subsequent reaction steps,
the end products IPP (9) and DMAPP (10), which are precursors of terpenoids, are formed.
Synthesis of all higher terpenoids occurs via formation of geranyl pyrophosphate (GPP) and
geranylgeranyl pyrophosphate (11) (GGPP).
In the proposed MEP pathway, steviol was synthesized from kaurene (13) 46
. The plant gene for
the first step in the MEP pathway is deoxyxyulose-5-phosphate (DXP) synthase (DXS), which
leads to the synthesis of DXP from pyruvate (1) and glyceraldehyde 3-phosphate (2). Once
synthesized, DXP can either be used for the production of vitamins like thiamin or in the MEP
pathway for isoprenoid synthesis.
The DXS amino acid sequence is highly conserved among plant species, which enabled Totte´ et
al. (2003)47
to design primers for RT-PCR and clone the DXS gene from Stevia. The author
suggested that steviol was synthesized via mevalonic acid pathway (MVA), involving mevalonic
acid in the biosynthesis of steviol, but no direct proof was given to support it. Brandle et. al
(2002)48
sequenced 5548 expressed sequence tags (ESTs) from stevia leaf cDNA library. The
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 11
Jacobs University Bremen
ESTs were classified according to their function in primary or secondary metabolism and many
genes specific to MEP pathway but not the MVA pathway were identified, which concludes that
the source of IPP for diterpenes is through the MEP pathway.
Steviol glycosides share four common steps in biosynthetic pathway with gibberellic acid
formation. After oxidation of ent-kaurene at the C-19 position to ent-kaurenoic acid, the
pathways to the steviol glycosides and the gibberellins diverge. Steviol is produced by
hydroxylation of ent-kaurenoic acid at the C-13 position. Steviol is then glycosylated by series of
UDP-glucosyltransferases (UGTs). UGTs are highly regiospecific and recognize particular
substructure of the acceptor molecule rather than the molecule in its entirety49
. The MEP
pathway of steviol glycosides and the enzymes involved in this pathway is presented in Table 2
and Figure 5.
Table 2.Enzymes involved in biosynthesis of steviol glycosides
Enzyme abbreviation Enzyme
DXS deoxyxyulose-5-phosphate synthase
DXR deoxyxyulose-5-phosphate reductoisomerase
CMS 4-diphosphocytidyl-2-C-methyl-D-erythritol synthase
CMK 4-diphosphocytidyl-2-C-methyl-D-erythritol kinase
MCS 4-diphosphocytidyl-2-C-methyl-D-erythritol 2,4-cyclodiphosphate synthase
HDS 1-hydroxy-2-methyl-2(E)-butenyl 4-diphosphate synthase
HDR 1-hydroxy-2-methyl-2(E)-butenyl 4-diphosphate reductase
GGDPS geranylgeranyl diphosphate synthase
CPS copalyl diphosphate synthase
KS kaurene synthase
KO kaurene oxidase
KAH kaurenoic acid 13-hydroxylase
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 12
Jacobs University Bremen
COO-
O
OPO32-H
OH
O
OPO32-
OH
OH
O
OH
OPO32-
OH
OH
O P2O52-
OO
OH OH
N
N
O
NH2
OH
OHOH
O P2O52-
OO
OH OH
N
N
O
NH2
O
OHOH
PO32-
O
O
OHOH
PO2-
PO2-
O
O
OH
P2O62-
OP2O63- OP2O6
3-
OP2O63-
OP2O63-
H
H
COOH
H
H
COOH
H
H
OH
COOH
H
H
O glc
COOH
H
H
O glc glc
COO
H
H
O glc glc
glc COO
H
H
O glc glc
glc
glc
+
+
DXS
Pyruvate glyceraldehyde-3-phosphate
1-deoxyxylulose-5-phosphate
DXR
CMS
2-C
4-diphosphocytidyl-2-C-methyl-D-erythritol
CMKMCS
4-diphosphocytidyl-2-C-methyl-D-erythritol-2-phosphateHDR
isopentenyl diphosphate dimethylallyldiphosphate1-hydroxy-2methyl-(E)butenyl-4-diphosphate
2-C-methyl-D-erythritol-2,4-cyclodiphosphate
HDS
(-)-copalyl diphosphate
KS
CDPS
GGDPS
geranylgeranyl diphosphate
KOKAH
(-)-kaurenoic acidsteviol
UGT85C2 (-)-kaurene
steviolmonoside steviolbioside stevioside rebaudioside A
UGT
Figure 5.Steviol glycoside biosynthesis via the MEP pathway50
.
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 13
Jacobs University Bremen
SCoA
O
SCoA
O O
SCoAO-O
HO
O
SCoAO-O
HO
O
O-O
HOP
O
O-
OH
O
O-O
HOP
O
O-
OP
O
OH-OO
P
O
O-
OP
O
OH-O
OP
O
O-
OP
O
-OOH
AACT HMGS HMGR
MK
PMKPPMD
IDI
1 2 34
56
7
8
Figure 6.Alternative MVA pathway. Enzymes of the MVA pathway are as follows: AACT,
AcAc-CoA thiolase; HMGS, HMG-CoA synthase; HMGR, HMG-CoA reductase; MK,
mevalonate kinase; PMK, phosphomevalonate kinase; PPMD, diphospho-mevalonate
decarboxylase. 1, Ac-CoA; 2, AcAc-CoA; 3, HMG-CoA; 4, MVA; 5, mevalonate 5-phosphate;
6, mevalonate 5-diphosphate. Both MPE and MVA pathways lead to the formation of compound
8, dimethylallyl diphosphate; 7, isopentenyl diphosphate. The interconversion of IPP into
DMAPP is catalyzed by IDI, isopentenyl diphosphate isomerase51
.
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 14
Jacobs University Bremen
2.5. Analysis of Steviol Glycosides of Stevia rebaudiana
A wide range of analytical techniques have been used to determine the diterpenoid glycosides in
stevia. These techniques include thin layer chromatography (TLC) 52
, capillary electrophoresis
(CE) 53
, near infrared reflectance 54
, and enzymatic methods 55
. But the most used and most
efficient method is high performance liquid chromatography (HPLC or LC). The use of the
hyphenated technique coupled with mass spectrometry (LC-MS) in the analysis of plant extracts
provides important advantages because of the combination of the separation capabilities of LC
and the power of MS as an identification and confirmation method.
In many modern HPLC separations, prepacked columns are used and many types are available
from the manufacturers. However, it is possible to carry out most separations using silica column
for non-polar compounds or reversed phase C18 bonded phase column for polar compounds. The
solvent systems used in the analytical HPLC usually include gradient elutions using solvents of
aqueous acetic, formic or phosphoric acids with methanol or acetonitrile as an organic modifier.
The pH and ionic strength of the mobile phase are known to influence the retention of phenolics
in the column depending on protonation, dissociation, or a partial dissociation. A change in pH
which increases the ionization of a sample could reduce the retention in a reversed phase
separation. Thus, small amounts of acetic (2– 5%), formic, phosphoric or trifluoroacetic acid
(0.1%) are included in the solvent system to enhance ionization of phenolic and carboxylic
groups and hence to improve peak shapes, resolution and reproducubility of chromatographic
runs.
However, for steviol glycosides chromatographic separations in HPLC are not so straight
forward due to the structural similarity of steviol glycosides. Especially for isomer pairs of
stevioside/rebaudioside B and rebeaudioside A/rebaudioside E with the same molecular formula,
resulting in very close retention times in LC thus, resulting in selectivity problems due to peak
overlap and irreproducibility. Therefore, it is still a challenge to achieve efficient separation and
identification for steviol glycoside extracts.
Detection and separation of steviol glycosides on liquid chromatography (LC) were performed
employing amino (NH2)3, 56-60
, C18 61, 62
, and hydrophilic interaction chromatography (HILIC) 63
,
64 columns in combination with mostly UV or MS detection.
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 15
Jacobs University Bremen
Among those, two-dimensional LC 65, 66
and ultra high pressure liquid chromatography (UHPLC)
64, 67 systems were used as well.
Liquid chromatography with amino columns provides good separation of steviol glycosides and
have selectivity for the isomers of stevioside/rebaudioside B and rebaudioside A/rebaudioside E
but they are having disadvantage of reproducibility and long equilibration times 57
.
C18 column exhibits longer retention time and more robustness if compared to amino columns
but also poor selectivity for the separation of stevioside and rebaudioside A. Gradient elution is
not enough to overcome this problem, thus two dimensional systems either with connection of
two C18 columns 68
or C18 with amino column 66
were used.
Hydrophilic interaction chromatography is still new and a useful technique for the retention of
more polar analytes with increased selectivity if compared to reversed phase chromatography.
The interaction of the analytes is believed to be with the water rich layer forming on the surface
of the polar stationary phase against the water poor mobile phase. HILIC can offer a tenfold
increase in sensitivity over reversed-phase chromatography for detection of polar compounds
with mass spectrometry, due to more volatile organic solvent 69
. Some papers 63, 64
describe the
use of HILIC column for stevia extract. In those studies, steviol glycosides were separated with
isocratic elution using 5–20% water in acetonitrile with buffer or formic acid and the robustness
of the separation against changes of buffer concentration and percentage of water differ 64
.
Methods that use UV detection for steviol glycoside quantification are most popular; however
suffer from a series of disadvantages. Detection is typically carried out at 200-210 nm using the
carboxylic acid and olefinic chromophores. These wavelengths are very close to the UV cutoff of
acetonitrile as a solvent and particularly problematic if gradient elution is used. Additionally
many further stevia constituents from the matrix and other impurities absorb at these
wavelenghths. On no occasion was the absence or presence of co-eluting impurities established
in any published steviol glycoside UV method.
Despite some efforts in the development of methods aimed at the identification and
quantification of steviol glycosides until today no validated and certified method exists.
Furthermore no interlaboratory trials were ever conducted on steviol glycoside quantification
allowing a reliable assessment of method validity. Accordingly despite many contributions
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 16
Jacobs University Bremen
published in the field of steviol glycoside analysis, there is still urgent need for critical
assessment of published methods and the development of a generally accepted standard method.
2.6. Phenolic Acids
Polyphenols are secondary metabolites that constitute one of the most widespread groups of
compounds in plants. They are derivatives of the pentose phosphate, shikimate and
phenylpropanoid pathways in plants70
. Polyphenolic compounds contribute to pigmentation of
flowers, fruits, leaves or seeds and play important role in the growth, reproduction and adaptative
strageties of plants71
. In food, phenolics contribute to the bitterness, astringency, color, flavor,
odor, and oxidative stability of products. Moreover, health-protecting capacity and antinutritional
properties of plant phenolics are of great importance to producers, processors and consumers72
.
The antioxidant activity of the dietary polyphenolics is considered to be much greater than that
of the essential vitamins, therefore contributing significantly to the health benefits of fruits73
Phenolic compounds are present in almost all foods of plant origin. Fruits, vegetables, and
beverages are the main sources for these compounds in the human diet. The level of phenolics in
plant sources depend on cultivation techniques, cultivar, growing conditions, ripening process, as
well as processing and storage conditions. In addition, the content of some phenolics may
increase under stress conditions such as UV radiation, infection by pathogens and parasites,
wounding, air polution and exposure to extreme temperatures74
.
Fruit and beverages such as coffee, tea and red wine constitute the main sources of polyphenols.
Certain polyphenols such as quercetin are found in all plant products (fruit, vegetables, cereals,
leguminous plants, fruit juices, tea, wine, infusions, etc), whereas others are specific to particular
foods (flavanones in citrus fruit, isoflavones in soya, phloridzin in apples). In most cases, foods
contain complex mixtures of polyphenols, which are often poorly characterized 75
.
Onions are rich sources of flavonoids76
. Flavonols, the predominant phenolics, are located
mostly in the tomato skin. Cherry tomatoes contained a much higher level of flavonols than
larger size tomato cultivars76, 77
. Anthocyanins are located in the red onion skin and the outer
fleshy layer78
.
The main group of polyphenols includes simple phenols, phenolic acids (benzoic and cinnamic
acid derivatives), coumarins, flavonoids, stilbenes, hydrolyzable and condensed tannins, lignans,
and lignins.
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 17
Jacobs University Bremen
Phenolic acids consist of two subgroups; the hydroxybenzoic and hydroxycinnamic acids.
Hydroxybenzoic acids include gallic, p-hydroxybenzoic, protocatechuic, vanillic and syringic
acids, which in common have the C6–C1 structure. Hydroxycinnamic acids, on the other hand,
are aromatic compounds with a three-carbon side chain (C6–C3), with caffeic, ferulic, p-coumaric
and sinapic acids (Figure 7) being the most common 71, 79
.
OH
R2R1
COOH OH
OH
COOH
COOH
R1
HO
hydroxybenzoic acid protocatechuic acid hydroxycinnamic acid
HO
HO COOH
HO
COOH
HO
H3CO COOH
caffeic acid p-coumaric acid ferulic acid
Figure 7.Examples for hydroxybenzoic and hydorxycinnamic acids.
Caffeic acid is the major representative of hydroxycinnamic acids and occurs in foods mainly as
chlorogenic acid (5-caffeoylquinic acid). Chlorogenic acids (CGAs) are a family of esters
formed between one or more residues of certain trans-cinnamic acids and quinic acid (1L-1
(OH),3,4/5-tetrahydroxycyclohexane carboxylic acid) which have axial hydroxyls on carbons 1
and 3 and equatorial hydroxyls on carbons 4 and 5. During processing, trans isomers may be
partially converted to cis isomers 80, 81
(Figure 8). The main classes of CGAs found in nature are
the caffeoylquinic acids (CQA), dicaffeoylquinic acids (diCQA), and, less commonly,
feruloylquinic acids (FQAs), each group with at least three isomers82
. CGAs are antioxidant
components produced by plants in response to environmental stress conditions such as infections
by microbial pathogens, mechanical wounding, and excessive UV or visible light levels83
Chlorogenic acids make up 5-10% of the weight of coffee beans and plays a significant role in
coffee color and aroma formation84
.
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 18
Jacobs University Bremen
OR1
C
OR3
OR4
OR5O
HO
6
1
OH
C
OH
O
OHO
HO
OH
OH
O
quinic acid trans -4-caffeoylquinic acid
OH
C
OH
O
OHO
HO
O
OHOH
cis-4-caffeoylquinic acid
Figure 8.General structure of quinic acid and one of chlorogenic acids as an example.
In addition to being found in coffee, these compounds are also found at significant levels in plant
foods such as apples, pears, tomato, potato, and eggplant85
. Coffee is a major source of
chlorogenic acid in the human diet; daily intake in coffee drinkers is 0.5–1 g; coffee abstainers
will usually ingest < 100 mg/d 86
.
In the last few years, CGAs has been the subject of several investigations in their potentially
beneficial effects in humans involving their antioxidant activity, among other beneficial effects.
Several pharmacological activities of CGAs including antioxidant activity, the ability to increase
hepatic glucose utilization,87-94
inhibition of the HIV-1 integrase,95-97
antispasmodic activity,98
and inhibition of the mutagenicity of carcinogenic compounds99
have been revealed by in vitro,
in vivo, and human intervention studies so far. CGAs and their metabolites display additional
highly favorable pharmacokinetic properties.100-102
Flavonoids are low molecular weight compounds, consisting of fifteen carbon atoms, arranged in
a C6–C3–C6 configuration. Essentially the structure consists of two aromatic rings, A and B,
joined by a 3-carbon bridge, usually in the form of a heterocyclic ring, C. The aromatic ring A is
formed via glucose metabolism with condensation of malonyl-coenyme A (CoA) catalyzed by
chalcone synthetase, while ring B and C is derived from phenylalanine through the shikimate
pathway, which is converted to cinnamic acid and to coumaric acid. Coumaric acid CoA and
three malonyl CoAs are condensed in a single enzymatic step to form naringenin chalcone. The
C-ring closes and becomes hydrated to form 3-hydroxyflavonoids (e.g., catechins), 3,4-diol
flavonoids (e.g., quercetin), and procyanidins103
.
Variations in the heterocyclic ring C give rise to the major flavonoid classes, i.e., flavonols,
flavones, flavanones, flavanols (or catechins), isoflavones, flavanonols, and anthocyanidins
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 19
Jacobs University Bremen
(Figure 9), while individual compounds within a class differ in the pattern of substitution of the
A and B rings. These substitutions may include oxygenation, alkylation, glycosylation, acylation,
and sulfation 79, 104
.
O
A C
B
1'6'
5'
4'3'
2'
2
3
45
7
8
6
O
O
O
O
OH
O
O
O
OH
O+
OH
Flavone Flavonol Flavanone
Flavanol Antocyanidin
Generic structure
Figure 9.Generic structure of major classes of flavonoids.
Within different subclasses of flavonoids, further differentiation is based on the number, position
and nature of substituent groups attached on the rings. Mostly they are sugars, such as glucose,
galactose, rhamnose, arabinose, xylose and rutinose. Flavonoid glycosides have many isomers
with the same molecular weight but different aglycone and sugar component at different
positions attaching on the aglycone ring 72, 105, 106
. Flavonoid glycosides as well are commonly
encountered in plant material and following ingestions these glycosides are hydrolysed by the
human microbial gut flora into their aglycones, which are subsequently absorbed and show
significant bioavailability.
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 20
Jacobs University Bremen
Flavonoids play different roles in the ecology of plants. Due to their attractive colors, flavones,
flavonols, and anthocyanidins may act as visual signals for pollinating insects. Because of their
astringency, catechins and other flavanols can represent a defense system against harm of insects
to the plant. Furthermore flavonoids protect plants from UV radiation of sun with their favorable
UV-absorbing properties104
.
Apart from their roles in plants, flavonoids play important role in human diet. Flavonoids are
important antioxidants (hydrogen-donating radical scavengers) due to their high redox potential,
which allows them to act as reducing agents, hydrogen donors, and singlet oxygen quenchers.
The antioxidant property of flavonoids may protect tissues against oxygen free radicals and lipid
peroxidation. Thus, flavonoids might contribute to the prevention of atherosclerosis, cancer and
chronic inflammation107
. In addition, they have a metal chelating potential, which play an
important role in oxygen metabolism and are essential for many physiological functions 108
. The
proposed binding sites for trace metals to flavonoids are the catechol moiety in ring B, the 3-
hydroxyl, 4-oxo groups in the heterocyclic ring, and the 4-oxo, 5-hydroxyl groups between the
heterocyclic and the A rings. However, the major contribution to metal chelation is due to the
catechol moiety, as exemplified by the more pronounced bathochromic shift produced by
chelation of copper to quercetin compared to that of kaempferol (similar in structure to quercetin
except that it lacks the catechol group in the B ring)104
.
Flavonoids and phenolic acids have protective role in carcinogenesis, inflammation,
atherosclerosis, thrombosis and have high antioxidant capacity. Furthermore, flavonoids have
been reported as aldose reductase inhibitors blocking the sorbitol pathway that is linked to many
problems associated with diabetes106
.
Phenolic acids in stevia were analyzed by HPLC on a C18 column by Kim et. al 109
and the main
phenolic compounds found were pyrogallol, 4-methoxybenzoic acid, p-coumaric acid, 4-
methylcatechol, sinapic and cinnamic acid. The flavonoids detected in stevia leaves belong to the
subgroups of flavonols and flavones. They were identified using two-dimensional UHPLC-DAD
and LC-MS/MS and spectroscopic methods (1H and
13C NMR, IR, and 2D NMR)
110, 111.
There is currently great interest in phenolic acids research due to the possibility of improved
public health through diet, where preventative health care can be promoted through the
consumption of fruit and vegetables. Therefore, the presence of such compounds with proven
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 21
Jacobs University Bremen
health benefits in stevia extracts would affect the the organoleptic properties of stevia based
products and would add health aspects to the use of stevia as sweenetening agent.
2.7. Proteomics of Stevia rebaudiana
Determination of amino acid content and potential allergenic proteins in food is very significant
and necessary research in the food formulation processes. Currently, there has been no published
data related to stevia allergens and there is only one published paper for the protein analysis in
stevia112
. Therefore, screening and quantitative analysis of stevia proteins and any potential
allergen is crucial. Detailed and comprehensive characterization of plant-derived food allergens
can be carried out using proteomics. In proteomics, after the separation and purification of the
protein, proteins are identified by mass spectrometry. Proteomic technologies using 2D-PAGE
and immunoblotting are then applied in the identification of new allergens113
.
In the past, protein determination was carried out by mRNA analysis, but later it was found that
there was no correlation with protein content as gene expression is regulated post-
transcriptionally and translation from mRNA cause differences 114, 115
. Most proteins are
chemically modified through post-translational modifications, mainly through the addition of
carbohydrate and phosphate groups. Such modifications play an important role in modulating the
function of many proteins. The most common post-translational modifications include
glycosylation, phosphorylation, ubiquitination, methylation, acetylation, and lipidation115
.
The major methods to study proteins include, high quality separation of proteins in two
dimensions (2D-SDS PAGE), characterization of separated proteins by mass spectrometry and
information collection using bioinformatic tools and databases115
. Matrix-assisted laser
desorption/ionization (MALDI) and electrospray ionization mass spectrometry (ESI) are widely
used techniques for proteomic studies.
Sample preparation is the most important step in the analysis of proteins from plants due to the
low protein content relative to other systems and the large quantities of polysaccharides, lipids,
phenolics and other secondary metabolites. Pretreatment of samples for 2D electrophoresis
involves solubilization, denaturation and reduction to completely break up the interactions
between the proteins and removal of all interfering compounds (phenolic compounds, nucleic
acids) to ensure efficient separation115
.
The most common protein extraction protocol is based on precipitating proteins from
homogenized tissue or cells with trichloracetic acid (TCA) in acetone. An alternative protocol is
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 22
Jacobs University Bremen
based on the solubilization of proteins in phenol, followed by their precipitation with ammonium
acetate in methanol. No one protocol is necessarily more appropriate than the other; studies have
suggested that they are complementary116
. In order to evaluate the effectiveness of a given
extraction protocol, protein quantification is needed. Bradford, Lowry and BCA assays are the
most common colorimetric methods.
Identification of proteins proceeds with separation of proteins from the extract with 2D-SDS
PAGE and subsequent digestion of the individual separated proteins or digestion of the entire
protein mixture followed by separation of the resulted peptides117
.
2D SDS PAGE separates proteins in two dimensions; the first dimension in a pH gradient
according to their isoelectric point (pI), and in the second dimension, the proteins is separated
according to their molecular weight.
The first dimension of electrophoresis involves denaturing isoelectric focusing using
immobilized pH gradient gels (IPG). IPG strips have a gradient of charge imbedded in
acrylamide. IPG strips improve the reproducibility and reliability and overcome pH gradient
instability. Strips come in a variety of pH ranges and lengths (from 7 to 24 cm). Samples can be
applied on the strips by cup loading or by in-gel rehydration. In cup loading method, the strips
are pre-rehydrated with rehydration buffer and the samples are applied into the loading cup at
either acidic or basic end. In in-gel rehydration, the sample in lysis buffer is diluted with the
rehydration buffer. The IPG matrix absorbs the proteins. Isoelectric focusing (IEF) is carried out
on a first dimension electrophoresis unit consisting of five phases of stepped voltage from 500 to
3500 V (Multiphor) or 500 to 8000 V (IPGPhor)115
. After completion of the first dimension the
proteins are separated according to their mass in second dimension using SDS-PAGE. Separation
of proteins in second dimension is based on differences in their electrophoretic mobility due to
differences in their size. SDS is a very effective solubilising agent for a wide range of proteins.
The majority of proteins bind SDS at a ratio of 1.4 g SDS / 1 g protein to form negatively
charged complexes118
. Proteins are transferred electrophoretically from the IEF strip into a
narrow starting zone prior to entering the main separating gel. This concentrates the proteins and
results in very sharp bands or spots. Once the protein samples have entered the separating gel,
the negatively charged protein-SDS complexes continue to move towards the anode. As they
pass through the separating gel the proteins are resolved on the basis of their size because of the
molecular sieving properties of the gel. After 2D electrophoresis, the protein spots can be
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 23
Jacobs University Bremen
visualized by staining with either coomassie brilliant blue stain, which will detect proteins
present in amount greater than 100 ng or with silver staining for amounts in ng range 119
.
The end point of any proteomics expression is to identify and characterize the proteins. Edman
degradation was the standard method for protein sequencing for the last 25 years120
. Other
traditional approaches for protein identification include the use of antibodies to perform Western
blots. However, this method has restricted use due to antibodies non-specific binding and the
availability of antibodies to all proteins121
.
Development in mass analysis techniques for mass spectrometry (MS) and the ability to correlate
MS data of proteins to sequences in databases have opened up new possibilities in protein
sequencing.
Protein identification via MS can be carried out in the form of whole-protein analysis (‘top-
down’ approach) or analysis of enzymatically or chemically produced peptides (‘bottom-up’
approach). To date, one of the most common methods of identifying proteins is through peptide-
mass fingerprinting (PMF). The proteins are digested with a proteolytic enzyme such as trypsin,
to produce a set of tryptic fragments unique to each protein115
. Summary of other MS-based
proteomic strategies is presented in Figure 10.
Figure 10.Strategies for MS-based protein identification122
.
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 24
Jacobs University Bremen
Mass spectrometry analysis of peptides and proteins relies exclusively on soft ionization
techniques that create intact gas-phase ions from biomolecules. The creation of intact molecular
ions enables accurate measurement of molecular weight. The electrospray ionization (ESI) and
matrix-assisted laser desorption/ionization (MALDI) techniques are widely used in proteomic
studies. The most common mass analyzers used with MALDI are time-of-flight (TOF) mass
spectrometers.
MALDI-TOF allows the analysis of high molecular weight compounds with high sensitivity and,
soft ionization with little or no fragmentation. MALDI uses a solid matrix to co-crystallize with
peptides/proteins on a sample plate and a laser light as its ionizing beam. Ionization occurs when
these matrix molecules absorb the energy provided by a laser (usually 337 nm). Release of the
energy causes a rapid thermal expansion of matrix and analyte into the gas phase. Proton transfer
from analyte to matrix may result in charge reduction to the singly charged ion observed in the
gas-phase123
The matrix is typically a small energy absorbing molecule such as α-cyano-4-
hydroxycinnamic acid (HCCA) or 2,5,-dihydroxybenzoic acid (2,5-DHB). The molecular weight
values of the trypsinized peptides or intact proteins obtained by MALDI-TOF are then used to
identify the predicated proteins using web-based search engines such as MASCOT.
In cases where the protein is not present in the database, the proteins may be analyzed by tandem
mass spectrometry (ESI-MS). The fragment ions observed in MS/MS spectrum is analyzed to
derive the order of the amino acids in the tryptic peptides or in the intact protein. This method is
known as de novo sequencing. If the fragment ion carries its charge on the N-terminal, the ion is
categorized as a, b or c. If the charge is on the C-terminal of the fragment the type of the ion can
be x, y or z (Figure 11). The difference in the mass between adjacent y- or b-ions corresponds to
that of an amino acid. This can be used to identify the amino acid and, hence the peptide
sequence115
.
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 25
Jacobs University Bremen
N C C N C
R1
H
O R2
H
C
O
N C
R3
H
C
O
HN C
R4
H
C
O
OH
H
H
H H
H+
N C C N C
R1
H
O R2
H
C
O
N C
R3
H
C
O
HN C
R4
H
C
O
OH
H
H
H H
H+
Figure 11.Peptide fragmentation momenclature.
2.8. Lipid Analysis
Lipids play an important role in physiology and pathophysiology of living systems. All plant
cells produce fatty acids from acetyl-CoA by a common pathway localized in plastids124
. Fatty
acyls (FAs) are group of molecules synthesized by chain-elongation of an acetyl-CoA primer
with malonyl-CoA or methylmalonyl-CoA groups. Structures with a glycerol group are
represented by two categories: the glycerolipids (GLs), composed mainly of mono-, di- and tri-
substituted glycerols, and the glycerophospholipids (GPs), which are defined by the presence of
a phosphate (or phosphonate) group esterified to one of the glycerol hydroxyl groups. Other
compounds including fatty chains [e.g., fatty acids, fatty alcohols and inter-fatty esters (waxes)]
are also considered in this category. The sterol lipids (STs) and prenol lipids (PRs) share a
common biosynthetic pathway via the polymerization of dimethylallyl
pyrophosphate/isopentenyl pyrophosphate but have obvious differences in terms of their eventual
Y3
Y2 Y1
b3 b2 b1
a1 a2 a3
x3
a1
z3
c1
x2
a2
z2
c2
x1
a3
c3
z1
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 26
Jacobs University Bremen
structure and function. Another well-defined category comprises sphingolipids (SPs), which
contain a long-chain base as their core structure. Saccharolipids (SLs) comprise lipids in which
fatty acyl groups are linked directly to a sugar backbone. The final category comprises
polyketides (PKs), which are a diverse group of metabolites from plant and microbial sources125
(Figure 12). Although they are different in their chemical composition, they all share one
characteristic, which is solubilization in non-polar solvents, such as chloroform and hexane.
OH
O
O O
O
HO
O
P
O
O
OHN+
HO
H
H
H H
H
O
OP
HO
O
O P
O
HOO
O
N
NH
O
O
OHOH
HNO
O
HO
HO
O
OH
O OH
O
HO
O
OH
HNH
O
OHH
OH
O
O
O
H
H
O O
O
Fatty acyls: hexadecanoic acid
Glycerophospholipids: 1-hexadecanoyl-2-(9Z-octadecenoyl)-sn-glycero-3-phosphocholine
Sterol lipids:cholest-5-en-3-b-ol
Saccharolipids:UDP-3-O-(3R-hydroxy-tetradecanoyl)-a-D-N-acetylglucosamine
Glycerolipids: 1-hexadecanoyl-2-(9Z-octadecenoyl)-sn-glycerol
Sphingolipids: N-(tetradecanoyl)-sphing-4-enine
Prenol lipids: 2E,6E-farnesol
Polyketides: aflatoxin B1
Figure 12.Examples of lipid categories.
Lipids are one of the most important metabolites of the organism. Essential fatty acids and fat-
soluble vitamins, which are required by organism, can be supplied from lipids. Terpenes and
steroids like vitamin A, D, E, K, cholic acid, and steroid hormones are related with nutrition,
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 27
Jacobs University Bremen
metabolism, and regulation function. Lipids on the surface of the organisms protect its surface
from mechanical damages and heat losing, and for the cells, it is closely related to cell
recognition 126
. Apart from their biological functions, fatty acids, triacylglycerols and
phospholipids are the primary classes of lipids of interest in foods.
In general, lipid analysis involves three basic steps which are; 1) extraction of lipids from the
sample, 2) analytical separation, 3) identification and quantification of lipids.
Lipids are mainly extracted from cells, plasma, and tissues. The components obtained depend on
the method of extraction used, especially the solvent. All lipids have a polar head and a nonpolar
tail. Therefore, mixture of chloroform and methanol in a two-step extraction was chosen to
obtain better dissolution of lipids. This approach was developed in 1950s by Folch127
. Lipids of
all major classes could be recovered via chloroform/methanol extraction, typically according to
the Folch, or Lees, and Sloane Stanley or Bligh and Dyer protocols, in which they are mostly
enriched in the chloroform phase 128
.
The most widely used method for the extraction of solid samples is Soxhlet extraction. Purified
lipid extracts are susceptible to oxidation. They should be dissolved in a non-polar solvent (e.g.,
hexane or chloroform) and stored at −20°C in a glass container in a nitrogen atmosphere. They
can be stored in refrigeration temperature (0–4°C) for short-term 125
.
Several separation techniques have been used for the determination of lipids. Long-chain fatty
acids have been determined by gas chromatography (GC) or liquid chromatography (LC).
Supercritical fluid chromatography (SFC) 129
and thin layer chromatography (TLC) 130
was also
utilized for lipid separation.
Traditionally, lipids have been analyzed using gas chromatographic (GC) separation with flame-
ionization detection (FID) or mass spectrometry (MS) detection. Compounds must be thermally
stable with high vapor pressure to be volatilized during the injection in to the GC. Therefore
lipids have to be converted into derivatives with lower boiling points, such as alcoholic esters.
Lipids can be analyzed after hydrolysis, derivatization, or pyrolysis with GC technique. Fatty
acid methyl ester derivatization is the most common method used for analysis with GC
technique. Transesterification is one mechanism that can be employed to form FAMEs from
fatty-acid esters in foods. Alkali- or acid-catalyzed transesterification procedures can be used to
form FAMEs in a methanolic medium. The separation of FAMEs is usually achieved on highly
polar liquid phases and the analytes are separated according to their chain length and degree of
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 28
Jacobs University Bremen
saturation. The identification of the fatty acids in the sample can be achieved with comparison of
the retention times of FAME reference compounds. 125
. Although the derivatization solves the
problem of sample volatility, problems with ester formation may include the incomplete
conversion of fatty acids to FAME, loses of highly volatile short chain fatty acids or formation of
contaminations which can overlap with the FAME peak in the GC chromatogram131
.
The wide choice of mobile and stationary phase makes selectivity extremely powerful in HPLC.
Depending on the mobile phases and stationary phases used, there are two modes in LC; which
are normal phase (NP) and reversed phase (RP). RP-LC methods have been developed utilizing
both aqueous and non-aqueous solutions for lipid analysis, whereas in NP-LC, non-polar
solvents are used for separation. RP-LC separates lipids according to their fatty acyl composition
and in NP-LC separation occurs on the basis of their class of compounds125
. Lipids with
molecular weights of 100 – 2000 Da can be detected by LC-MS126
.
Electrospray ionization-mass spectrometry (ESI-MS) has been successfully applied to lipid
analysis. Especially, the combination of chromatographic techniques with MS provides the
technical support for the analysis of lipids and accelerates the emergence of the lipidomics.
Lipidomics, focuses on the global analysis of lipids and their metabolites. The concept
lipidomics was raised by Han et al. in 2003126, 132
.
Strategies currently used in lipidomics include direct-infusion ESI–MS and ESI–MS/MS, LC
coupled with ESI–MS or MS/MS, and MALDI combined with Fourier transform ion cyclotron
resonance MS (MALDI–FTICR–MS) or time-of-flight–MS (MALDI–TOF–MS). In addition, for
some classes of lipids, LC coupled with APCI–MS was also used133
. The analytes can be directly
injected to MS without prior separation and soft ionization. ESI-MS has the advantage that the
structural identification of lipids is more straightforward using different MS/MS experiments
such as precursor ion scan, product ion scan and neutral loss scan. However, ion suppression can
be the major complication in the direct injection MS experiments. MALDI-MS is a laser-based,
soft-ionization method that is often used for analysis of large proteins, but has also been used
successfully with lipids. Important advantages of MALDI–MS in lipid analysis are the speed of
analysis and simplicity of operation: the analytes are ionized under relative soft conditions by
laser desorption using an ultraviolet-absorbing matrix. One important disadvantage of MALDI is
the presence of a lot of background in the lower mass range due to the matrix molecules. In
addition, MALDI–MS is generally less quantitative compared to ESI–MS technique133
.
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 29
Jacobs University Bremen
Currently, a wide range of analytical techniques are used for analyzing the lipids, however none
of them provides a global lipid profiling. Therefore, the more different techniques we use, the
wider aspect we can have about our sample.
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 30
Jacobs University Bremen
3. RESEARCH OBJECTIVE
The project aims to carry out the chemical profiling of plant called Stevia rebaudiana. Stevia
produces, as its main secondary metabolite diterpene glycosides (steviol glycosides), which are
natural sweeteners. As a sweetener it has two advantages. Firstly, as a terpene it does not cause
allergic reactions unlike most peptide based sweeteners. Secondly, stevia grows almost on any
soils, in particular on fields where tobacco used to be grown. Due to pH of soil become acidic
after tobacco cultivation, very few plants have this specialty and the stevia cultivation in
European Union would offer the tobacco farmers an alternative crop. Most importantly, the
advantages and health benefits of natural sweeteners from stevia make it promising as a novel
food in near future.
Within the EU project (DIVAS) Stevia rebaudiana was cultivated in variety of locations in the
Mediterranean region of Europe over a period of two years using seven different botanical
varieties of stevia. The leaves were harvested between two and three times annually producing a
total of 166 different samples of Stevia rebaudiana leaves. The objective of the project was to
provide scientific basis for a future specification of stevia as novel food in Europe. For this
purpose, chemical specification of stevia leaves has been carried out. Methods have been
developed and used to study the major classes of secondary metabolites. Analysis of the
chemical composition of stevia leaves will allow definition of upper and lower limits of all
relevant stevia plant constituents that are appropriate to chemical analysis including proteins,
lipids and secondary plant metabolites comprising polyphenols and terpene glycosides.
As stevia may be cultivated within various regions of EU with different soil and climatic
conditions it is important to know whether an EU-common specification will be achieved and
how stevia leaves from regions outside EU can be distinguished on a scientific basis. For this
purpose, dataset obtained from stevia leaves from various parts of Europe were analyzed
statistically.
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 31
Jacobs University Bremen
4. STEVIOL GLYCOSIDES ANALYSIS BY LC-MS
4.1. Overview
Steviol glycosides extracted from Stevia rebaudiana leaves were subjected to LC- tandem MS
and LC-TOF MS for characterization. The separation of steviol glycosides were compared using
C18 reverse phase column and hydrophilic interaction chromatography (HILIC) column.
Additionally tandem MS data of steviol glycosides is presented using an ion trap instrument,
taking advantage of low energy collision induced fragmentation and multi stage fragmentation
up to MS4 to identify steviol glycosides. LC-TOF method using HILIC column was validated
and used for quantification of steviol glycosides in 166 stevia leaves extracts harvested in
Europe.
4.2.Materials & Methods
4.2.1. Extraction Method
Extraction of steviol glycosides, as well as phenolic acids, lipids and volatile terpenes were
achieved using soxhlet extraction system. Soxhlet conditions were optimised with respect to
solvent volume, extraction cycles and time. Prior to the steviol glycoside extraction a chloroform
extraction was carried out to remove the lipid fraction. In addition to the steviol glycosides the
methanolic fraction contained phenolics in quantitative amounts used for quantification.
Extraction of Steviol glycosides and Phenolic acids: Two grams of S. rebaudiana leaves were
immersed in liquid nitrogen, ground in a hammer mill, and extracted first with 150 mL of
chloroform in a Soxhlet apparatus (Buchi B-811 extraction system) for 2 h and then with 150 mL
of methanol for another 2 h. Solvents were removed from the methanolic extract in vacuo, and
extracts were stored at - 20 oC until required.
4.2.2. LC-MS Analysis of Steviol glycosides
Compound identification was carried out using high resolution mass spectrometry (HR-MS) and
tandem MS using an ion trap mass spectrometer. A HR-MS using an ESI-TOF-MS experiment
allowed determination of molecular formulae based on the accurate mass measurements.
Molecular formulas were in general accepted if an error below 5 ppm was experimentally
observed, as accepted by all peer reviewed chemistry journals.
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 32
Jacobs University Bremen
LC-TOF MS: The LC equipment (Agilent 1100 series, Bremen, Germany) comprised a
binary pump, an autosampler with a 100 μL loop, and a diode array detector with a light-pipe
flow cell (recording at 210 nm and scanning from 200 to 600 nm). This was interfaced with a
MicroTOF Focus mass spectrometer (Bruker Daltonics) fitted with an ESI source. The MS
parameters were: nebulizer 1.6 bar, dry gas 12.0 L/min, dry temperature 220 0C. The
MicroTOF was operated in negative ion mode and the mass range was 150 – 1200 m/z.
Internal calibration was achieved with 10 mL of 0.1 mol/L sodium formate solution injected
through a six-port valve prior to each chromatographic run. Calibration was carried out using
the enhanced quadratic calibration mode.
LC-MSn
(tandem MS): The LC equipment (Agilent 1100 series, Bremen, Germany)
comprised a binary pump, an autosampler with a 100 μL loop, and a diode array detector with
a light-pipe flow cell (recording at 210 nm and scanning from 200 to 600 nm). This was
interfaced with an ion-trap mass spectrometer fitted with an ESI source (Bruker Daltonics
HCT Ultra, Bremen, Germany) operating in Auto-MSn mode to obtain fragment ions m/z.
Tandem mass spectra were acquired in Auto-MSn mode (smart fragmentation) using a
ramping of the collision energy. Maximum fragmentation amplitude was set to 1 V, starting at
30% and ending at 200%. MS operating conditions (negative mode) were capillary
temperature of 365 oC, a dry gas flow rate of 10 L/min, and a nebulizer pressure of 50 psi.
4.2.3. HPLC conditions
HILIC conditions: Separation was achieved on a 4,6 x 150 mm Dionex Acclaim Mixed Mode
Wax-1 column with 5 μm particle size. Solvent A was 10 mM ammonium formate buffer at pH
3 and solvent B was acetonitrile (ACN). Solvents were delivered at a total flow rate of 0.5
mL/min and the column temperature was set to 40 oC. 5 μL of samples in 80% ACN/water were
injected in to LC-MS system, unless stated otherwise. The isocratic profile was 85 %ACN and
15% water (10 mM ammonium formate buffer).
Reverse phase (C18) conditions: Separation was on a 250 x 3 mm C18 column (Varian Pursuit
XRS) with 5 μm particle size. Solvent A was water/formic acid (1000+0.005 v/v), and solvent B
was acetonitrile (ACN). Solvents were delivered at a total flow rate of 0.5 mL/min and the
column temperature was set to 25 oC. 5 μL of samples were injected in to LC-MS system,
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 33
Jacobs University Bremen
unless stated otherwise. The gradient profile was 10 to 80% B in 60 min and a return to 10% B
at 65 min and 5 min isocratic to re-equilibrate.
4.2.4. Calibration Curve of Steviol Glycoside Standards
Stock solutions of the standard compounds of stevioside, rebaudioside A and steviolbioside were
prepared in 80% ACN/water. A series of standard solutions was injected (5μL) into the LC-MS
system. The areas of the peaks of each standard from extracted ion chromatograms (EIC) were
used to make the respective standard curves.
4.2.5. Method Validation
Selectivity of the method was determined by comparing the chromatograms of leaf extract and
reference compounds. Precision was determined by intra and inter-day measurements with three
different concentration of standard solution of stevioside and rebaudioside A on the HILIC
column and evaluated by the relative standard deviation (%RSD). Accuracy of the method was
determined by spiking two different stevia leaf extracts with three different amounts of
stevioside and rebaudioside A, separately and RSD was calculated. Quantification was achieved
by applying calibration curve equations obtained by the least square method.
4.2.6. Solid Phase Extraction (SPE) of Steviol glycosides
Extraction of stevia leaves: 0.8 g of stevia pulverized leaves were sonicated and heated with 30
mL of ACN/water (70:30 v/v) for 15 minutes. Then, the extract was filtrated through a 0.45 μm
filter.
SPE Material I Cartridges were filled with the steviaclean stationary phase specially produced
for Stevia rebaudiana (Knauer GmbH) in the amounts of 0.2 g, 0.4 g and 0.6 g. Each was
condinitioned with water (1mL) and 3 mL of ACN/water (90:10 v/v). 1mL of stevia extract was
loaded on the cartridge. The steviol glycosides were eluted with 2 mL of ACN/water (90:10 v/v).
The eluate was filtered and subjected to HPLC analysis with amino column (Knauer, Eurospher
100 NH2, 5 μm, 150 x 3mm). 5 μL of samples were injected in to LC-MS system. Solvents were
delivered at a total flow rate of 1.0 mL/min and the column temperature was set to 35 oC. UV
detection was at 210 nm. 134
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 34
Jacobs University Bremen
SPE Material II Bond Elut C18 cartridges were conditioned with 3 mL of ACN/water (90:10
v/v). 1 mL of stevia extract was loaded on the cartridges. The steviol glycosides were eluted with
ACN/water (90:10 v/v). The eluate was subjected to HPLC analysis with amino and HILIC
columns.
4.3. Results & Discussion
An optimized analytical method was developed for steviol glycoside analysis in stevia leaf
methanol extracts using LC-MS.
Typically for steviol glycoside extraction, solvent extraction using hot water or acetonitrile is
employed followed frequently by further solid phase extraction (SPE) sample clean up. The
challenge of steviol glycoside extraction lies in the different solubilities of steviol glycosides in
aqueous and organic solvents. Good solubility of all steviol glycosides has been reported for
water, however co-extraction of phenolic constituents and carbohydrates exacerbate separation
problems and therefore analysis.
For this reason, first, optimization to use different organic solvents for steviol glycoside
extraction from dried leaf material using Soxhlet extraction was performed. Despite reports that
rebaudioside A shows moderate methanol solubility, we first optimized for Soxhlet extraction
using methanol. Prior to Soxhlet extraction 2 g of dried leaves were treated with liquid nitrogen
and crushed and milled using a blade mill. Extraction times, extraction cycles and extraction
volume were optimized by multiple extraction experiments and it was concluded that using 2 g
of leaf material allowed for a reproducible amount of steviol glycosides being extracted from
notoriously heterogenic plant material.
For development of the LC-MS method, a standard reversed phase C18 column to a HILIC
column was compared. C18 columns have the advantage that steviol glycosides and all phenolic
constituents can be analysed and quantified present in stevia leaves, however retention times are
long and selectivity is not satisfactory. Co-elution of rebaudioside A with stevioside was
observed using a C18 column (Figure 13). At retention times up to 25 min. a total of twelve
chlorogenic acids and nine flavanone glycosides were detected from the analysis of
commercially obtained stevia leaves (please refer to article attached in appendix).
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 35
Jacobs University Bremen
Figure 13.Total ion chromatogram in negative ion mode using a C-18 column of methanolic
Stevia rebaudiana extract showing phenolics (chlorogenic acids, flavonoids and steviol
glycosides).
A total of eight steviol glycosides could be observed at retention times from 30 to 42 minutes
with rebaudioside A and stevioside co-eluting. Efficient separation and resolution of this isomer
pair and rubusoside/steviolbioside were achieved with HILIC column using acetonitrile/water
(10 mM ammonium formate) as solvent in the HPLC method. In contrast to C18 column, the
elution order is inverted on the HILIC column. The more glucose units attached to the backbone
structure resulted in later retention times on HILIC. Thus, it was not possible to detect steviol on
the HILIC column. A base peak chromatogram (BPC) is presented in Figure 14 showing baseline
separation of all naturally occurring steviol glycosides. Characterization of steviol glycosides
was achieved by ion-trap mass spectrometry with selected ion monitoring (SIM), and
confirmation of elemental composition was provided by ESI-TOF measurements (Table 1).
Figure 14.Base peak chromatogram of steviol glycosides obtained using HILIC column.
RebA
0.00
0.25
0.50
0.75
1.00
1.25
7 x10
0 5 10 15 20 25 30 35 40 45 Time [min]
stevioside
Chlorogenic acids and flavonoids Diterpene glycosides
Intens
.
Steviolbioside
Stevioside
RebA RebC
Dulcoside
A
Rubusoside
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 36
Jacobs University Bremen
4.3.1. Identification of Steviol Glycosides
Compound identification was carried out using high resolution mass spectrometry (HR-MS)
followed by tandem MS using an ion trap mass spectrometer. HR-MS values for all steviol
glycosides are given in Table 1.
COORH3C
CH3
OR1
CH2
General structure ofsteviol glycosides
Table 1.High resolution mass spectrometrical data of stevia terpene glycosides in negative ion
mode from LC-TOF MS analysis
Compound R R1 Molecular
Formula
Experimental
m/z (M-H+)
-
Theoretical
m/z (M-H+)
-
Relative Error
(ppm)
Steviol
Steviolbioside
H
H
H
glc2 - 1glc
C20H30O3
C32H50O13
317.0819
641.3181
317.0717
641.3179
9.0
0.4
Rubusoside Glc glc C32H50O13 641.3166 641.3179 2.0
Stevioside Glc glc2 - 1glc C38H60O18 803.3751 803.3707 5.5
Rebaudioside A Glc glc32 -1glc
1glc
C44H70O23 965.425 965.4235 1.6
Rebaudioside B H glc32 -1glc
1glc
C38H60O18 803.368 803.3707 2.8
Rebaudioside C
(Dulcoside B)
Glc glc32 -1rham
1glc
C44H70O22 949.427 949.4286 1.7
Rebaudioside D glc2-1glc glc32 -1rham
1glc
C50H80O28 1127.4726 1127.4763 3.3
Rebaudioside E glc2-1glc glc2-1glc C44H70O23 965.4199 965.4235 3.7
Rebaudioside F Glc glc32 -1xyl
1glc
C43H68O22 935.4097 935.4129 3.5
Dulcoside A Glc glc2 - 1rham C38H60O17 787.3732 787.3758 3.3
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 37
Jacobs University Bremen
Representative tandem MS spectra and fragmentation pathways are shown in Figure 15 and
Figure 16 for rebaudioside A ([M-H+]
- m/z 965) and rebaudioside E ([M-H
+]
- m/z 965). These
two isomers were differing in their MS2 fragmentation, resulting with m/z 803 peak [M-H
+-glc]
-
with the loss of one glucose unit ([M-H2O]- m/z 162) for rebaudioside A, and m/z 641 peak with
the loss of two glucose units for rebaudioside E ([M-H+-2glc]
-). It is most probable that the both
rebaudiosides are losing the glucose first from carboxylic acid moiety due to the increased
stability of the resulting resonance stabilized anion. In general steviol glycosides could be
characterized up to MS4. Rebaudioside A loses one glucose unit each in MS
3 and MS
4 resulting
with the m/z 479 and m/z 317 peaks corresponding to [M-H+-3glc]
- and [M-H
+-4glc]
- ions,
respectively.
Figure 15.Mechanism of fragmentation in tandem MS spectra of Rebaudioside A and
Rebaudioside E illustrating how isomeric compounds can be distinguished by tandem MS.
O
OH
HO
O O
HO
O
CH3
H3C
O
CH2
O
OH
HO OH
O
O
OH
OH
OH
OH
MS2 -324
Rebaudioside EO
HO
HO
OH
HO
O
OH
HO
HO O
HO
O
CH3
H3C
O
CH2
O
OH
O OH
OO
O
HO
OH OH
OH
OH
OH
OH
HO
MS2 -162
MS3 -162
MS4 -162
Rebaudioside A
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 38
Jacobs University Bremen
Figure 16.Tandem MS spectra of Rebaudioside A (above) and Rebaudioside E (below) in
negative ion mode.
Tandem mass spectra for rebaudioside D is presented in Figure 17. The structure losses one
glucose unit ([M-H2O]- m/z 162) in MS
2 resulting with the [M-H
+-glc]
- ion with m/z 787 and in
MS3 [M-H
+-2glc]
- with m/z 625. In MS
4 rebaudioside D loses one rhamnose sugar unit ([M-
H2O]- m/z 146) resulting with the [M-H
+-2glc - rham]
- with m/z 479 ion in negative mode.
Further tandem MS data of steviol glycosides are presented in appendix A.
Figure 17.Tandem MS spectra of Rebaudioside D in negative ion mode.
949.6 985.6
-MS
787.4 -MS2
479.1
625.2 -MS3
317.1 479.1
-MS4
0
1 7 x10
Intens.
0.0
0.5
0 2 4 6 x10
0
1 6 x10
200 400 600 800 1000 m/z
322.9 479.1 803.4
641.2 -MS2
317.0
479.1 -MS3
317.0 -MS4
0
2
5 x10 Inten
s.
0.0 0.5
1.0
0
1
2
4
200 400 600 800 1000 m/z
965.6 1001.6
-MS,
803.4 -MS2
317.1 413.1 479.1
641.2 -MS3
317.1
479.1 -MS4
0 1 2
7 x10
Intens
.
0.0 0.5
1.0 1.5
0.0
0.5
0
1
200 400 600 800 1000 m/z
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 39
Jacobs University Bremen
4.3.2. Method Validation
The validation procedure involved determination of limit of quantitation (LOQ), determination
of linear range for quantitation, repeatability studies including multiple injection and day to day
repeatability. Additionally, inter sample repeatability experiments were carried out.
Sensitivity & Selectivity
Retention times of the reference compounds of rebaudioside A and stevioside were compared to
the chromatograms obtained from the leaf extracts, apart from that, the accurate masses were
obtained for each peak in the chromatogram from HR-MS measurements.
The calibration curve was linear in the range of 10 – 500 μg/mL for stevioside and 5 – 500
μg/mL for rebaudioside A and steviolbioside. The equations of calibration curves obtained by the
least square method were as follows;
Stevioside: y = 35999x - 845055 R² = 0.9925
Rebaudioside A: y = 39360x + 287715 R² = 0.9935
Steviolbioside: y= 246929x –100000 R2 = 0.9985
where y is the peak area from the LC chromatogram and x is the μg/mL for rebaudiosideA and
stevioside.
Precision
Precision was calculated based on intra and inter-day (n=3) repeatability. Standard solution of
rebaudioside A at the concentrations of 5, 10, 100 and 500 μg/mL and 10, 50, 300 and 350
μg/mL for stevioside were measured on three different days on the HILIC column and the results
were evaluated by calculating the %RSD. The repeatability of the inter-day measurements was in
the range of 4.1 - 6.7 % for rebaudioside A and 1.9 – 5.4 % for stevioside.
Intra-day measurements were evaluated by calculating the %RSD of three injections of each
concentration of 50, 200 and 300 μg/mL of rebaudioside A and 50, 250, 350 μg/mL of
stevioside. Intra-day precision was in the range of 2.4 – 5.6 % for rebaudioside A and 0.9 – 3.7
% for stevioside.
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 40
Jacobs University Bremen
Limit of detection was determined as the concentration of the component providing signal to
noise ratio (S/N) of three and for limit of quantification, the concentration resulting in a S/N of
10. According to that, for stevioside LOD was obtained as 2.5 ppm and LOQ as 10 ppm. For
rebaudioside A, LOD was 5 ppm and LOQ was obtained as 10 ppm. For steviol glycoside LOD
was 2 ppm and LOQ was 5 ppm.
Accuracy
The accuracy of the method was determined by calculating the relative error observed in a
standard addition experiment. Stevia leaf extracts were spiked with different amounts of
stevioside and rebaudioside A separately. The relative errors was in the range of 0.043 – 0.074
μg/mL for rebaudioside A and 0.056 – 0.14 μg/mL for stevioside
Comparison to UV data quantification
Calibration curves were also obtained from UV measurements (210 nm) for stevioside. The
sensitivity of the method was less compared to results of LC-MS using EIC. The linearity range
was 50 – 500 μg/mL. S/N of 2:1 was achieved with concentration of 75 ppm. The equation of
calibration curve obtained by the least square method was y = 2.8799x – 90.019 (R2 = 0.9972).
The %RSD of triple injections of 50, 250, 350 μg/mL of the standard was respectively, 19.1, 9.9,
and 3.4.
4.3.3. Comparison to SPE sample clean up
Conventional methods for steviol glycoside quantification frequently employ SPE sample
pretreatment followed by UV based quantification. Two problems might arise here, which have
never been addressed: Firstly, do after SPE treatment analytes co-elute in the steviol glycosides
not observed but co-quantified by UV and secondly, do SPE materials retain steviol glycosides.
Two different SPE stationary phases reported in the literature were tested for presample
treatment of stevia extracts prior to HPLC-MS analysis. SPE with material I was resulting with
decrease in the peaks in the LC chromatogram especially with rebaudioside A, with the increased
amount of material I in the cartridge (Figure 18). SPE cleanup with material II did not have a
dramatic effect on the HPLC analysis with both amino and HILIC columns and gave the similar
results if compared with material I based procedure (Figure 19).
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 41
Jacobs University Bremen
Consequently, great care is required when using SPE based protocols since retention of
rebaudioside A, the most lipophilic steviol glycoside, might lead to non-satisfactory accuracy.
Figure 18.Total ion chromatograms for comparison of different amounts of material I in SPE
cleanup procedure.
Figure 19.Total ion chromatograms for comparison of SPE cleanup of the stevia extract with
materials I and II cartridges.
0.2 g Material I
0.6 g Material I
stevioside
Rebaudioside C Rebaudioside A
0.2 g Material II
0.2 g Material I
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 42
Jacobs University Bremen
Additionally, it was shown that for both SPE materials I and II if using identical HPLC-
conditions to those reported in the literature using an MS detector co-elution of several analytes
with Rebaudioside A, E and Dulcoside were observed (appendix B, page 122). The nature of the
co-eluting analytes remains largely unknown, however, it can be anticipated that some of them
display UV absorption at 210 nm, again leading to non-satisfactory accuracy. These critical
assessments of SPE based LC-UV quantification methods clearly show that this type of method
require urgent improvement and amendments and are inferior to LC-MS based methods without
SPE sample pretreatment.
4.3.4. Quantification of Steviol Glycosides
Steviol glycoside levels were quantified in all 166 samples made available within the project.
Quantification of steviol glycosides were performed using the validated method with HILIC
column. For three selected steviol glycosides (rebaudioside A, stevioside and steviol), calibration
curves were obtained using six-point calibration from the extracted ion chromatogram (EIC) of
LC-TOF measurement. The quantities of other steviol glycosides were calculated relatively
according to stevioside values for each sample. Please refer to section 4.3.2 for the calibration
curve data.
Data analysis reveals that there are distinct differences between the steviol glycoside profile in
Stevia rebaudiana leaves in the seven different varieties analyzed and distinct differences
between Stevia rebaudiana leaves from different origins. The average amounts and range
(min&max values) of quantity of each steviol glycoside in all 166 samples is presented in Table
3. Detailed quantification data can be found in appendix B.
Table 3.Steviol glycosides values from 166 samples
Steviol glycoside Average (mg/100 g leaves) Range (mg/100 g leaves)
Rebaudioside A 1.017 79 - 5336
Stevioside 6071 252 - 17509
Dulcoside A 239 5 - 680
Rubusoside 111 5 - 459
Rebaudioside C 282 26 - 820
Total 9036 554 - 18067
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 43
Jacobs University Bremen
Additionally, from the obtained quantification data, the samples were analyzed on a radar plot
based on their variety and origin (Figure 20 and 21). Within the project seven defined botanical
varieties of Stevia rebaudiana were cultivated and their steviol glycoside profile was determined.
The distinct differences in the stevioside profile of different variety of stevia leaves can be easily
recognized. From the radar plot (Figure 20), it can be seen that variety 5 and non-EU samples
have the maximum value of stevioside, whereas variety 7 having the minimum value for
stevioside but maximum value for rebaudioside A.
Stevia rebaudiana was within this project cultivated in nine different locations within the EU.
Additionally samples from outside the EU were available for comparison. In Figure 21, EU
cultivated Stevia rebaudiana shows higher concentrations of stevioside (Amiflikeia and
Argentinie having the maximum end values). However for rebaudioside A, Conaga cultivated
stevia shows higher end concentration values.
Figure 20.Radar plot of steviol glycoside concentrations varying between seven varieties
(average values taken within +/- 3 σ) and in comparison to non-EU samples. Concentrations are
given on radial axis in g/100g dry leaf material. Outer numbers are indicating the 7 varieties and
non-EU samples; numbers inside the plots are indicating the concentrations.
0.000
2.000
4.000
6.000
8.000
10.000 1
2
3
4
5
6
7
Non-EU
RebA
Stevioside
Dulcoside A
Rubusoside
RebC
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 44
Jacobs University Bremen
Figure 21.Radar plot of steviol glycoside concentrations varying between all origins (average
values taken within +/- 3 σ) and in comparison to non-EU samples. Concentrations are given on
radial axis in g/100g dry leaf material.
4.4. Conclusion
In conclusion, steviol glycosides were successfully analyzed and quantified using LC-MS
technique. Stevia leaves can be analyzed using a variety of different LC-MS methods. While LC-
MS on a C18 column allows analysis of steviol glycosides, however suffering from selectivity
problems, analysis on a HILIC column coupled to ESI-TOF detection allows separation and
quantification of all known naturally occurring steviol glycosides. Linear range, sensitivity and
reproducibility were excellent. Using both high resolution MS and tandem MS on an ion trap
instrument reliable structure confirmation can be carried out based on characteristic MSn
fragment spectra of all steviol glycosides.
Distinct differences in the quantity of steviol glycosides within the stevia samples were observed.
From the data it was observed that variety 7 is having the maximum concentration for
rebaudioside A but minimum value for stevioside concrentration. EU origin cultivated stevia
samples have the maximum concentration for stevioside but average value for rebaudioside A
concentration.
0.000
2.000
4.000
6.000
8.000
10.000
12.000 TCV
Uconor
Agrinion
Toumpa
Portugal
Amfilia
Argentinie
Granada
Turkei
Amiflikeia
APTTB
Conaga
RebA
Stevioside
Dulcoside A
Rubusoside
RebC
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 45
Jacobs University Bremen
5. POLYPHENOLS in STEVIA REBAUDIANA
5.1. Overview
The hydroxycinnamate derivatives of S. rebaudiana have been investigated qualitatively and
quantitatively by LC-MSn. Chlorogenic acids and flavonoid glycosides of Stevia rebaudiana
from different origins in all around Europe with seven different botanical varieties were profiled
and quantified. The correlation study between CGAs, differences between stevia samples and
effect of origin and variety on the CGA profile were tested statistically from the obtained dataset.
5.2. Materials & Methods
5.2.1. Sample Preparation
Two grams of S. rebaudiana leaves was immersed in liquid nitrogen, ground in a hammer mill,
and extracted first with 150 mL of chloroform in a Soxhlet apparatus (Buchi B-811 extraction
system) for 2 h for removal of lipid fraction and then with 150 mL of methanol for another 2 h.
Solvents were removed from the methanolic extract in vacuo, and extracts were stored at - 20 oC
until required.
5.2.2. LC-MS Analysis of Polyphenols
LC-TOF MS: The LC equipment (Agilent 1100 series, Bremen, Germany) comprised a binary
pump, an autosampler with a 100 μL loop, and a diode array detector with a light-pipe flow cell
(recording at 254 nm and scanning from 200 to 600 nm). This was interfaced with a MicroTOF
Focus mass spectrometer (Bruker Daltonics) fitted with an ESI source. The MS parameters were:
nebulizer 1.6 bar, dry gas 12.0 L/min, dry temperature 220 0C. The MicroTOF was operated in
negative ion mode and the mass range was 150 – 1200 m/z. Internal calibration was achieved
with 10 mL of 0.1 mol/L sodium formate solution injected through a six-port valve prior to each
chromatographic run. Calibration was carried out using the enhanced quadratic calibration mode.
LC-MSn: The LC equipment (Agilent 1100 series, Bremen, Germany) comprised a binary
pump, an autosampler with a 100 μL loop, and a diode array detector with a light-pipe flow cell
(recording at 254 nm and scanning from 200 to 600 nm). This was interfaced with an ion-trap
mass spectrometer fitted with an ESI source (Bruker Daltonics HCT Ultra, Bremen, Germany)
operating in Auto-MSn mode to obtain fragment ions m/z. Tandem mass spectra were acquired in
Auto-MSn mode (smart fragmentation) using a ramping of the collision energy. Maximum
fragmentation amplitude was set to 1 V, starting at 30% and ending at 200%. MS operating
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 46
Jacobs University Bremen
conditions (negative mode) were capillary temperature of 365 oC, a dry gas flow rate of 10
L/min, and a nebulizer pressure of 50 psi.
HPLC: Separation was achieved on a 250 x 3 mm C18 column (Varian Pursuit XRS) with 5 μm
particle size. Solvent A was water/formic acid (1000+0.005 v/v), and solvent B was acetonitrile
(ACN). Solvents were delivered at a total flow rate of 0.5 mL/min and the column temperature
was set to 25 oC. 5 μL of samples were injected in to LC-MS system, unless stated otherwise.
The gradient profile was 10 to 80% B in 60 min and a return to 10% B at 65 min and 5 min
isocratic to re-equilibrate.
5.2.3. Calibration Curve of Standard Compounds
Most abundant chlorogenic acid derivatives (3-CQA, 4-CQA, 5-CQA, 3,5-diCQA, 4,5 diCQA)
and two flavonoid glycosides (quercetin-3-glycoside and kaempferol-7-glycoside) were chosen
for calibration curves.
Stock solutions of the standard compounds were prepared in 80% ACN/water. A series of
standard solutions was injected (5μL) into the LC-MS system. The areas of the peaks of each
standard from extracted ion chromatograms (EIC) were used to make the respective standard
curves.
5.2.4. Hydrolysis of Flavonoid Glycosides
5 mg crude extract was dissolved in 2 ml 2M HCl and heated at 90 0C for 40 min. Sample was
then directly used for LC-MS or diluted with MeOH.
5.2.5. Statistical Analysis
Statistical analyses of the data were performed using IBM SPSS 20. The distributions of the
variables were tested for normality using the Kolmogorov-Smirnov test. Associations between
the variables were investigated using both parametric (Pearson’s correlation) and non-parametric
(Spearman’s correlation) techniques. Results were interpreted using the widely accepted 5%
level of significance.
To test whether there were differences on each chlorogenic acid with respect to its origin or
variety, separate one-way ANOVA analyses was employed, followed by two post-hoc tests:
Fisher’s Least Significant Difference (LSD) as the least conservative test where equal variances
are assumed and Games-Howell test where non-equal variances are assumed for the multiple pair
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 47
Jacobs University Bremen
wise comparison tests. All empirical results were interpreted using the widely accepted 5% level
of significance (p < 0.05).
5.3. Results & Discussion
Stevia extract was analyzed by LC-MSn in the negative ion mode using an ESI ion-trap mass
spectrometer, allowing assignments of compounds to regioisomeric level, and also by HR-MS
using ESI-TOF in negative ion mode connected to LC, that allowed determination of molecular
formulae based on the accurate mass measurements. Molecular formulas were in general
accepted if an error below 5 ppm was experimentally observed, as accepted by all peer reviewed
chemistry journals. In a second experiment tandem MS experiments were carried out and the
observed fragmentation patterns compared to those of authentic reference materials (either
obtained commercially or from our own laboratory). After obtaining multi-dimensional
information of four parameters on chromatographic retention times, UV-spectra (UV-VIS DAD
detector coupled to LC-MS system), HR-MS and tandem-MS data comparison to authentic
reference material did allow compound identification. Peak assignments of CGAs have been
made on the basis of structure diagnostic hierarchical keys previously developed81, 135
.
Chlorogenic acids and flavonoids were quantified using an established reversed phase LC-MS
method on a C-18 column using ESI-TOF-MS in the negative ion mode. All required analytes
showed baseline separation with exception of the pair 3,4- and 4,5 dicaffeoyl quinic acid. A
typical chromatogram using a reversed phase C-18 column of a stevia extract showing polar
polyphenols at early retention times and more lipophilic steviol glycosides at later retention
times is shown in Figure 22. Abbreviations and numbering of CGAs and flavonoids are
presented in Figure 23, Table 4.
In all cases quantitation was carried out using extracted ion chromatograms only. Additionally
flavonoids were quantified using kaempferol-7-glucoside and quercetin-3-glucoside as reference
standards, resulting for flavonoids in relative values rather than absolute values.
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 48
Jacobs University Bremen
Figure 22.Base peak chromatogram in negative ion mode using a C18 column of methanolic
Stevia rebaudiana extract showing phenolics (CGAs, flavonoids and steviol glycoside).
Table 4.Chromatographic and MS data on flavonoid glycosides and CGAs present in stevia
leaves
Compound Number Compound* m/z [M-H+]
-
1 3-caffeoylquinic acid (3CQA) 353
2 5-caffeoylquinic acid (5CQA) 353
3 4-caffeoylquinic acid (4CQA) 353
4 Rutin 609
5 Quercetin-galactoside 463
6 Kaempferol-glucopyranoside
Quercetin-rhamnoside 447
7 Quercetin-fructoside; Luteolin-glucuronide 461
8 Quercetin-pentoside 433
9 kaempferolxylosylglucoside
Naringin 579
10 Apigenin-galactoside 431
11 3,5-di-caffeoylquinic acid (3,5diCQA) 515
12 4,5-dicaffeoylquinic acid(4,5diCQA) 515
13 Quercetin-diglucoside-rhamnoside 771
14 Kaempferol-glucosylrhamnosyl-glucoside/galactoside 755
15
Kaempferol-rhamnopyranosyl-glucopyranoside(rutinoside) isomers
Quercetin-dirhamnoside
Apigenin-diglucoside/galactoside
593
16 Quercetin-trisaccharide 741
17 Kaempferol 3-rhamnopyranosyl-rhamnopyranosyl-glucopyranoside 739
*Compounds named for flavonoid glycosides are only possible structures, which were not identified or confirmed.
0.0
0.5
1.0
1.5
2.0
2.5
6 x10
5 10 15 20 25 30 35 40
Steviol glycosides CGAs & Flavonoid glycosides
6 5 8
9
1 4
7
10
11
12
13
14
2
3
Time [min]
Intens.
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 49
Jacobs University Bremen
Figure 23.Structures of caffeoylquinic acids and flavonoid glycosides.
OH
OH
OOH
HOOC
O OH
O
OHOH
HOOC
O
OH
3
O
OH
OHOH
HOOC
O
OH
OH
5 OH
OH
O
OOH
HOOC
O
OH
OH
O
O
OHOH
HOOC
O
OH
OH
43O
OH
HO
4
O
OH
OH
5
O
OH
OOH
HOOC
O
OH
OH
5
O
OH
OH
3
3-CQA 5-CQA
3,5-diCQA 3,4-diCQA 4,5-diCQA
O
O
OHOH
HOOC
O
OH
OH
4
O
5
HO
HO
cis-4,5-diCQA
4-CQA
HO O
O
OOH
OHOH
OHO
OHOH
OH
Quercetin-3-O-beta-D-glucoside
O
O OO
OH
O
HO
OH
OH
OH
OOH
HO
OH
OHOH
OH
OOH
HO
OH
OOH
HO
Kaempferol
Quercetin Luteolin Apigenin
O O
OH
OOH
OHO
HOOH
OH
Kaempferol-7-O-beta-D-glucoside
OH
OH
OH
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 50
Jacobs University Bremen
5.3.1. Characterization of Chlorogenic acids
Mono-caffeoylquinic acids (mono-CQA) and di-caffeoylquinic acids (di-CQAs) were identified
in stevia using the hierarchial keys previously developed81
. It is possible to discriminate between
the isomers of caffeoylquinic acids and feruloylquinic acids by using LC-MSn. The
fragmentation pattern depends on the stereochemical relationships between the substituents on
the quinic acid moiety. Four peaks were detected at m/z 353.1 and assigned as well-known 3-
CQA, trans-5-CQA, and 4-CQA and cis-5CQA. Three dicaffeoylquinic acid isomers were
identified by their parent ion m/z 515.2 and were assigned as 3,5-diCQA, 3,4-diCQA, and 4,5-
diCQA using the hierarchial keys81, 135
. Two further peaks present as minor components showed
fragmentation patterns similar to that of 4,5-diCQA, which were identified as cis isomers of 4,5-
diCQA. Figures 24 - 26 show selected data including an extracted ion chromatogram for
monocaffeoyl quinic acids and two tandem mass spectra for selected regioisomeric stevia
caffeoyl quinic acids as typical secondary metabolites. All other fragmentation patterns are
provided in appendix C.
Twenty-four hydroxycinnamic acid derivatives of quinic and shikimic acid were detected in the
work using stevia leaves not cultivated in this project and the results have been published in the
course of this project (Please refer to the article attached in appendix for detailed insight).
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 51
Jacobs University Bremen
OH
HOOC
O
OH
OH
OH
OH
O
1 3
OH
HOOC
OH
OH
O
1
O
OH
5
OH
HOOC
OH
O
OH
OH
O
OH
HO
Figure 24.Extracted ion chromatogram of m/z 353 of three mono-caffeoylquinic acids, from left
to right: 3-Caffeoylquinic acid (1), 5-Caffeoylquinic acid (2) and 4-Caffeoylquinic acid (3) in
negative ion mode.
All monoacyl CGA (m/z ~ 353) gave the expected parent ion (monoacyl CGA - H+) (Figure 25
and 26) in tandem MS analysis. 3-CQA and 5-CQA produce an MS2 base peak at m/z ~ 191
corresponding to [quinic acid-H+]- (Q1 ion) and in MS
3 it fragments to Q2 ion with m/z 85.1 and
[quinic acid – H2O - H+]- (Q3 ion) at m/z 172.8 (Figure 25). 3-CQA might be discriminated by its
MS2 peak at m/z ~135 and by slight difference of the intensity of m/z ~178 in MS
2. It is easy to
distinguish the 4-substititued CGA by it is dehydrated quinic acid moiety which gives MS2 base
peak at m/z ~173.
EIC 353.0 -All MS
0.0
0.5
1.0
1.5
2.0
8 Inte
ns.
0 10 20 30 40 50
1
2
3
Cis-2
x10
Time
[min]
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 52
Jacobs University Bremen
Figure 25.Consecutively MS, MS2and MS
3 spectra of 3-Caffeoylquinic acid in negative ion
mode.
Figure 26.Consecutively MS, MS2and MS
3 spectra of 4-Caffeoylquinic acid in negative ion
mode.
The diacyl CGAs behave similarly, giving the parent ion [diacyl CGA – H+]- at m/z 515.
DiCQAs loses the caffeic acid moiety in MS2 yielding a [diacyl CGA – cinnamoyl – H
+]
- and
these ions are identical to the parent ions obtained from monoCQAs. 4,5-diCQA give dehydrated
quinic acid moiety as base peak at m/z ~173 in MS3 as previously seen for 4-CQA. This ion was
not observed for 3,5-diCQA, instead MS3 base peak was at m/z 191 as previously observed for 3-
CQA (Figure 27 and 28). Thus, 3,5-diCQA can be distinguished easily from the 4-acylated
caffeoylquinic acid isomers.
472.8
352.9
374.9
-MS
172.7 -MS2(352.9)
71.3 154.7
93.0 -MS3(353.1->172.7)
0 1 2 3 7 x10
Intens.
0 2 4 6
6 x10
0.0
0.5
1.0
5 x10
200 400 600 800 1000 m/z
472.9 729.2
352.9 -MS
134.8
190.7 -MS2(352.9)
172.8 85.1 -MS3(353.1->190.6)
0 1 2 3 7 x10
Intens.
0.0
0.5
1.0
7 x10
0
2
4
6 4 x10
200 400 600 800 1000 m/z
OH
HOOC
OH
OH
O
OH
OH
O
Q1
OH
HOOC
OH
O
Q2 Q3
CH
CHHO
HO Caffeic m/z 135
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 53
Jacobs University Bremen
Figure 27.Consecutively MS, MS2and MS
3 MS
4 spectra of 3,5-dicaffeoylquinic acid in negative
ion mode.
Figure 28.Consecutively MS, MS2and MS
3 MS
4 spectra of 4,5-dicaffeoylquinic acid in negative
ion mode.
1031.4
515.0 -MS
172.7 254.8 298.9
352.9 -MS2(515.0)
134.8
172.7 -MS3(515.3->352.9)
93.0
0 1
8 x10 Intens.
0.0
0.5
8 x10
0
1
7 x10
0 2 4 5 x10
200 400 600 800 1000 m/z
352.9 613.0
515.0
1031.3
-MS
190.7
352.9 -MS2(515.0)
134.7
190.7 -MS3(515.3->352.9)
85.1 126.8
172.7
-MS4(515.3->353.1->190.7)
0
1 8 x10
Intens.
0.0
0.5
8 x10
0
2
7 x10
0 1 2 5 x10
200 400 600 800 1000 m/z
-MS4(515.3->353.1->172.8)
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 54
Jacobs University Bremen
5.3.2. Characterization of Flavonoid Glycosides
Flavonoid glycosides are an important group of plant natural products in which different type of
sugars are linked to an aglycone. Determination of the identity and position of linkage to the
aglycone of sugars by mass spectrometry alone is challenging. From the mass spectra of
flavonoid glycosides using tandem MS we can obtain molecular mass, structure of the aglycone
(from its m/z), number of sugar rings and their configuration. Negative ion mode was selected for
the analyses because previous results suggested that negative mode was more sensitive than
positive ion mode.
A total of twelve peaks in the chromatogram corresponding to flavonoid glycosides were
identified. All compounds could be identified as belonging to this class of compounds due to
their characteristic fragmentation patterns in tandem MS showing neutral losses of sugar
moieties following by characteristic fragment spectra of the aglycones. The nature of the
aglycones was further substantiated by hydrolysis of the total phenol extract followed by LC-MS
analysis revealing that four flavonoid aglycones quercetin, kaempferol, luteolin and apigenin
(Figure 29 and Table 4) are present in Stevia rebaudiana leaves.136
O O O O
OH
O
HO
OH
OH OH
OOH
HO
OH
OH
OH
OH
OOH
HO
OH
OOH
HO
KaempferolMw 286
QuercetinMw 302
LuteolinMw 286
ApigeninMw 270
Figure 29.Chemical structure of four flavonoid aglycones identified in Stevia rebaudiana leaves.
Although the aglycone and the glycane were identified for an observed m/z, the accurate
structure of the flavonoids glycoside could not be determined because identity and the site of
connection of monosaccharide cannot be determined by LC-MS. The possible structures of
compounds were determined by comparison of the mass spectral data obtained with literature
data. However, the detailed chemical structure of these flavonoids could not be established
unambiguously by LC-MS. None of the compounds present were shown to be identical to any of
the twelve reference standards used commercially or to reference compounds available in our
laboratory. A preparative LC isolation and full structure elucidation of the compounds was
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 55
Jacobs University Bremen
outside the scope and timeline of the project but needs further attention. A table showing the
chromatographic and MS characteristics of the flavonoids was given previously in Table 4 and
extracted ion chromatogram and tandem mass spectra for selected compound is shown in Figure
30 and 31. All other tandem mass spectral data of flavonoid glycosides are presented in appendix
C.
The possible fragmentation pattern and ion nomenclature of flavonoid glycosides is illustrated on
luteolin-7-O-rutinoside in Figure 32.
Figure 30.Extracted ion chromatogram of m/z 447.0 in negative ion mode.
Figure 31.An example of tandem MS spectra for compound 1, revealing its identity as
kaempferol glucopyranoside.
447.0 -MS
284.8 -MS2(447.0)
174.7 216.7
-MS3(447.3->286.8)
0 50
100 Intens.
[%]
0 50
100
0 50
100
100 200 300 400 500 600 700 m/z
1
EIC 447.0 -All MS
0.00
0.25
0.50
0.75
1.00
1.25 8 x10 Intens.
5 1
0 1
5 2
0 2
5 3
0
1
3
2
Time
[min]
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 56
Jacobs University Bremen
O
OH
OO
OH
OH
OH
CH2OH
O
OH
OH
Figure 32.Fragmentation illustration on luteolin-7-glucoside137
The most useful fragmentation for flavonoid aglycone identification are the cleavage of two C—
C bonds of the C-ring, resulting in structural information for A and B ions (Figure 32). These
ions can be rationalized by retro-Diels–Alder (RDA) reactions and are the most diagnostic
fragments for flavonoid identification since they provide information on the number and type of
substituents in the A- and B-rings. The flavonoid aglycone fragment ions can be designated
according to the nomenclature proposed by Ma et al 138, 139
.
5.3.3. Quantification of Chlorogenic acids & Flavonoid Glycosides
Phenolics including chlorogenic acids and flavonoids were quantified using an established
reversed phase LC-MS method on a C-18 column using ESI-TOF-MS in the negative ion mode.
Standard solutions were analyzed using the same chromatographic method as used for stevia leaf
extracts as indicated before (chapter 5.2.2). The calibration curves were obtained by the external
standard method on six levels of concentration of reference compounds. For quantitation the six
most abundant chlorogenic acid derivatives (3-CQA, 4-CQA, 5-CQA. 3,4-diCQA, 3,5-diCQA,
4,5 diCQA) were chosen and calibration curves were obtained with excellent linearities. In all
cases quantitation was carried out using extracted ion chromatograms only. Cis isomers of 5-
CQA and 4,5-diCQA were quantified based on the corresponding calibration curves of trans
isomers, resulting in relative values of cis isomers (please refer to appendix D for quantification
data, page 140). Additionally flavonoids were quantified using kaempferol-7-glycoside (k7g) and
B0
A0
Z0
X0
A1
Y0
B1
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 57
Jacobs University Bremen
quercetin-3-glycoside (q3g) as reference standards, resulting for flavonoids in relative values
rather than absolute values (appendix D).
The linearity and the equations of calibration curves obtained by the least square method were as
follows:
Reference compound Linearity (μg/mL) Equation R2
3-CQA 1-100 82912x+346549 0.99
5-CQA 5-500 49062x+907693 0.99
4-CQA 1-150 129889x+395550 0.99
3,5-diCQA 10-200 138064x+7000000 0.98
4,5-diCQA 10-400 138243x+4000000 0.98
K7g 50-1000 26105x+6000000 0.98
Q3g 2-150 97003x+2000000 0.98
Where y is the peak area from the LC chromatogram and x is the μg/mL for the reference
compound.
5.3.3.1. Sample Variation
From the obtained quantification data, a series of statistical analysis was carried out. As a first
step, for each variety, origin and harvest average values and standard deviations were
determined. Additionally, minimum and maximum values for each sample subgroups are given
in the tables.
Within the project seven defined botanical varieties of Stevia rebaudiana were cultivated and
their phenolic profile was determined. From the data variations between different batches and
average values averaged over all samples from a single variety can be compared. Additionally
variations in single compound quantities, quantities of groups of compounds (e.g. mono-caffeoyl
quinic acids, dicaffeoyl quinic acids) or ratios of two single compounds can be compared.
From the data for example it can be seen that the average concentration of all monocaffeoyl
quinic acids remains rather constant over all varieties (2.123 - 2.686 g/100g), whereas a more
spread of data is observed for dicaffeoyl quinic acids (1.484 – 2.432 g/100g). Varieties 5, 6, 7
and 3 show on average increased levels of chlorogenic acids compared to varieties 2. Average
values are given in Table 5, for detailed quantification data please refer to appendix D. All EU
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 58
Jacobs University Bremen
Stevia rebaudiana varieties show considerable higher levels of chlorogenic acids and flavonoids
if compared to some samples obtained from outside the EU.
Variations can be displayed in a radar plot shown in Figure 33, 34, 35. Here the lower
chlorogenic acid content in non-EU samples as well as in variety 2 can be appreciated.
Table 5.Average values (taken within +/- 3 σ) for chlorogenic acids in seven different varities
Variety
3CQA
(g/100g
leaves)
4CQA
(g/100g
leaves)
5CQA
(g/100g
leaves)
Total mono
(g/100g
leaves)
3,5-diCQA
(g/100g
leaves)
4,5-diCQA
(g/100g
leaves)
TotaldiCQA
(g/100g
leaves)
1 0.279 0.096 1.921 2.262 0.925 1.159 1.969
2 0.267 0.094 1.804 2.208 0.880 1.208 1.807
3 0.290 0.114 2.267 2.650 1.101 1.213 2.022
4 0.294 0.118 2.258 2.616 1.101 1.259 2.020
5 0.279 0.105 2.108 2.466 1.365 1.313 2.432
6 0.261 0.105 2.100 2.686 1.309 1.396 2.272
7 0.178 0.113 2.262 2.507 1.277 1.488 2.108
Non-EU 0.195 0.092 1.836 2.123 0.881 0.700 1.484
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 59
Jacobs University Bremen
Figure 33.Radar plot of individual chlorogenic acid concentrations varying between seven
varieties (average values taken within +/- 3 σ) and in comparison to non-EU samples.
Concentrations are given on radial axis in g/100g dry leaf material. Outer numbers indicating the
7 varieties and non-EU samples; numbers inside the plot are indicating the average
concentrations of individual chlorogenic acids.
Figure 34.Radar plot of total mono- and di-acyl quinic acids (chlorogenic acids) concentrations
varying between seven varieties (average values taken within +/- 3 σ) and in comparison to non-
EU samples. Concentrations are given on radial axis in g/100g dry leaf material. Outer numbers
indicating the 7 varieties and non-EU samples; numbers inside the plots are indicating the
concentrations.
0.000
0.500
1.000
1.500
2.000
2.500 1
2
3
4
5
6
7
Non-EU
3-CQA
4-CQA
5-CQA
3,5 diCQA
4,5-diCQA
0
0.5
1
1.5
2
2.5
3 1
2
3
4
5
6
7
Non-EU
monoCQA
diCQA
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 60
Jacobs University Bremen
Figure35. Bar plot of total mono- and di-acyl quinic acids (chlorogenic acids) concentrations
varying between seven varieties (average values taken within +/- 3 σ) and in comparison to non-
EU samples. Concentrations are given on radial axis in g/100g dry leaf material.
Variations by Origin
Stevia is capable of growing almost anywhere under poor soil conditions, in particular where
tobacco used to grown. There are only few plants possessing this feature. Therefore, stevia can
serve as an alternative crop to the tobacco farmers and discourage tobacco cultivation inside EU.
Stevia rebaudiana was within this project cultivated in nine different locations (Figure 36) within
the EU (e.g. Conaga,TCV, Italy; Granada, Spain; and Toumpa, Agrinio, Amfikleia, Greece).
Additionally samples from outside the EU were available for comparison (e.g. Paraguay,
Argenitine). Stevia leaves obtained from these origins were analyzed for studying the effect of
growth conditions (e.g. sun, soil, and climate) on the metabolite profile.
According to the literature polyphenol concentrations are due to their physiological function as
UV protection agents a direct function of growth altitude and climatic conditions, in particular
sunshine hours. Accordingly variations of chlorogenic acid concentrations between different
origins should be expected.
0.000 0.500 1.000 1.500 2.000 2.500 3.000
1
2
3
4
5
6
7
Non-EU
di CQA
mono CQA
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 61
Jacobs University Bremen
Figure 36.Map showing the origins of stevia cultivation within the project.
Indeed the data reveal significant variations in CGA concentrations varying from 3.090 -1.637
g/100g for total monocaffeoyl quinic acids and 2.890 - 1.144 g/100g for dicaffeoyl quinic acids.
EU cultivated Stevia rebaudiana shows concentrations of CGAs nicely sandwiched between
extreme values at both ends observed in samples from outside the EU (e.g. highest for
Argentinian samples with an average value of 2.890 g/100g dicaffeoyl quinic acids and APTTB
and Portugal samples with a lowest average value of 1.448 g/100g and 1.144 g/100g
respectively). Again a radar plot shown in Figure 37 was used to display variations between
different origins. Average values are given in Table 6.
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 62
Jacobs University Bremen
Figure 37.Radar plot of total mono- and di-acyl quinic acids (chlorogenic acids) concentrations
varying between all origins (average values taken within +/- 3 σ) and in comparison to non-EU
samples. Concentrations are given on radial axis in g/100g dry leaf material.
A bar plot shown in Figure 38 allows further direct comparison between samples of different
origins.
Figure 38.Bar plot of total mono- and di-acyl quinic acids (chlorogenic acids) concentrations
varying between all origins (average values taken within +/- 3 σ) and in comparison to non-EU
samples. Concentrations are given on radial axis in g/100g dry leaf material.
0.000
1.000
2.000
3.000
4.000 TCV
Uconor
Agrinion
Toumpa
Portugal
Amfilia
Argentinie
Granada
Turkei
Amiflikeia
APTTB
Conaga
mono CQA
di CQA
0.000 0.500 1.000 1.500 2.000 2.500 3.000 3.500
TCV
Uconor
Agrinion
Toumpa
Portugal
Amfilia
Argentinie
Granada
Turkei
Amiflikeia
APTTB
Conaga
di CQA
mono CQA
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 63
Jacobs University Bremen
Table 6.Average values (taken within +/- 3 σ) for chlorogenic acids between origins
Variation between harvests
Stevia rebaudiana was harvested three times within this project. From the data for example it can
be seen that the average concentration of all dicaffeoyl quinic acids and monocaffeoyl quinic
acids remains rather constant over first and second harvests, whereas a significant decrease can
be observed in the third harvests for monocaffeoyl quinic acids. Overall, for individual
compounds and for total amounts, second harvests are having the highest values 2.595g /100g for
monocaffeoyl quinic acids and third harvests are for dicaffeoylquinic acids 2.242 g/100 g (Table
7). However, decrease in the quantities of CGAs with the later harvest could be observed if the
three harvests are compared while the origin and the variety of stevia are kept constant (Table 8).
In particular mono-CQAs are showing higher decreases from harvest I to harvest III, whereas
diCQAs values are rather constant within chosen three origins with same variety presented in
Table 8.
Origin 3CQA
(g/100g leaves)
5CQA
(g/100g leaves)
4CQA
(g/100g leaves)
Total mono
(g/100g leaves)
3,5-diCQA
(g/100g leaves)
4,5-diCQA
(g/100g leaves)
TotaldiCQA
(g/100g leaves)
TCV 0.224 2.015 0.101 2.336 0.909 0.987 1.757
Uconor 0.323 2.060 0.119 2.575 1.373 1.480 2.079
Agrinion 0.278 2.078 0.104 2.604 1.233 1.339 2.460
Toumpa 0.269 2.450 0.123 2.780 1.351 1.255 2.471
Portugal 0.202 1.580 0.079 1.861 0.555 0.589 1.144
Amfilia 0.342 2.208 0.110 2.674 0.898 1.043 1.941
Argentinie 0.284 2.446 0.122 2.883 1.435 1.520 2.890
Granada 0.197 1.855 0.093 2.145 0.686 0.703 1.389
Turkei 0.104 1.837 0.092 2.093 0.727 0.859 1.586
Amiflikeia 0.441 2.646 0.133 3.090 1.302 1.254 2.335
APTTB 0.206 1.363 0.068 1.637 0.746 0.703 1.448
Conaga 0.310 2.525 0.126 2.956 0.993 1.410 2.328
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 64
Jacobs University Bremen
Table 7.Average values (taken within +/- 3 σ) for chlorogenic acids between harvests
Harvest
3CQA
(g/100g
leaves)
4CQA
(g/100g
leaves)
5CQA
(g/100g
leaves)
Total
monoCQA
(g/100g
leaves)
3,5-diCQA
(g/100g
leaves)
4,5-diCQA
(g/100g
leaves)
TotaldiCQA
(g/100g
leaves)
I
0.306
0.110 2.096 2.562 1.004 1.164 1.967
II
0.276
0.113 2.251 2.595 1.209 1.182 2.142
III
0.145
0.091 1.828 2.064 1.226 1.155 2.242
Table 8.Comparison of average values (taken within +/- 3 σ) for chlorogenic acids between three
harvests of same variety and origin
Sample no. Origin Variety Harvest 3CQA 5CQA 4CQA Totalmono 3,5diCQA 4,5diCQA TotaldiCQA
71 Agrinion 3 I 0.413 3.266 0.163 3.843 1.008 1.282 2.290
123 Agrinion 3 II 0.247 2.626 0.131 3.004 1.570 1.342 2.913
162 Agrinion 3 III 0.168 1.372 0.069 1.608 0.974 0.938 1.912
98 Toumpa 3 I 0.491 4.099 0.205 4.795 1.208 1.535 2.743
127 Toumpa 3 II 0.175 2.490 0.124 2.789 1.650 1.271 2.921
126 Toumpa 3 III 0.187 1.961 0.098 2.246 1.684 1.567 3.251
137 TCV 3 I 0.441 3.153 0.158 3.752 1.578 1.266 2.844
139 TCV 3 II 0.267 2.692 0.135 3.094 1.854 1.155 3.010
161 TCV 3 III 0.161 1.777 0.089 2.026 0.987 1.077 2.064
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 65
Jacobs University Bremen
5.3.3.2. Flavonoid quantification
A selection of the two major flavonoids identified within the chromatogram was quantified as
their glycosylated derivatives using calibration curves for closely related compounds quercetin-
3-glycoside and kaempferol-7-glycoside for all 166 samples. Additionally, for ten selected
samples a flavonoid hydrolysis was carried out using acidic methanol with subsequent
quantification of the four flavonoid aglycones quercetin, apigenin, luteolin and kaempferol
(please refer to chapter 5.3.2 for identification). Relative values for flavonoids based on LC-MS
data are contained within Tables 9 and 10. Data for the flavonoid hydrolysis are given in Table
11.
Table 9.Flavonoid glycosides average values for two major flavonoids in samples between
origins determined by LC-MS directly from extracts without hydrolysis. Reference compounds
used were quercetin-3-glucoside and kaempferol-7-glucoside. Total flavone value designates
addition of all intensities of all flavonoid signal in chromatogram referenced to kaempferol-7-
glucoside.
Origin
Kaempferol-7-glucoside
(g/100g leaves)
Quercetin-3-glucoside
(g/100g leaves)
Total flavones
(g/100g leaves)
TCV 1.653 0.089 3.630
Uconor 2.762 0.103 6.397
Agrinion 3.431 0.047 7.865
Toumpa 3.602 0.063 8.170
Portugal 3.792 0.051 9.122
Amfilia 4.001 0.073 8.142
Argentinie 4.461 0.053 9.698
Granada 3.094 0.062 7.491
Turkei 2.885 0.088 7.444
Amiflikeia 2.956 0.066 6.984
APTTB 3.782 0.052 9.730
Conaga 2.506 0.101 5.726
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 66
Jacobs University Bremen
Table 10.Flavonoid glycosides average values between varieties
Table 11.Values for flavonoids quercetin, kaempferol, luteolin and apigenin determined after
hydrolysis of total polyphenol fraction using HCl/MeOH, determined by LC-MS. n.d.
indicatesthat value was outside linear range of method140
.
Sample
no Origin Variety Harvest Year
Kaempferol
(mg/100g)
Quercetin
(mg/100g)
Luteolin
(mg/100g)
Apigenin
(mg/100g)
21 TCV 1 II 28.09.2010 82.5 75 12 4
32 Uconor 2 I 11.08.2010 n.d 78 8 3
41 Uconor 3 I 11.08.2010 70 638 9 n.d.
47 Toumpa 3 II 10.09.2010 108 463 n.d. 5
61 Amfilia 3 II 15.09.2010 90 455 n.d n.d.
87 Uconor 7 I 13.07.2011 190 737.5 14 n.d.
89 Uconor 4 I 2011 65 353 n.d. n.d.
94 Uconor 5 I 13.07.2011 88 417.5 n.d. n.d.
114 Uconor 6 II 2011 n.d 445 7 2
115 Conaga 3 I 2011 73 195 8 4
123 Agrinion 3 II 2011 73 318 n.d. n.d.
Variety
Kaempferol-7-glucoside
(g/100g leaves)
Quercetin-3-glucoside
(g/100g leaves)
Total flavones
(g/100g leaves)
1 2.860 0.073 7.224
2 2.090 0.099 5.401
3 2.955 0.062 7.412
4 3.095 0.104 6.640
5 2.904 0.076 8.083
6 2.850 0.045 7.682
7 2.164 0.155 5.432
Non-EU 3.064 0.097 7.638
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 67
Jacobs University Bremen
If individual figures for flavonoids from LC-MS data are compared to those obtained by
hydrolysis an obvious differences in values is apparent, even if the g/100 g dry leaves values are
corrected for the increased molecular weight of the flavonoids. A direct comparison must be
viewed with great care since firstly the literature hydrolysis method has been validated only for
flavonoid glucosides. However, in the previous identification section we have shown that
compounds present in stevia are hexosides but not glucosides, and can hence show a greatly
reduced hydrolysis kinetic. Secondly, the quantitative data without hydrolysis are based on the
use of reference standards that are chemically different from the compounds present in the leaf.
Again it can be anticipated that ion enhancement effects in LC-MS might contribute to higher
levels of flavonoids determined here. However, the values obtained are highly useful since they
give a detailed insight into relative variations between flavonoids between different varieties and
origins and give a guideline towards absolute and accurate values.
5.3.3.3. Principal Component Analysis (PCA)
A principle component analysis (PCA) based on the LC-MS dataset of stevia phenols was carried
out to allow differentiation between different stevia varieties and geographic origins. Figure 39a
shows a representative analysis with score and loading plots where within the scores plot every
spot corresponds to a single leave sample with clear groupings apparent allowing distinction
between varieties based on their phenolic profile. The loading plot reveals in each data point a
pair of retention time and m/z ratio, therefore defining individual compounds whose quantities
can be used as phytochemical markers for variety distinction.
A second PCA analysis was carried out with an aim to distinguish EU cultivated samples from
non-EU cultivated samples. For this purpose 20 EU and 20 non EU samples were subjected to a
full PCA analysis. Score and loading plots are shown in Figure 39b.
From the score plot it can be seen that the samples fall in three groupings. Group A from South
American samples can clearly be distinguished from all other samples based on their high
diCQA content (from score plot).
A second group B contains exclusively European samples from the Uconor cooperative. The
final group C contains both EU and non-EU samples e.g. from Turkey, Ucraine and India, which
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 68
Jacobs University Bremen
group together. The sample distinction information from the loading plot suggests that a
combination of rebaudioside A concentrations and diCQA concentrations allows distinction here.
Figure 39a.PCA analysis of phenol profile of 35 stevia leaf LC-MS datasets. Score plot is on the
left with each colour representing a different stevia variety and loading plot on the right with
each data point corresponding to individual compounds allowing differentiation.
Figure39b.PCA analysis of phenol profile of 40 stevia leaf LC-MS datasets (red points non-EU
samples, blue points EU samples). Score plot is on the left with each colour representing a
different stevia variety and loading plot on the right with each data point corresponding to
individual compounds allowing differentiation.
-1.0 0.0 1.0 PC 2
-1
0
1
PC 4
-1.0 0.0 1.0 PC 2
-1
0
1
PC 4
-0.5 0.0 0.5 1.0 PC 1
-0.4
-0.2
0.0
0.2
PC 2
-0.5 0.0 0.5 1.0 PC 1
-0.4
-0.2
0.0
0.2
PC 2
A
B
C
3,5 diCQA
Reb A
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 69
Jacobs University Bremen
5.3.4. Statistical Evaluation of Quantification Data of Polyphenols in Stevia
From the obtained data, a series of statistical analysis was carried out. For each variety, origin
and harvest average values and standard deviations were determined. Additionally, the statistical
pattern and type of statistical distribution of the data was analyzed for each subgroup. The
correlation studies were performed by Pearson's correlation, with the significance value of p <
0.05.
Significant differences among stevia leaves for each variable were assessed with analysis of
variance (ANOVA).
5.3.4.1. Statistical Spread of Data
The distribution of the dataset is an essential step for examination of data in statistical analysis.
The most important and useful distribution of data is Gaussian (normal) distribution. A normal
distribution can be easily characterized by observing its symmetrical bell shaped curve on a
histogram (Figure 40). Skewness and kurtosis values (< 1) show also that the data is normally
distributed (Table 12). The Kolmogorov-Smirnov (KS) test was also used for the analysis of data
distribution. In this test, the significance value above 0.05 means the data is normally distributed.
Each mono and di-CQAs as well as total mono and di-CQAs quantities obtained from 166 stevia
samples showed normal distribution as judged by the KS test. In contrast quantitative data for
cis-5 CQA and cis-4,5 diCQA were found to exhibit a non-Gaussian distribution.
The KS test result, mean values, standard deviations, skewness and kurtosis of the curve for each
CGA is represented in Table 12. Histogram of 5-CQA is presented as an example in Figure 40.
Figure 40.Histogram of 5-CQA, showing the normal distribution of the dataset.
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 70
Jacobs University Bremen
Table 12.Descriptive statistics of caffeoylquinic acids
Statistics
3-CQA 5-CQA 4-CQA 3,5-diCQA 4,5-diCQA Total diCQA Total monoCQA
Mean 0.295 2.488 0.124 1.201 1.237 2.435 2.907
Std. Deviation 0.144 0.852 0.426 0.512 0.485 0.937 1.009
Skewness 0.411 0.151 0.151 0.049 0.476 0.111 0.137
Kurtosis -0.217 0.365 0.372 -0.848 0.170 -0.513 0.212
Minimum 0.002 0.193 0.010 0.173 0.210 0.311 0.205
Maximum 0.719 4.986 0.249 2.476 2.609 4.575 5.608
KS test, Asymp sig. (2 tailed) 0.521 0.951 0.916 0.468 0.746 0.584 0.817
5.3.4.2. Correlations
The correlation analysis was performed to measure the degree of relation between two chosen
CGAs. There are several different correlation methods, the most commons are Pearson
(parametric) test, which is used in the case of normally distributed populations and the Spearman
(nonparametric) test, in which there is no requirement for the assumption of normality or
homogeneity of variance. The main result obtained from these tests is the correlation coefficient
(r) and it ranges from -1.0 to +1.0. The closer “r” is to -/+1, the more closely the two variables
are related.
Correlations of CGAs grouped as mono to monoCQA, diCQA to diCQA and as well as
monoCQA to diCQAs were tested according to Pearson correlation. However, correlations of cis
derivatives were tested according to Spearman correlation due to their non-Gaussian data
distribution. Overall, correlation was observed for all CGAs with each other (Table 13).
However, the strong correlation was observed between the mono-CQAs and 3,5-diCQA with
4,5-diCQA (Pearson correlation of 0.762 and 0.791 respectively) and 5-CQA with 4,5-diCQA as
well as 4-CQA with 4,5-diCQA (Pearson correlation of 0.573 for both cases). The rest of the
correlations were slightly weaker. It needs to be pointed out that, all correlation coefficients
closer to +1, indicates that increasing quantity of e.g. 5-CQA leads to quantity increase in the
4,5-diCQA. The square of the correlation coefficient (r) is the percent of the variation in one
variable which is related to the variation in the other. In the case of correlation of 3,5-diCQA
with 4,5-diCQA 62% of the variance is related (or is correlated) (Figure 41).
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 71
Jacobs University Bremen
Table 13.Correlation coefficients of mono and di-CQAs
Correlations
3-CQA 5-CQA 4-CQA 3,5-diCQA 4,5-diCQA
3-CQA
Pearson Correlation (r) 1 0.762** 0.762** 0.416** 0.514**
Sig. (2-tailed) 0.000 0.000 0.000 0.005
R2 0.581 0.581 0.173 0.264
5-CQA
Pearson Correlation (r) 0.762** 1 1.000** 0.532** 0.573**
Sig. (2-tailed) 0.000 - 0.000 0.000 0.000
R2 0.581 - 1 0.283 0.328
4-CQA
Pearson Correlation (r) 0.762** 1.000** 1 0.532** 0.572**
Sig. (2-tailed) 0.000 0.000 - 0.000 0.000
R2 0.581 1 - 0.283 0.327
3,5-diCQA
Pearson Correlation (r) 0.416** 0.532** 0.532** 1 0.791**
Sig. (2-tailed) 0.000 0.000 0.000 - 0.000
R2 0.173 0.283 0.283 - 0.625
4,5-diCQA
Pearson Correlation (r) 0.514** 0.573** 0.572** 0.791** 1
Sig. (2-tailed) 0.000 0.000 0.000 0.000 -
R2 0.264 0.328 0.327 0.625 -
** Correlation is significant at the 0.01 level (2-tailed).
b) a)
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 72
Jacobs University Bremen
Figure 41.Graph showing the correlation between a) 3,5-diCQA/4,5-diCQA b) 3-CQA/5-CQA,
c) 5-CQA/4,5-diCQA and d) 4-CQA/3,5-diCQA.
Furthermore, another study was performed on whether the formation of cis-caffeoylquinic acids
plays a role in agricultural practice and whether they would serve as a useful marker for UV
exposure of plant tissues. Agricultural parameters including climatic conditions were obtained
through records of the nearest official weather station.
To investigate the biosynthetic origin of the cis-CQA derivatives, initially the statistical pattern
and type of statistical distribution of the quantitative data was analyzed. Both 5-CQA and 3,4 di-
CQA concentration show a Gaussian distribution profile over all samples analyzed as judged by
the Kolmogorov Smirnov test (section 5.3.4.1). In contrast quantitative data for cis-5 CQA and
cis-3,4 diCQA were found to exhibit a non-Gaussian distribution.
For a large data set of quantitative data for three mono- and four di-CQA concentrations,
correlation of all chlorogenic acids isomers correlate linearly with each other pointing to a
common stimulus of biosynthetic production in the plant (Figure 41).
The correlation linearity test was applied to cis derivatives of 5-CQA and 4,5-diCQ with their
trans derivatives and there was no linear concentration dependency (Figure 42). This result
indicates towards two distinct stimuli and pathways of their biosynthetic production.
In contrast concentrations of the two cis-4,5-diCQA isomers and concentration of cis-5-CQA and
the cis-4,5 diCQA derivatives show a linear concentration dependency (Table 14 & Figure 42).
This should be interpreted as all cis derivatives sharing the same external stimulus for
c) d)
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 73
Jacobs University Bremen
production. This stimulus is different from that of trans-CQA biosynthesis. Due to that the
stimulus of cis-CQA production is direct irradiation by UV light of exposed leaves (appendix E).
Figure 42.Linear dependency of cis-5-CQA with 5-CQA and two isomers of cis-4,5-diCQAs.
Table 14.Correlation coefficients of cis-isomers according to Spearman’s rule
Correlations
Cis-5CQA Cis-4,5diCQA1 Cis-4,5diCQA2
Spearman's rho
Cis-5CQA
Correlation Coefficient 1.000 0.265** 0.183*
Sig. (2-tailed) . 0.003 0.041
N 126 126 126
Cis-4,5diCQA1
Correlation Coefficient 0.265** 1.000 0.731**
Sig. (2-tailed) 0.003 . 0.000
N 126 126 126
Cis-4,5diCQA2
Correlation Coefficient 0.183* 0.731** 1.000
Sig. (2-tailed) 0.041 0.000 .
N 126 126 126
** Correlation is significant at the 0.01 level (2-tailed).
* Correlation is significant at the 0.05 level (2-tailed).
In a dataset comprising 120 samples of Stevia rebaudiana grown in different locations, linear
correlations of trans-mono and di-CQAs with each other indicates a common stimulus in
production. However, non-linear correlation between cis and trans-CQAs and linear correlation
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 74
Jacobs University Bremen
between the cis-CQAs convincingly demonstrates that the production of cis and trans CQA
derivatives must follow two distinctly different pathways, where trans-CQAs production is not
UV dependent stimulus.
To substantiate this hypothesis, available UV irradiation data with cis-CQA concentrations were
correlated. As a first set of data, changes of cis-CQA concentrations in different harvests were
studied. For two set of samples three harvests were carried out from June to September of crops
grown at the same locations. For the last harvest in September the number of sunshine hours
affecting the plants was always considerably lower if compared to the two early harvests from
June to August. Sunshine hours were available through the weather databases of the nearest
airport weather station (www.weather.org and www.weather.online.co.uk). Representative data
are shown in Figure 43 where in two bar charts the total cis-CQA concentration is given for three
harvests in two different locations.
Figure 43.Amount of 5-CQA in mg/100g dry leaves from three harvests from location A (TCV)
and location B (Amfilikeia) during 2011.
0
50
100
150
200
250
300
350
400
Harvest 1 A Harvest 1B Harvest 2A Harvest 2B Harvest 3A Harvest 3B
mg/100g
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 75
Jacobs University Bremen
A more global picture including all available data was obtained by the log of trans/cis-5-CQA
concentrations against the number of sunshine hours in the month prior to sample collection for
six locations were in 2010 and 2011 a total of ten harvests were collected. The data plotted in
this manner display a linear relationship with the quotient trans/cis-5-CQA concentration clearly
depending on the number of sunshine hours. The more sunshine hours the plant leaves were
exposed to the smaller quotient, so the higher the relative amount of cis-5-CQA concentrations
(Figure 44).
Figure 44.log of trans/cis-5-CQA concentrations against the number of sunshine hours in the
month for a total of ten harvests from six locations.
0
0.5
1
1.5
2
2.5
3
0 50 100 150 200 250 300 350 400
log C(trans)/C (cis) vs sunshine hours in month
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 76
Jacobs University Bremen
5.3.4.3. Analysis of Variance (ANOVA)
One way ANOVA was performed to study the influence of origin and, in another test, variety on
the mono and di-CQA content of stevia leaves cultivated in various regions of EU with different
soil and climatic conditions. Post-hoc test was applied to find out which origin or varieties differ,
if any.
ANOVA is a statistical analysis used for identifying factors that are influencing a given dataset.
It is a statistical technique for comparing mean values for multiple independent populations.
ANOVA analysis selects the most discriminating variables by calculating an F factor which is
proportional to the ratio of the ‘within-group’ variance to the ‘between groups’ variance. The
higher this ratio, the more the groups are significantly different from each other141
.
The One-Way ANOVA, used in this study, compares the mean of one or more groups based on
one independent factor. However, there are two assumptions to be met within the dataset. First
one is, the data should be normally distributed (the data distribution was analyzed in chapter
5.3.4.1). The second assumption is that the variances of the groups to be compared should be
similar. This can be checked by Levene’s test (test of homogeneity of variances). If the
significance value is greater than 0.05 (found in the Sig. column, e.g. Table 15) then there is
homogeneity of variances and therefore, the assumption of homogeneity of variance is met. If the
Levene's test is significant (lower than 0.05), it would mean that there is no similar variances and
the assumption of homogeneity of variance is not met, therefore it would be necessary to use an
adjusted test such as the Welch statistic (the Robust Tests of Equality of Means) within ANOVA.
In general, when setting up the analysis, it is common and advantageous to select a test for either
situation since we do not know if the assumption is met or violated.
Influence of Origin: Levene’s test was performed to assess whether the assumption of
homogeneity of variance between groups is met within the data.
Test of homogeneity results revealed that for dependent variable 3-CQA and 4,5-diCQA
Levene’s test is not significant (F8/116 = 1.783, p=0.087), (F8/116 = 1.430, p=0.191),respectively.
However, for 5-CQA, 4CQA and 3,5-diCQA Levene’s test was significant at the level of 0.05
(Table 15), which states that the variances in the different groups of origins are different (the
groups are not equal variance on the dependent variable, that is origin in this case) and therefore,
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 77
Jacobs University Bremen
a modified procedure which do not assume equality of variance (Welch statistic) was also
applied to these data along with ANOVA.
Table 15. Results of test of homogeneity of variances
Levene
Statistic
df1 df2 Sig.
cqa3 1.783 8 116 0.087
cqa5 2.058 8 116 0.046
cqa4 2.073 8 116 0.044
totalmono 2.146 8 116 0.037
diCQA35 2.018 8 116 0.050
diCQA45 1.430 8 116 0.191
totaldi 1.733 8 116 0.098
*df: degrees of freedom
ANOVA results (Table 16) did not reveal significant differences between the average CGA
content of stevia leaves of different origins. The data only hints at some differences in the
average content of stevia leaves from different origin destinations. These, however, can be
considered only marginally significant at the 5% level (p= 0.054 for e.g. 4-CQA). Moreover,
from the ANOVA output, robust test of equality of means (Welch test) is considered for the
CQAs that are not meeting the assumption of homogeneity of variance. Like the ANOVA test, if
the significance value is less than 0.05 in Welch test, then it means there are statistically
significant differences between groups.
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 78
Jacobs University Bremen
Table 16.ANOVA results for effect of origin on stevia CGA content
From the results of Welch test and One-Way ANOVA, the conclusion would be that at least two
of the group means is significantly different from the each other. Beyond this, it is necessary to
conduct a post-hoc comparison test to see exactly which pairs of groups are significantly
different. There are varieties of post hoc tests available. The most commonly used LSD test is
applied for CQAs (3CQA, 4,5-diCQA) which are meeting the assumption of homogeneity of
variance. Games-Howell post hoc test is applied for the CQAs (5CQA, 4CQA, 3,5diCQA)
violating the assumption of homogeneity of variance (equal variances not assumed).
Multiple comparison (post hoc) results obtained from LSD test (Fisher's Least Significant
Difference) revealed that the 3-CQA amounts are significantly different at the 0.05 level between
the origins of Agrinion&Granada, Granada&TCV, Granada&Uconor and Granada & non-EU
origins. 4,5-diCQAs resulted to be significantly different at the 0.05 level for origin of Granada
and the all other origins except Portugal.
Multiple comparison (post hoc) results obtained from Games-Howell test revealed that there is
no significant difference at the 0.05 level between the origins for 5-CQA and 4-CQA. However,
significant difference observed for 3,5-diCQA between the origins of Amiflikeia and Toumpa
with the significance value of 0.008.
One Way Anova
Compound F Sig. (p)
3-CQA 1.022 0.138
5-CQA 1.850 0.054
4-CQA 1.853 0.054
Totalmono-CQA 1.985 0.064
3,5-diCQA 1.767 0.106
4,5-diCQA 1.727 0.127
Total-diCQA 1.881 0.089
Robust Test of Equality of Means
Compound F Sig. (p)
3-CQA Welch 1.823 0.132
5-CQA Welch 1.588 0.183
4-CQA Welch 1.590 0.182
Totalmono Welch 1.960 0.101
3,5-diCQA Welch 3.135 0.017
4,5-diCQA Welch 2.666 0.034
Total-diCQA Welch 3.120 0.017
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 79
Jacobs University Bremen
Influence of Variety: Test of homogeneity results revealed that for dependent variable 3-CQA
(F7/117 = 1.207, p=0.304), 3,5-diCQA (F7/117 = 0.545, p=0.799) and 4,5-diCQA (F7/117 = 1.193,
p=0.312) Levene’s test was not significant. However, for 5-CQA and 4-CQA Levene’s test was
significant at the level of 0.05 (p=0.018) (Table 17). Same as in the analysis of origins, Welch
test was also applied to data for the cases of violation of homogeneity of variances.
Table 17. Test of homogeneity of variances
Levene Statistic df1 df2 Sig.
cqa3 1.207 7 117 0.304
cqa5 2.542 7 117 0.018
cqa4 2.549 7 117 0.018
totalmono 2.242 7 117 0.036
diCQA35 0.545 7 117 0.799
diCQA45 1.193 7 117 0.312
totaldi 1.090 7 117 0.374
*df:degrees of freedom
ANOVA results did not reveal significant differences between the average CGA content of
stevia leaves of different varieties. Moreover, Welch test did not reveal any significant difference
at the level of 0.05. The results of ANOVA test and Welch test is presented in Table 18.
Table 18.ANOVA results for effect of variety on stevia CGA content
One Way ANOVA
Compound F Sig. (p)
3-CQA 0.849 0.549
5-CQA 0.459 0.862
4-CQA 0.464 0.859
Totalmono-CQA 0.463 0.860
3,5-diCQA 1.901 0.075
4,5-diCQA 1.446 0.194
Total-diCQA 1.577 0.149
Robust Test of Equality of Means
Compound F Sig. (p)
3-CQA Welch 0.926 0.500
5-CQA Welch 0.428 0.878
4-CQA Welch 0.433 0.874
Totalmono Welch 0.373 0.911
3,5-diCQA Welch 1.637 0.161
4,5-diCQA Welch 1.250 0.305
Total-diCQA Welch 1.402 0.238
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 80
Jacobs University Bremen
Post-hoc tests with Games-Howell method resulted with no pair-wise differences for 5-CQA and
4-CQA for each variety. However for 3-CQA with the LSD test, the mean difference between
variety 1-5 and variety 3-5 was significant at the 0.05 level (p=0.04). The varieties 4-6, 5-7 and
6-7 had significant difference (p=0.04) when 4,5-diCQA quantity in stevia leaves was taken as a
dependent variable. Variety 6 showed significant difference with varieties 2, 3, 4 and 7 for 3,5-
diCQA quantity in stevia leaves.
5.4. Conclusion
Mono-CQAs and di-CQAs including their corresponding cis-isomers and flavonoid glycoside of
120 samples (after removal of outliers and duplicates) of Stevia rebaudiana grown in different
locations and with different botanical varieties were profiled and quantified successfully. For the
first time a full quantitative data set from a large number of samples from different origins and
varieties, in which the full profile of all secondary metabolite quantities was determined in any
agricultural plant was obtained.
With the obtained dataset, differences between stevia leaves and influence of the origin of
cultivation and the botanical varieties on the CQAs profile were studied successfully with the
most common used statistical method ANOVA. Pair-wise comparisons of varieties and origins
for each CQA were achieved by less conservative statistical post hoc test (LSD) and Games-
Howell post hoc test to determine exact pair of variety/origin that are differentiating from each
other.
Correlation studies showed that the production of cis and trans CQA derivatives must follow two
distinctly different pathways and the stimulus of cis-CQA production is direct irradiation by UV
light exposed leaves. Therefore, cis-CQA concentrations may serve as useful markers of UV
exposure of plant material in agricultural practices.
Finally, with the PCA, it was possible to differentiate the EU and non-EU cultivated stevia
leaves according to their rebaudioside A and diCQAs profile.
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 81
Jacobs University Bremen
6. LIPID ANALYSIS of STEVIA
6.1. Overview
Lipid profile of stevia leaves was determined for 46 chosen samples by GC-MS. Total lipid
amounts of each sample were determined gravimetrically after chloroform extraction.
Identification was achieved for the fatty acids by comparison of retention times and mass spectra
with the commercially obtained fatty acid methyl esters (FAME) reference mixture. This study
was complemented by a MALDI-TOF-MS anaylsis for further lipid identification. Additionally,
a steam distillation extract was subjected to GC-MS analysis for the analysis of volatile terpenes.
Identification of the terpenes was achieved by software based NIST library search.
6.2. Materials & Methods
6.2.1. Extraction
Two grams of stevia leaves was immersed in liquid nitrogen, ground in a hammer mill, and
extracted with 150 mL of chloroform in a Soxhlet apparatus (Buchi B-811 extraction system) for
2 h. The solvent was removed in vacuo and total lipid amount was determined gravimetrically.
6.2.2. Methyl Ester Formation
Total lipid extract was dissolved in 2 mL of chloroform after gravimetric determination. 200 μL
of methanolic potassium hydroxide solution (2 mol/L) and 1 g of sodium hydrogen sulfate
monohydrate (NaHSO4) were added to 1 mL of lipid extract solution in chloroform to form the
methy esters of lipids for GC analysis.
6.2.3. GC-FID
GC analyses were performed on GC-2010 (Shimadzu, Kyoto, Japan) equipment with flame
ionization detector and split/splitless injector. Injector temperature was at 290 °C and samples
were injected using autosampler (1 µL) with split ratio of 1:10. Capillary columns was used Rxi -
5 ms (15 m × 0.25 mm, with film thickness of 0.25 µm) Restek. The temperature program was
raised from 80 °C (1min) up to 300 °C at rate 5°C/min, and the total run time was 50 mins.
Helium was used as carrier gas at flow rate of 5 mL/min. Detector temperature was set at 310 °C.
To form the flame, hydrogen gas flow, 40 mL/min, and air gas flow, 400 mL/min, were used.
GC solution 2.10 software was used for data collection, and calculation of all parameters.
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 82
Jacobs University Bremen
6.2.4. GC-MS
GC-MS analyses have been carried out with a Varian CP-3800 gas chromatograph (Palo Alto,
CA, USA) equipped with a split/splitless injector and coupled to a Saturn 2000 ion-trap Varian
mass spectrometer (MS). Data acquisition was performed using a Star Toolbar system (Varian).
Samples were injected manually with split ratio of 1:10 at 290 °C. The compounds were
analyzed on a 30 m, 0.25 mm I.D. fused-silica capillary column coated with a 0.1µm layer of
poly (5% phenyl/95% dimethylpolisiloxane) (Rxi-5Sil MS, Restek) using helium as the carrier
gas at flow rate of 1.3 mL/min, respectively. The oven temperature was heated from 80 °C (1
min) to 300 °C at the rate 5 °C/min and the total run time was 50 min. For the MS, the electron
multiplier was set to 1350 V and ionization was accomplished by electron impact (EI). The
transfer line temperature was set at 300 °C whilst 244 °C and 120 °C were the temperatures used
for the trap and the manifold, respectively. Mass spectra were recorded from m/z 40–600.
6.2.5. Calibration Curve of FAME
The quantitative analyses have been initiated by generation of calibration curves for FAME
standard mixture from the range of C14:0/C14:1 – C24:0/C24:1. Calibration curves were
generated by plotting peaks areas of FAME standard mixture (Marine Oil FAME Mix (20
components) from Restek) at different concentrations as function of peak areas. In doing so,
FAME standard solution was diluted with n-heptane to a series of standards with concentrations
of 5, 50, 100, 250, 500, 750, 1000, 1250, 1500, 2000, 3000 μg/L.
6.2.6. MALDI-TOF MS
As matrix solution 5g/L 2,5-dihydroxybenzoic acid (2,5-DHB) solution in acetonitrile containing
0.1% trifluoroacetic acid (TFA) was used due to robustness of DHB to impurities. 1 μl aliquot of
the organic extracts of stevia was mixed with 1 μl of the matrix solution on the maldi target
(MPT AnchorChip 600-384 target, Bruker Daltonics) and the matrix crystals were allowed to air-
dry.
MALDI-TOF spectra were acquired on an Autoflex II MALDI-TOF-TOF mass spectrometer
(Bruker Daltonics) equipped with a 337 nm nitrogen laser. The instrument was operated in the
reflector mode: source, 19.00 kV; lens, 8.95 kV; and reflector, 20 kV, using an optimized ion
extraction delay time of 80 ns. The laser frequency was set at 25 Hz with 50 laser shots per
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 83
Jacobs University Bremen
acquisition. The laser strength was kept about 40% above threshold to obtain optimum signal to
noise ratio. Spectra were obtained by summing, on average, 200 laser shots. Spectra were
acquired in the mass range 0–2500 amu. The instrument was externally calibrated in the
enhanced quadratic calibration mode prior to acquisition using a peptide tune-mix sample
(Bruker Daltonics).
6.3. Results & Discussion
46 Stevia rebaudiana samples were chosen for lipid analysis. Firstly, total lipids were
determined gravimetrically after extraction with a non-polar solvent. Since hexane extraction
provided values below 1 weight %, a Soxhlet chloroform extraction was selected for this
purpose. The total lipid sample was in a following step subjected to GC-MS analysis. Table 19
shows total lipid values obtained for 46 representative samples comprising at least three samples
from all seven varieties and samples from all origins. Secondly the total lipid fraction was
subjected to basic hydrolysis and derivatisation followed by GC-MS analysis to identify the
individual fatty acid spectrum of all seven stevia varieties and of representative samples from all
origins. Derivatisation chosen included the formation of fatty acid methylesters. Identification of
fatty acid methylesters was achieved through GC-MS by comparison of retention time and mass
spectra with a commercial certified reference compound mixture. A representative GC
chromatogram is shown in Figure 45 for total lipid profile of stevia after chloroform extraction
and in Figure 46, the GC chromatogram of FAME standard mixture can be observed142, 143.
Figure 47 presents the fragmentation of fatty acid by electron impact ionization obtained in GC-
MS measurement of stevia leaves chloroform extract.
Additionally a steam distillation extract was obtained and subjected to GC-MS analysis to profile
the volatile terpenes and to compare the lipid composition with chloroform extract.
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 84
Jacobs University Bremen
Figure 45.GC.MS chromatogram of total lipid extracts from Stevia rebaidiana leaves from
sample (Uconor, Var.4).
Figure 46.GC-MS chromatogram of FAME standard mixture.
C14:0
C14:1
C16:0
C16:1
C18:1 C18:1 C18:2 C18:3
C18:0
C20:0 C20:1 C20:2
C20:4 C20:3
C20:5
C22:0
C22:1
C22:6
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 85
Jacobs University Bremen
Figure 47.Representative EI-MS spectra obtained from GC-MS measurement of stevia extract.
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 86
Jacobs University Bremen
Table 19.Total lipid values in weight % from 46 samples
Sample no Origin Variety Harvesting Year %
119 Agrinion 4 II 2011 3.42
120 conaga 6 I 2011 7.71
121 Conaga 7 I 2011 4.19
122 Agrinion 6 II 2011 5.14
123 Agrinion 3 II 2011 5.45
124 Toumpa 7 I 2011 3.05
125 Toumpa 4 III 2011 4.43
126 Toumpa 3 III 2011 6.67
127 Toumpa 3 II 2011 5.59
128 Agrinion 4 II 2011 0.50
129 Toumpa 4 II 2011 4.32
132 Agrinion 3 II 2011 5.92
133 Toumpa 5 II 2011 4.26
134 Toumpa 6 II 2011 4.15
135 Toumpa 7 II 2011 3.92
136 TCV 5 I 30.06.2011 1.68
137 TCV 3 I 30.06.2011 5.48
138 TCV 6 I 30.06.2011 5.16
139 TCV 3 II 11.08.2011 7.23
140 TCV 6 II 17.08.2011 4.90
141 TCV 4 II 24.08.2011 4.09
142 TCV 7 I 18.08.2011 2.11
143 TCV 4 I 07.07.2011 3.42
144 TCV 5 II 17.08.2011 8.23
147 Amiflikeia 5 I 2011 3.25
148 Amiflikeia 6 I 2011 4.23
149 Amiflikeia 1 I 2011 4.27
154 Toumpa 5 I 2011 3.24
8 TCV 4 I 04.08.2010 1.52
49 Toumpa 3 I 30.07.2010 3.37
43 Agrinion 2 II 20.09.2010 2.63
75 Granada 3 I 09.09.2010 3.41
54 Amfilia 4 II 15.09.2010 2.71
30 Uconor 4 II 07.05.2010 0.54
53 Amfilia 4 I 04.08.2010 2.99
5 TCV 1 I 04.08.2010 0.89
16 TCV 2 I 04.08.2010 0.98
19 TCV 3 II 10.09.2010 6.74
6 TCV 2 I 04.08.2010 5.11
10 TCV 2 II 28.09.2010 7.05
14 TCV 2 I 04.08.2010 2.27
32 Uconor 2 I 11.08.2010 3.14
43 Agrinion 2 II 20.09.2010 3.69
37 Uconor 4 I 11.08.2010 2.60
34 Uconor 4 II 07.09.2010 5.13
36 Uconor 4 I 11.08.2010 1.46
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 87
Jacobs University Bremen
For determination of the fatty acid profile GC-FID was applied using the GC-MS data as a
reference point. Oleic acid was found to be the predominant fatty acid in stevia with an average
ratio of saturated to unsaturated fatty acids of 1:3.8. Among the saturated fatty acids 16:0
palmitic acid was the predominant acid with varying quantities of stearic acid present (24 - 2 %).
Saturated fatty acids with shorter and longer chain length were not identified. In terms of
unsaturated fatty acids 16:1 was identified next to three co-eluting isomers of 18:1, with oleic
acid being the predominant compound. Values stated for 18:1 acids represent a sum over all
three isomers. A series of polyunsaturated fatty acids were identified as well comprising 16:2,
18:3 and 18:2 acids with variable quantities. Values are given in Table 20.
Table 20.Quantities of polyunsaturated fatty acids
Sample Origin Variety Harvest Year % 16:2 %16:1 %16:0 %18:3 % 18:2 % 18:1 % 18:0
S10 TCV 2 II 28.09.2010 15.45 9.97 12.04 15.56 10.25 35.09 1.64
S6 TCV 2 I 04.08.2010 20.41 20.52 11.42 2.51 7.57 34.59 2.98
S14 TCV 2 I 04.08.2010 22.03 14.73 9.50 13.49 6.87 30.39 2.99
S34 Uconor 4 II 07.09.2010 9.29 1.71 12.26 6.00 11.29 52.06 7.39
S37 Uconor 4 I 11.08.2010 3.54 2.10 14.70 8.41 12.44 55.97 2.84
S5 TCV 1 I 04.08.2010 2.09 0.72 14.59 4.70 13.04 57.64 7.23
S10 TCV 2 II 28.09.2010 14.56 9.02 13.01 11.90 10.32 36.11 5.09
S16 TCV 2 I 04.08.2010 4.00 0.51 14.56 5.50 12.61 55.03 7.78
S19 TCV 3 II 10.09.2010 1.21 0.57 12.57 8.35 12.77 48.69 15.83
S32 Uconor 2 I 11.08.2010 2.18 1.11 14.86 6.80 11.80 53.53 9.71
S36 Uconor 4 I 11.08.2010 5.15 3.38 14.18 9.03 13.75 46.55 7.96
S43 Agrinion 2 II 20.09.2010 1.49 0.39 11.12 10.81 10.28 49.80 16.11
S45 Portugal 1 I 07.07.2010 1.42 0.68 10.35 11.98 10.40 40.32 24.86
S121 Conaga 7 I 2011 7.91 4.77 14.20 6.00 13.59 47.71 5.82
S122 Agrinion 6 II 2011 8.44 5.68 15.47 4.10 9.43 52.31 4.57
S123 Agrinion 3 II 2011 10.69 6.36 13.41 8.63 9.44 48.57 2.90
S124 Toumpa 7 I 2011 0.93 0.46 21.06 5.77 16.42 54.17 1.20
S127 Toumpa 3 II 2011 9.05 5.67 16.23 6.39 13.06 43.56 6.03
S136 TCV 5 I 30.06.2011 1.11 0.61 14.03 6.84 10.42 52.11 14.88
S143 TCV 4 I 07.07.2011 8.05 5.38 14.04 4.43 11.29 50.25 6.56
S147 Amiflikeia 5 I 2011 8.27 5.48 15.26 8.71 8.41 45.93 7.94
S148 Amiflikeia 6 I 2011 10.93 7.22 15.29 7.20 9.95 42.27 7.13
S149 Amiflikeia 1 I 2011 7.26 4.38 16.83 4.95 10.70 49.61 6.27
S154 Toumpa 5 I 2011 0.63 0.39 15.93 4.80 11.48 60.90 5.86
S120 Conaga 6 I 2011 1.71 0.55 14.79 5.37 11.58 58.93 7.07
S142 TCV 7 I 18.08.2011 2.23 1.31 17.64 6.13 15.83 51.55 5.30
Average 6.93 4.37 14.21 7.48 11.35 48.22 7.46
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 88
Jacobs University Bremen
It should be noted that in plants fatty acid composition is reported to be a consequence of
climatic conditions, in particular temperature, rather than botanical variation. The total lipid
content determined is well in line with values from other leafy dietary plants. For example values
for green tea has been determined as 3-5% total lipids, spinach for 4.5 % total lipids and the
botanically related Asteraceae plant lettuce at 4-6 % total lipids. The average fatty acid
distribution and their structures are shown in Figure 48 and Figure 49.
Figure 48.Structures of fatty acids in Stevia rebaudiana extract.
OH
O
16:0 Palmitic acid (hexadecanoic acid)
OH
O
16:1 Palmitoleic acid (hexadec-9-enoic acid)
HO
O
16:2 9,12-Hexadecadienoic acid
OH
O
18:0 Stearic acid (octadecanoic acid)
OH
O
18:1 Oleic acid (octadec-9-enoic acid)
HO
18:2 Oleic acid (9,12 - octadecadienoic acid)
O
OHO
18:3 Linoleic acid (9,12,15 - octadecatrienoic acid)
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 89
Jacobs University Bremen
Figure 49.Fatty acid profile of average stevia leaf in %. X:Y denominates the number of carbon
atoms in the fatty acid (X) and the number of double bonds in the fatty acid (Y). Data were
obtained from 46 samples after lipid hydrolysis, derivatisation and GC-MS analysis
Average fatty acid and lipid values determined were as follows:
• Total lipids: 4.52 % (0.89-8.23 %)
• Ratio saturated/unsaturated fatty acids: 1:3.8 (1:6-1:1.8)
This work was complemented by a MALDI-TOF-MS anaylsis of the lipid fraction allowing
identification of selected intact lipids, mainly triacylglycerides of oleic acid. A representative
MALDI-MS spectrum is shown in Figure 50.
0
10
20
30
40
50
60
% 16:2 %16:1 %16:0 %18:3 % 18:2 % 18:1 % 18:0 % X:Y
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 90
Jacobs University Bremen
Figure 50.MALDI-MS spectrum of total lipid extract in positive ion mode using 2,5-DHB as a
matrix. Main peak at m/z 872.3 corresponding to trioleic acid glyceride.
Finally direct GC-MS analysis of the lipid fraction was carried out supplemented by a GC-MS
analysis of a steam distillation extract of Stevia rebaudiana leaves. This analysis allowed
positive identification of twelve mono- and sesquiterpenes using a NIST library search. The
NIST library search identification of compounds was made if the database spectrum showed a
match with the experimental spectrum with a NIST score of 800 or above. Additionally a library
A database search was carried out and compounds accepted if a library A score of above 25 %
and a NIST score of above 800 was observed. A NIST score of above 800 is generally accepted
in the literature for positive compound identification. All compounds identified were additionally
present in both steam distillation extract and total lipid fraction. Selected structures identified are
shown below in Figure 51 and retention times with NIST score for few terpenes are presented in
Table 21.
It should be noted that previously an additional 25 volatile terpenes have been reported in Stevia
rebaudiana leaves. These compounds are shown in the appendix F, but their presence could not
be confirmed in this study144
.
229.4
326.7
361.5
478.8
533.8
593.9
872.3
0_O20\1: +MS
0
20
40
60
80
100
Intens.
[%]
200 400 600 800 1000 1200 1400 1600 1800 2000 m/z
H2C O (CH2)8 CH CH (CH2)7 CH3
HC O CO (CH2)7 CH CH (CH2)7 CH3
H2C O CO (CH2)7 CH CH (CH2)7 CH3
Ether type lipid: 1,2-dioleoyl-3-oleylglycerol (C57H106O5)
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 91
Jacobs University Bremen
O
myrtenal alpha pinenesabinene terpinene
O
cumin aldehyde
O
1,8 cineol
3-carene copaene humulene (caryophyllene)
gamma cadinene
-pinenelimonene
Figure 51.Chemical structures of terpenes identified in Stevia rebaudiana leaves.
Table 21.Retention time and NIST scores of some terpenes identified in stevia extract
Terpene RT (min) NIST score m/z
sabinene 9.5 838 136
α-pinene 10.4 920 136
3-carene 10.3 925 136
caryophyllene 10.5 727 204
γ-cadinene 11.2 887 204
copaene 12.8 848 204
In addition to the mono- and sesquiterpenes the GC-MS data show a series of triterpenes at
longer retention times. NIST search clearly indicates their identity as triterpenes with low NIST
scores for steroid and related structures, however, no match in the database allowed positive
compound identification. None of the terpenes identified was reported to show a toxicologically
problematic profile. Compounds have been reported in many additional dietary plants. General
volatile terpene levels in the plant are very low. The total FID or TIC (total ion current in MS
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 92
Jacobs University Bremen
data) are around 3-5 % of the total intensity of the lipid fraction, which allows an estimate 1.5 -
2.5 × 10-3
weight % of total volatile lipids.
Additionally the GC-MS data revealed the presence of giberillic acid (structure below) a well
known plant hormone in the lipid fraction. To confirm its existence, methylester formation of
reference compound of steviol (having the closest chemical structure) was performed and
subjected to GC-MS analyses (Figure 52). The retention times and mass spectral data were
compared between stevia extract and the steviol standard. Furthermore, increase in the intensity
was observed after standard steviol addition to the stevia extract.
Figure 52.GC chromatogram of methylesterified steviol and stevia extract.
O
OC
COOHHO
OH
giberillic acid
OR
CH3
RO2C CH3
steviol
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 93
Jacobs University Bremen
Further secondary metabolite search
The groups of Nash and Fleet recently reported on the isolation of novel indolizidine alkaloid
they named steviamine from Stevia rebaudiana leaves.145
Within the LC-MS data available ions
corresponding to its m/z ratio was searched, however, an ion corresponding to this mass at a
significant level was not found in any sample analyzed in neither positive nor negative ion mode
(above S/N 20). The absence of the ion can be confirmed later on in future studies with an
authentic reference samples requested from the Fleet group.
N
OHHO
HO
Steviamine (not found)
6.4. Conclusion
Lipid profile and quantification data for chosen stevia samples were successfully obtained.
Steam distillation and chloroform extracts of stevia leaves analyzed on GC-MS and GC-FID did
not have significant differences in lipid profile. Furthermore, volatile terpenes were identified by
NIST library search.
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 94
Jacobs University Bremen
7. PROTEOMICS of STEVIA
7.1. Overview
The main aim of stevia leave protein analysis was to purify, separate and sequence stevia leaf
proteins with an aim to identify potentially allergenic proteins. Isolation and purification of
stevia leaf proteins were achieved using 2D gel electrophoresis. The resulting 2D gels were
stained with coumassie blue staining to visualise and identify individual proteins. Susbsequent
trypsin digestion and MALDI-TOF MS analysis was carried out on selected proteins. For
MALDI-MS sample preparation anchor chip MALDI targets were used in conjunction with 2,5-
DHB (2,5-dihydroxy benzoic acid) matrices. Using data base search algorithms the obtained
mass spectrometrical data were used in attempt to sequence the proteins.
7.2. Materials & Methods
7.2.1. Extraction of Proteins
Fresh plant tissues were crushed in pestle and mortar in presence of liquid nitrogen in order
to prevent the degradation of protein by the release of protease enzyme. The leaves were
crushed in fine powder, and to precipitate the proteins, the powder was suspended in TCA
extraction buffer at -20 oC for overnight (Table 22). The proteins precipitated as white flakes
after the overnight incubation. The supernatant is collected and centrifuged at 5000g for 30
min. The supernatant was discarded, and the protein pellets settled at bottom was washed
with ice cold acetone and centrifuged again at 5000g for 15 min. The washing step was
repeated for 3 to 4 times and then this protein pellet was dried by passing nitrogen gas at
slow stream, after the drying process these pellets can be stored at -80°C for few months for
later identification processes (Figure 53).116
Table22. Amount and properties of chosen stevia leaves
Sample
Number Harvest Variety Origin
Tissue
(g)
Extraction buffer
(mL/g)
Approx.protein weighed
after extraction (mg)
8 I 4 TCV 10 20 20
10 I 2 TCV 10 20 15
36 I 4 Uconor 10 20 12
7 II 7 TCV 10 20 15
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 95
Jacobs University Bremen
Figure 53.Extraction procedure of proteins. Red arrow shows the process of TCA extraction and
Blue arrow follows the phenol extraction116
.
7.2.2. Protein Analysis
SDS-PAGE Protocol: The protein pellet (200μg) was resuspended in IEF buffer/sample buffer
(125 μL for 7 cm strip). 100 μL of IEF buffer with the protein sample was vortexed for 10 min at
10000 rpm and the supernatant was collected. The remaining volume of 25 μL of IEF buffer was
added on the remaining pellet and vortexed for 20 mins and centrifuged at 10000 rpm. The
supernatants were collected and added to the previous collected supernatant.
The collected supernatant (IEF buffer with protein sample) was added in the rehydration tray
uniformly and then the IPG strip (strip for isoelectric focusing, ph range was 3 - 10 and pH 4 - 7)
was placed in the rehydration tray and it was kept overnight for sample absorption in to the strip.
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 96
Jacobs University Bremen
For running the IPG strip, the voltage gradient of 250 volt for 30 min, and then to 3500 volts for
4 hours was used.
Equilibration buffer I:
6M urea, 0.375M Tris-HCl pH8.8, 2% SDS, 20% Glycerol, 2%(w/v) DTT.
Equilibration buffer II :
6M urea, 0.375M Tris-HCl pH8.8, 2% SDS, 20% Glycerol, 2.5%(w/v) iodoacetamide.
Sample buffer:
CHAPS (Roche), Pharmalyte (pH 3-10) (GE Healthcare Lifesciences), dithiothreitol (DTT)
(Sigma-Aldrich), Serdolit MB-1 (Serva), urea, Pefabloc® (VWR), Thiourea (Fluka)146
Rehydration buffer was purchased from Biorad ReadyPrep™ 2D starter kit.
2D SDS-PAGE Protocol: The gel casting chamber was filled from the bottom to a height of
about 2 cm below the top of the glass plates with separation gel and stacking gel (Table 23) on
the top. The gels were carefully overlaid with 1.0-1.5 mL buffer-saturated 2-butanol to allow for
complete polymerization. In order to ensure good contact between the IPG strip and the gel an
agarose solution was added. The agarose solution was kept at 70 °C and added first on top of the
gel. Immediately after, the equilibrated strip was placed on the gel. The IPG strip gel was then
subjected to electric field at 121volt and 45Amp for 1 to 2 hours for separation of proteins
according to their molecular weight. The gel was stained overnight in Coomassie brilliant blue
and later on destained with dd H2O. The protein bands were excised for destaining and trypsin
digestion147
.
Table 23.Preparation of separation and stacking gel for 2D SDS PAGE
A. 12.5 % Separating Gel B. 4% Stacking gel
Water 3.3 ml
30% Acrylamide 4.2 ml
1.5M Tris (pH 8.8) 2.5 ml
10% SDS 100 μl
10% APS 50 μl
TEMED 5.0 μl
Water 6.1 ml
30% Acrylamide 1.3 ml
0.5M Tris (pH 6.8) 2.5 ml
10% SDS 100 μl
10% APS 50 μl
TEMED 10 μl
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 97
Jacobs University Bremen
Destaining gel pieces excised from Coomassie stained 2D SDS gel: The protein bands were
excised from 2D-SDS and the each excised band was cut in to smaller pieces. The gel pieces
were transferred to a microcentrifuge tube and 100 μL of 100 mM ammonium
bicarbonate/acetonitrile (1:1, v/v) was added, incubated with occasional vortexing for 30 min.
500 μL of neat acetonitrile was added and incubated at room temperature with occasional
vortexing, after gel pieces become white and shrink, the acetonitrile was removed. The destained
gel pieces were then subjected to in-gel digestion. Alternatively, they can be stored at -20 0C for
a few weeks until they needed.
Trypsin Digestion: 50 μL of trypsin buffer (Promega gold mass spectrometry grade, Germany)
was added on the destained gel pieces and it was left in an ice bucket for 30 min, and more
trypsin buffer was added if all solution was absorbed. After 90 mins, 10 μL of ammonium
bicarbonate buffer was added to cover the gel pieces and to keep them wet during the enzymatic
cleavage. The gel pieces were kept in an air circulation thermostat for incubation overnight at 37
0C. The tubes were chilled to room temperature, and 1μL aliquot of supernatant was directly used
for MALDI-TOF MS analysis. For further analysis 10 μL of 0.1 % (v/v) TFA was added in to
the tube, vortexed and centrifuged at 10,000 rpm, aliquot was withdrawn and dried down in a
vacuum centrifuge and stored at -20 0C till it was needed for further MS/MS analysis.
7.2.3. MALDI-TOF MS
MALDI-TOF MS analysis was carried out on selected trypsin digested proteins. For MALDI-
MS sample preparation anchor chip MALDI targets (MPT AnchorChip 600-384 target, Bruker
Daltonics.) were used in conjunction with DHB (dihydroxy benzoic acid) matrices (Sigma
Aldrich). Using data base search algorithms the obtained mass spectrometrical data were used in
attempt to sequence the proteins.
As matrix solution 5g/L 2,5-dihydroxybenzoic acid (DHB) solution in acetonitrile containing
0.1% trifluoroacetic acid (TFA) was used since DHB is more robust to impurities. 1 μl aliquot of
the organic extracts of stevia was mixed with 1 μl of the matrix solution on the maldi target
(MPT AnchorChip 600-384 target, Bruker Daltonics) and the matrix crystals were allowed to air-
dry.
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 98
Jacobs University Bremen
MALDI-TOF spectra were acquired on an Autoflex II MALDI-TOF-TOF mass spectrometer
(Bruker Daltonics) equipped with a 337 nm nitrogen laser. The instrument was operated in the
reflector mode: source, 19.00 kV; lens, 8.95 kV; and reflector, 20 kV, using an optimized ion
extraction delay time of 80 ns. The laser frequency was set at 25 Hz with 50 laser shots per
acquisition. The laser strength was kept about 40% above threshold to obtain optimum signal to
noise ratio. Spectra were obtained by summing, on average, 200 laser shots. Spectra were
acquired in the mass range 0–7500 amu. The instrument was externally calibrated in the
enhanced quadratic calibration mode prior to acquisition using a peptide tune-mix sample
(Bruker Daltonics).
7.3. Results and Discussion
7.3.1. SDS Results
SDS gel was performed to separate the proteins according to their molecular weight, but SDS gel
resulted with only one big band which corresponds to molecular weight of 55 kDa (Figure 54)
and other bands appeared as a light background, which can be due to insufficient separation or
due to low concentration of proteins with low range molecular weight. The bands on SDS gel
was subjected to trypsin digestion and analyzed by MALDI-TOF. The results from MALDI were
not sufficient for characterization of stevia proteins, thus, 2D-SDS separation of proteins were
performed.
There was no significant difference observed in the protein profile of four chosen stevia samples,
judged by their SDS gel, 2D gel bands and MALDI-TOF analysis.
In addition, only for one stevia sample phenol extraction method was tested and SDS gel bands
were compared with TCA/acetone extraction. Essentially, there was no benefits obtained from
phenol extraction, though the procedure was more time consuming.
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 99
Jacobs University Bremen
Figure 54.SDS Gel, for sample number 8 TCV harvest I, loaded on gel at different
concentrations 1mg/mL and 0.5mg/mL.
7.3.2. 2D-SDS
Proteins extracted by TCA/acetone method was separated by 2D SDS−PAGE are shown in
Figure 55. Essentially, the same results were obtained when the samples were prepared by
phenol precipitation (data not shown). Resolved protein spots were more concentrated at 55kDa
band, thus, more interest was given on these protein spots for further analysis by MALDI-TOF.
The selected spots were cut and treated with trypsin digestion separately and subjected to
MALDI-TOF analysis.
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 100
Jacobs University Bremen
Figure 55.2D-SDS separation of stevia total protein extract. 7cm strip of pH 4-7, where spot 1
and 2 are at 55kDa, and spot 5 at ~15kDa
7.3.3. MALDI-TOF MS Results
Proteins isolated by 2-D electrophoresis were digested by trypsin, and the resulting peptides were
mass analyzed by MALDI-TOF. The masses obtained from the spectral data were compared with
expected values computed from sequence database entries according to the enzyme's cleavage
specificity. The results were scored, and the ranking suggests the protein being identified or not.
Enzyme cleavage specificity, number of detected cleavage peptides, and mass accuracy are the
critical parameters148
. Moreover, peptide mass tolerance was set to 50 ppm; the mode of
proteolytic digestion was chosen as ‘trypsin digestion’ the searching database used was NCBI
and the searching taxonomy was “other green plants”. These parameters play a very crucial role
in MALDI –TOF for an exact surveillance of related protein sequence and reducing the false
positive results. The list of peptide masses were transferred into the peptide mass fingerprint
(PMF) search program Mascot.
The result of a peptide mass search with Mascot contains a lot of information. First, the
probability based score is very important. A protein is identified with a score higher than 100.
Second, the full protein summary report should be considered. By clicking onto the accession
number of the first hit more detailed protein information is displayed. The nominal mass must be
in accordance with the experimental data obtained from the gel electrophoresis. If this is not the
case, protein fragments or adducts should be considered. Furthermore, the sequence coverage
70kD
A 55kD
A
40kD
A
35kD
A
25kD
A
1
5
2
3
4
6
9
10
00
9
11
00
9
8
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 101
Jacobs University Bremen
(SC) and the number of mass values matched (MM) are very important. The difference between
the number of mass values searched and the number of mass values matched should be as small
as possible149
.
The results in Figure 56, shows that the PMF spectra generated has a close proximity to
ribulose-1,5-bisphosphate carboxylase (RuBisCO) enzyme150
, the result has the score of 232
with the expect value of 5.3E-18 at 50 ppm, and the sequence coverage is 45% (Figure 56 & 57).
The sequence which appears in bold black represents the matched peaks. Each matched peak
(m/z) defines a particular type of amino acid sequence, which is identified by database search.
The MS/MS fragmentation was processed for the unmatched and matched peaks with respect to
RuBisCO enzyme (Figure 58).
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 102
Jacobs University Bremen
Protein sequence coverage: 45%
Matched peptides shown in bold black.
1 KDYKLTYYTP EYETKDTDIL AAFRVTPQPG VPPEEAGAAV AAESSTGTWT
51 TVWTDGLTSL DRYKGRCYGI EPVPGEDNQY IAYVAYPLDL FEEGSVTNMF
101 TSIVGNVFGF KALRALRLED LRIPTAYVKT FDGPPHGIQV ERDKLNKYGR
151 PLLGCTIKPK LGLSAKNYGR ACYECLRGGL DFTKDDENVN SQPFMRWRDR
201 FLFCAEAIYK AQAETGEIKG HYLNATAGTC EDMMKRAVFA RELGVPIVMH
251 DYLTGGFTAN TSLAHYCRDN GLLLHIHRAM HAVIDRQKNH GMHFRVLAKA
301 LRMSGGDHIH SGTVVGKLEG EREITLGFVD LLRDDFIETD RSRGIYFTQD
351 WVSLPGVLPV ASGGIHVWHM PALTEIFGDD SVLQFGGGTL GHPWGNAPGA
401 VANRVALEAC VQARNEGRDL ATEGNEIIRE ATKWSPELAA ACEVWKEIKF
451 EFQAMDTLDG DKDKDKKR
Figure 56.MALDI-TOF MS spectra and mass list of trypsin digested 2D-SDS spot (spot number
8) and Mascot search result showing the sequence information for RuBisCO enzyme with the
score of 45%. The sequence which appears in bold black represents the matched peaks (known in
the database and matched with experimental data) and sequences which are in black represents
the unmatched peaks (unknown peaks).
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 103
Jacobs University Bremen
Figure 57.Mascot search result of MALDI spectra, the score is 232 and expectation rate is 5.3e-
18; this data is generated by using Matrix science which acts as search engine. NCBI database
and other green plant taxonomy was selected, the peptide mass tolerance was kept at 50 ppm and
two partials. The gene bank accession number is “gi| 38146633|”.
MALDI TOF peptide mapping of the stevia protein yielded partial identification of the sequence
from MASCOT database search. To prove the sequence obtained from the database and to have
more information on the unknown peaks MS/MS fragmentation of chosen peaks with significant
intensity was performed.
The MS/MS fragmentation of one of the unmatched peak m/z 842 was performed and the peptide
was tried to be sequenced by de-novo sequencing and database search. The de-novo sequence for
this peptide resulted with the sequence of AVAETVPR, however when this sequence was
subjected for MASCOT search, the result was 100% sequence coverage for a hypothetical
protein without any correlation with Stevia rebaudiana. Furthermore, the MS/MS fragmentation
of one of the matched peak with m/z 1230 was performed (Figure 58) and again the peptide was
sequenced by de-novo sequencing (Figure 59) and the sequence obtained was subjected to
MASCOT database search and indeed, the suggested sequence which was DLATEGNEIIR
resulted in 100% sequence coverage with that of RuBisCO sequence belonging to Stevia
rebaudiana as it was shown in Figure 56, with the gene bank accession number of gi| 38146633|.
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 104
Jacobs University Bremen
Figure 58.MS/MS de novo sequencing of the m/z 1230. Series of y and b fragments are labeled.
Figure 59.Structure and fragmentation of m/z 1230 based on de novo sequencing.
y1
b10
y10
b1
y6
b2
y7
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 105
Jacobs University Bremen
The database information on plant proteomics is too limited. The only plant which has entire
genome had been sequenced is Arabidopsis thaliana. Apart from that genome of few plants, such
as rice maze tomato and wheat had been sequenced. Currently, there is no plant genetically
related to the Asteraceae family available in the databases. Therefore, only insufficient sequence
coverage is being obtained from the database searches. The missing parts of the sequence were
partially covered by studying the tandem mass spectra of unknown peaks and sequencing by de-
novo method. However, successful de novo sequencing requires full sequence coverage, thus
demanding better quality spectra than those typically used for data base searching and sequences
obtained by de novo needs confirmation and this can be done either by chemically synthesizing
the obtained sequence and comparing their mass spectrical information or by complete
sequencing of the Stevia rebaudiana genome.
7.4. Conclusion
This study was the first attempt for sequencing leaf proteins of a plant from Asteraceae family
and for Stevia rebaudiana. MALDI-TOF has given a breakthrough in this research and once
again proved to be a very crucial technique in field of proteomics, for the first time ever Stevia
proteins are being characterized. The total of 75 peaks were generated by MALDI-TOF out of
which 33 matched peaks yielded protein score of 232 and 45% of sequence coverage of
RuBisCO enzyme, and 42 of them were considered to be unmatched with the native sequence.
The satisfactory match obtained in the database with the only protein sequence established in
stevia available, clearly indicates that the method employed was valid. However, further studies
are necessary to have the entire sequence of proteins exist in stevia leaves.
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 106
Jacobs University Bremen
8. SUMMARY
In summary, a unique and novel dataset comprising the full analysis of around sixty secondary
metabolites from an agricultural plant using seven different varieties from nine different origins
were obtained. In addition to this, correlation studies between the phytochemical information and
climatic metadata obtained within the project on growth conditions provide a new insight in to
agricultural plant science in general.
In more detail, secondary metabolite profile of 166 Stevia rebaudiana leave samples is analyzed
and ten volatile terpenes have been identified and fatty acid profile and quantities have been
obtained.
First attempt to sequence leaf proteins of a plant from Asteraceae family and for Stevia
rebaudiana was performed. Proteins were separated and analyzed successfully and efficient
results were obtained.
Furthermore, around fourty phenolic secondary metabolites, of the class of chlorogenic acids and
flavonoid glycosides were identified and quantified. Additionally, ten steviol glycosides have
been analyzed and quantified.
For a total of ten compounds accurate quantitative data have been obtained for all 166 samples
and for a further fourty compounds relative concentration variations. The data allow a full
description of variations between plants of different varieties and of different origins. Both for
phenolics and steviol glycosides significant variations between origins, varieties and harvests
have been observed as well as variations between stevia samples cultivated in EU and outside.
As conclusion, the quantitative data allow a scientifically sound and state of the art specification
of stevia leaves for licenscing as a novel food in the EU.
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 107
Jacobs University Bremen
9. REFERENCES
1. M. S. Bertoni, Revista de Agronomia de l'Assomption, Editon edn., 1899, vol. 1:35. 2. C. C. Shock, in California Agriculture, University of California, Editon edn., 1982, vol. 36, pp. 4-5. 3. N. Kolb, J. L. Herrera, D. J. Ferreyra and R. F. Uliana, Journal of Agricultural and Food Chemistry,
2001, 49, 4538-4541. 4. I. Prakash, G. E. DuBois, J. F. Clos, K. L. Wilkens and L. E. Fosdick, Food and Chemical Toxicology,
2008, 46, S75-S82. 5. C. Gardana, P. Simonetti, E. Canzi, R. Zanchi and P. Pietta, Journal of Agricultural and Food
Chemistry, 2003, 51, 6618-6622. 6. N. A. Ibrahim, S. El-Gengaihi, H. Motawe and S. A. Riad, European Food Research and
Technology, 2007, 224, 483-488. 7. Y. Oshima, J. Saito and H. Hikino, Tetrahedron, 1986, 42, 6443-6446. 8. Y. Oshima, J.-I. SAITO and H. HIKINO, Phytochemistry, 1988, 27, 624-626. 9. K. Yasukawa, A. Yamaguchi, J. Arita, S. Sakurai, A. Ikeda and M. Takido, Phytotherapy Research,
1993, 7, 185-189. 10. A. D. Kinghorn, Stevia The Genus Stevia, CRC Press, 2001. 11. J. Brandle, Agriculture and Agri-Food Canada, Editon edn., 2004, Retrieved 2006. 12. M. Bridel and R. Lavielle, Academie des Sciences Paris Comptes Rendus (Parts 192), Editon edn.,
1931, pp. 1123–1125. 13. C. Hellfritsch, A. Brockhoff, F. Stähler, W. Meyerhof and T. Hofmann, Journal of Agricultural and
Food Chemistry, 2012, 60, 6782-6793. 14. H.U. Mitchell, Sweeteners and Sugar Alternatives in Food Technology, Wiley, 2008. 15. C. Trocho, R. Pardo, I. Rafecas, J. Virgili, X. Remesar, J. A. Fernández-López and M. Alemany, Life
Sciences, 1998, 63, 337-349. 16. B. A. Magnuson, G. A. Burdock, J. Doull, R. M. Kroes, G. M. Marsh, M. W. Pariza, P. S. Spencer,
W. J. Waddell, R. Walker and G. M. Williams, in Crit Rev Toxicol, United States, Editon edn., 2007, vol. 37, pp. 629-727.
17. A. G. Renwick, Food Additives and Contaminants, 2006, 23, 327-338. 18. V. L. Grotz and I. C. Munro, Regulatory Toxicology and Pharmacology, 2009, 55, 1-5. 19. I. M. Baird, N. W. Shephard, R. J. Merritt and G. Hildick-Smith, Food and Chemical Toxicology,
2000, 38, Supplement 2, 123-129. 20. Saccharin warning AP via Telegraph-Herald. 1973-05-22. Retrieved 2011-06-09. 21. J. Whysner and G. M. Williams, Pharmacology & Therapeutics, 1996, 71, 225-252. 22. E. Dybing, Toxicology, 2002, 181–182, 121-125. 23. R. L. Myers and R. L. Myers, The 100 most important chemical compounds: a reference guide,
Westport, Conn: Greenwood Press. 24. L. J. Navia, R. E. Walkup, N. M. Vernon, D. S. Neiditch, Production of sucralose without
intermediate isolation of crystalline sucralose-6-ester, US Patent 536/124, Mar 12, 1996. 25. L. Burnett, Artificial Sweetener controversies from saccharin to sucralose. leda.law.harvard.edu.
Retrieved 20 December 2010 26. H. Misra, M. Soni, N. Silawat, D. Mehta, B. Mehta and D. Jain, J Pharm Bioallied Sci, Editon edn.,
2011, vol. 3, pp. 242-248. 27. J. Hawke, in Nexus magazine, Editon edn., February–March 2003, Retrieved 20 December 2010,
vol. 10. 28. J. M. Pezzuto, C. M. Compadre, S. M. Swanson, D. Nanayakkara and A. D. Kinghorn, Proceedings
of the National Academy of Sciences, 1985, 82, 2478-2482. 29. P. J., Elsevier, New Trends in Natural Products Chemistry, Editon edn., 1986, pp. 371-385. 30. E. Procinska, B. A. Bridges and J. R. Hanson, Mutagenesis, 1991, 6, 165-167.
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 108
Jacobs University Bremen
31. H. R. Compadre CM, Nanayakkara NP, Pezzuto JM, Kinghorn AD., Biomed Environ Mass Spectrom, Editon edn., 1988, vol. 15, pp. 211-222.
32. M. Matsui, K. Matsui, Y. Kawasaki, Y. Oda, T. Noguchi, Y. Kitagawa, M. Sawada, M. Hayashi, T. Nohmi, K. Yoshihira, M. Ishidate and T. Sofuni, Mutagenesis, 1996, 11, 573-579.
33. C. Toskulkao, L. Chaturat, P. Temcharoen and T. Glinsukon, Drug Chem Toxicol, 1997, 20, 31-44. 34. L. Xili, B. Chengjiany, X. Eryi, S. Reiming, W. Yuengming, S. Haodong, H. Zhiyian, Food Chem
Toxicol, 1992, 30, 957-965. 35. Y. Aze, K. Toyoda, K. Imaida, S. Hayashi, T. Imazawa, Y. Hayashi, M. Takahashi, Eisei Shikenjo
Hokoku, 1991, 48-54. 36. K. Toyoda, H. Matsui, T. Shoda, C. Uneyama, K. Takada, M. Takahashi, Food and Chemical
Toxicology, 1997, 35, 597-603. 37. R. M. Oliveira-Filho, O. A. Uehara, C. A. Minetti and L. B. Valle, Gen Pharmacol, 1989, 20, 187-
191. 38. C. Wasuntarawat, P. Temcharoen, C. Toskulkao, P. Mungkornkarn, M. Suttajit, T. Glinsukon,
Drug Chem Toxicol, 1998, 21, 207-222. 39. Official Journal of the European Union, Editon edn., 11 November 2011, vol. 1131/2011, p. 26. 40. V. Chatsudthipong and C. Muanprasat, Pharmacology & Therapeutics, 2009, 121, 41-54. 41. C. Toskulkao, M. Sutheerawatananon, C. Wanichanon, P. Saitongdee and M. Suttajit, Journal of
Nutritional Science and Vitaminology, 1995, 41, 105-113. 42. J. M. C. Geuns, J. Buyse, A. Vankeirsbilck, E. H. M. Temme, F. Compernolle and S. Toppet, Journal
of Agricultural and Food Chemistry, 2006, 54, 2794-2798. 43. G. Brahmachari, L. C. Mandal, R. Roy, S. Mondal and A. K. Brahmachari, Archiv der Pharmazie,
2011, 344, 5-19. 44. E. Koyama, K. Kitazawa, Y. Ohori, O. Izawa, K. Kakegawa, A. Fujino, Food and Chemical
Toxicology, 2003, 41, 359-374. 45. A. Wheeler, A. C. Boileau, P. C. Winkler, J. C. Compton, I. Prakash, X. Jiang, D. A. Mandarino,
Food and Chemical Toxicology, 2008, 46, S54-S60. 46. N. Totté, L. Charon, M. Rohmer, F. Compernolle, I. Baboeuf and J. M. C. Geuns, Tetrahedron
Letters, 2000, 41, 6407-6410. 47. N. Totté, W. V. d. Ende, E. J. M. Van Damme, F. Compernolle, I. Baboeuf, J. M. C. Geuns,
Canadian Journal of Botany, 2003, 81, 517-522. 48. J. E. Brandle, A. Richman, A. K. Swanson and B. P. Chapman, Plant Molecular Biology, 2002, 50,
613-622. 49. T. Humphrey, A. Richman, R. Menassa and J. Brandle, Plant Molecular Biology, 2006, 61, 47-62. 50. J. E. Brandle and P. G. Telmer, Phytochemistry, 2007, 68, 1855-1863. 51. A. Hemmerlin, J. L. Harwood and T. J. Bach, Progress in Lipid Research, 2012, 51, 95-148. 52. J. Metivier and A. M. Viana, Journal of Experimental Botany, 1979, 30, 805-810. 53. P. Mauri, G. Catalano, C. Gardana and P. Pietta, ELECTROPHORESIS, 1996, 17, 367-371. 54. P. Nishiyama, M. Alvarez and L. G. E. Vieira, Journal of the Science of Food and Agriculture, 1992,
59, 277-281. 55. H. Mizukami, K. Shiiba and H. Ohashi, Phytochemistry, 1982, 21, 1927-1930. 56. M. Jamaluddin Ahmed and R. M. Smith, Journal of Separation Science, 2002, 25, 170-172. 57. U. Woelwer-Rieck, C. Lankes, A. Wawrzun and M. Wüst, European Food Research and
Technology, 2010, 231, 581-588. 58. H. C. Makapugay, N. P. D. Nanayakkara and A. D. Kinghorn, Journal of Chromatography A, 1984,
283, 390-395. 59. , P. Karásek, M. Roth, K Beneš K řík Č ký Analytical
and Bioanalytical Chemistry, 2007, 388, 1847-1857. 60. J. Liu and S. F. Y. Li, Journal of Liquid Chromatography, 1995, 18, 1703-1719.
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 109
Jacobs University Bremen
61. T něk Nep ı mp x A m ı nd ček Journal of Food Composition and Analysis, 2001, 14, 383-388.
62. V. J. Y. Minne, F. Compernolle, S. Toppet and J. M. C. Geuns, Journal of Agricultural and Food Chemistry, 2004, 52, 2445-2449.
63. e -Rieck, W. Tomberg and A. Wawrzun, Journal of Agricultural and Food Chemistry, 2010, 58, 12216-12220.
64. B. Zimmermann, U. Woelwer-Rieck and M. Papagiannopoulos, Food Analytical Methods, 2012, 5, 266-271.
65. F. Cacciola, P. Delmonte, K. Jaworska, P. Dugo, L. Mondello and J. I. Rader, Journal of Chromatography A, 2011, 1218, 2012-2018.
66. J. Pol, B. Hohnova and T. Hyotylainen, Journal of Chromatography A, 2007, 1150, 85-92. 67. C. Gardana, M. Scaglianti and P. Simonetti, Journal of Chromatography A, 2010, 1217, 1463-
1470. 68. J.M.C. Geuns, Validated techniques of analysis. In Stevia and Steviol Glycosides,
Euprint:Heverlee, Belgium, 2010; ISBN 978-90-742-53116, p61-113. 69. D. M. W.-D. Eric S. Grumbach , Pamela C. Iraneta , Bonnie Alden , Jeffrey R. Mazzeo, in LCGC
North America, Editon edn., 2004, vol. 22, p. 1010. 70. M. Murata, E. Tanaka, E. Minoura and S. Homma, Bioscience, Biotechnology, and Biochemistry,
2004, 68, 501-507. 71. L. Bravo, Nutrition Reviews, 1998, 56, 317-333. 72. M. Naczk and F. Shahidi, Journal of Chromatography A, 2004, 1054, 95-111. 73. H. Wang, G. Cao and R. L. Prior, Journal of Agricultural and Food Chemistry, 1996, 44, 701-705. 74. M. Naczk and F. Shahidi, Journal of Pharmaceutical and Biomedical Analysis, 2006, 41, 1523-
1542. 75. C. Manach, A. Scalbert, C. Morand, C. Rémésy and L. Jiménez, The American Journal of Clinical
Nutrition, 2004, 79, 727-747. 76. A. Crozier, M. E. J. Lean, M. S. McDonald and C. Black, Journal of Agricultural and Food
Chemistry, 1997, 45, 590-595. 77. A. J. Stewart, S. Bozonnet, W. Mullen, G. I. Jenkins, M. E. J. Lean and A. Crozier, Journal of
Agricultural and Food Chemistry, 2000, 48, 2663-2669. 78. L. Gennaro, C. Leonardi, F. Esposito, M. Salucci, G. Maiani, G. Quaglia and V. Fogliano, Journal of
Agricultural and Food Chemistry, 2002, 50, 1904-1910. 79. N. Balasundram, K. Sundram and S. Samman, Food Chemistry, 2006, 99, 191-203. 80. M. N. Clifford, Journal of the Science of Food and Agriculture, 2000, 80, 1033-1043. 81. M. N. Clifford, K. L. Johnston, S. Knight and N. Kuhnert, Journal of Agricultural and Food
Chemistry, 2003, 51, 2900-2911. 82. M. N. Clifford and J. R. Ramirez-Martinez, Food Chemistry, 1990, 35, 13-21. 83. V. Marques and A. Farah, Food Chemistry, 2009, 113, 1370-1376. 84. J. H. Chen and C.-T. Ho, Journal of Agricultural and Food Chemistry, 1997, 45, 2374-2378. 85. M. Lepelley, G. Cheminade, N. Tremillon, A. Simkin, V. Caillet and J. McCarthy, Plant Science,
2007, 172, 978-996. 86. M. R. Olthof, P. C. H. Hollman and M. B. Katan, The Journal of Nutrition, 2001, 131, 66-71. 87. Y. Kono, K. Kobayashi, S. Tagawa, K. Adachi, A. Ueda, Y. Sawa and H. Shibata, Biochimica Et
Biophysica Acta-General Subjects, 1997, 1335, 335-342. 88. M. R. Luzia, K. C. C. DaPaixao, R. Marcilio, L. C. Trugo, L. M. C. Quinteiro and C. A. B. DeMaria,
International Journal of Food Science and Technology, 1997, 32, 15-19. 89. A. W. Herling, H. J. Burger, D. Schwab, H. Hemmerle, P. Below and G. Schubert, American Journal
of Physiology-Gastrointestinal and Liver Physiology, 1998, 274, G1087-G1093.
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 110
Jacobs University Bremen
90. J. Shearer, A. Farah, T. de Paulis, D. P. Bracy, R. R. Pencek, T. E. Graham and D. H. Wasserman, Journal of Nutrition, 2003, 133, 3529-3532.
91. H. Hemmerle, H. J. Burger, P. Below, G. Schubert, R. Rippel, P. W. Schindler, E. Paulus and A. W. Herling, Journal of Medicinal Chemistry, 1997, 40, 137-145.
92. W. J. Arion, W. K. Canfield, F. C. Ramos, P. W. Schindler, H. J. Burger, H. Hemmerle, G. Schubert, P. Below and A. W. Herling, Archives of Biochemistry and Biophysics, 1997, 339, 315-322.
93. K. L. Johnston, M. N. Clifford and L. M. Morgan, American Journal of Clinical Nutrition, 2003, 78, 728-733.
94. M. D. del Castillo, J. M. Ames and M. H. Gordon, Journal of Agricultural and Food Chemistry, 2002, 50, 3698-3703.
95. W. E. Robinson, M. Cordeiro, S. AbdelMalek, Q. Jia, S. A. Chow, M. G. Reinecke and W. M. Mitchell, Molecular Pharmacology, 1996, 50, 846-855.
96. W. E. Robinson, M. G. Reinecke, S. AbdelMalek, Q. Jia and S. A. Chow, Proceedings of the National Academy of Sciences of the United States of America, 1996, 93, 6326-6331.
97. B. McDougall, P. J. King, B. W. Wu, Z. Hostomsky, M. G. Reinecke and W. E. Robinson, Antimicrobial Agents and Chemotherapy, 1998, 42, 140-146.
98. A. Trute, J. Gross, E. Mutschler and A. Nahrstedt, Planta Medica, 1997, 63, 125-129. 99. H. F. Stich, M. P. Rosin and L. Bryson, Mutation Research, 1982, 95, 119-128. 100. N. Hanlon, N. Coldham, A. Gielbert, N. Kuhnert, M. J. Sauer, L. J. Kingi and C. Ioannides, British
Journal of Nutrition, 2008, 99, 559-564. 101. A. Stalmach, W. Mullen, D. Barron, K. Uchida, T. Yokota, C. Cavin, H. Steiling, G. Williamson and
A. Crozier, Drug Metabolism and Disposition, 2009, 37, 1749-1758. 102. A. Farah, M. Monteiro, C. M. Donangelo and S. Lafay, Journal of Nutrition, 2008, 138, 2309-2315. 103. H. M. Merken and G. R. Beecher, Journal of Agricultural and Food Chemistry, 2000, 48, 577-599. 104. P.-G. Pietta, Journal of Natural Products, 2000, 63, 1035-1042. 105. E. de Rijke, P. Out, W. M. A. Niessen, Journal of Chromatography A, 2006, 1112, 31-63. 106. A z nić F Buc Ž M eš A M n B Nig ić Molecules, Molecules, Editon edn., 2009,
vol. 14, pp. 2466-2490. 107. P. C. H. Hollman and M. B. Katan, Food and Chemical Toxicology, 1999, 37, 937-942. 108. I. Ignat, I. Volf and V. I. Popa, Food Chemistry, 2011, 126, 1821-1835. 109. I.-S. Kim, M. Yang, O.-H. Lee and S.-N. Kang, LWT - Food Science and Technology, 2011, 44, 1328-
1332. 110. S. Ghanta, A. Banerjee, A. Poddar and S. Chattopadhyay, Journal of Agricultural and Food
Chemistry, 2007, 55, 10962-10967. 111. U. Wölwer-Rieck, Journal of Agricultural and Food Chemistry, 2012, 60, 886-895. 112. G. Li, R. Wang, A. J. Quampah, Z. Rong, C. Shi and J. Wu, Journal of Agricultural and Food
Chemistry, 2011, 59, 13065-13071. 113. J. M. González-Buitrago, L. Ferreira, M. Isidoro-García, C. Sanz, F. Lorente and I. Dávila, Clinica
Chimica Acta, 2007, 385, 21-27. 114. S. Rogers, M. Girolami, W. Kolch, K. M. Waters, T. Liu, B. Thrall and H. S. Wiley, Bioinformatics,
2008, 24, 2894-2900. 115. V. Dhingra, M. Gupta, T. Andacht and Z. F. Fu, International Journal of Pharmaceutics, 2005, 299,
1-18. 116. C. M. D. Tal Isaacson, Ramu S Saravanan,Yonghuae, Carmen Catala, Montserrat Saladie &
Jocelyn KC Rose, Nature protocol, Editon edn., 13 July 2006. 117. R. M. LoPachin, R. C. Jones, T. A. Patterson, W. Slikker Jr and D. S. Barber, NeuroToxicology,
2003, 24, 761-775. 118. J. A. Reynolds and C. Tanford, Proc Natl Acad Sci U S A, 1970, 66, 1002-1007. 119. V. Neuhoff, N. Arold, D. Taube and W. Ehrhardt, Electrophoresis, 1988, 9, 255-262.
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 111
Jacobs University Bremen
120. J. E. Shively, EXS, 2000, 88, 99-117. 121. D. Lin, D. L. Tabb, Biochimica et Biophysica Acta (BBA) - Proteins&Proteomics, 2003, 1646, 1-10. 122. X. Han, A. Aslanian and J. R. Yates Iii, Current Opinion in Chemical Biology, 2008, 12, 483-490. 123. R. Krüger, A. Pfenninger, I. Fournier, M. Glückmann and M. Karas, Analytical Chemistry, 2001,
73, 5812-5821. 124. J. B. Ohlrogge and J. G. Jaworski, Annual Review of Plant Physiology and Plant Molecular Biology,
1997, 48, 109-136. 125. A. Carrasco-Pancorbo, N. Navas-Iglesias and L. Cuadros-Rodríguez, TrAC Trends in Analytical
Chemistry, 2009, 28, 263-278. 126. C. Zhu, P. Hu, Q.-L. Liang, Y.-M. Wang and G.-A. Luo, Chinese Journal of Analytical Chemistry,
2009, 37, 1390-1396. 127. J. Folch, S. Arsove and J. A. Meath, Journal of Biological Chemistry, 1951, 191, 819-831. 128. V. Matyash, G. Liebisch, T. V. Kurzchalia, A. Shevchenko and D. Schwudke, Journal of Lipid
Research, 2008, 49, 1137-1146. 129. A. Nomura, J. Yamada, T. Yarita and T. Maeda, The Journal of Supercritical Fluids, 1995, 8, 329-
333. 130. T. White, S. Bursten, D. Federighi, R. A. Lewis and E. Nudelman, Analytical Biochemistry, 1998,
258, 109-117. 131. K. Eder, Journal of Chromatography B: Biomedical Sciences and Applications, 1995, 671, 113-
131. 132. X. Han and R. W. Gross, in J Lipid Res, United States, Editon edn., 2003, vol. 44, pp. 1071-1079. 133. C. Hu, R. van der Heijden, M. Wang, J. van der Greef, T. Hankemeier and G. Xu, Journal of
Chromatography B, 2009, 877, 2836-2846. 134. K. U. Borstel René, LCGC chromatography online, Application note, Editon edn., 2011. 135. M. N. Clifford, S. Knight and N. Kuhnert, Journal of Agricultural and Food Chemistry, 2005, 53,
3821-3832. 136. A. J. Day, F. Mellon, D. Barron, G. Sarrazin, M. R. A. Morgan, Free Radical Research, 2001, 35,
941-952. 137. A. Plazonic, F. Bucar, Z. Males, A. Mornar, B. Nigovic and N. Kujundzic, in Molecules, Switzerland,
Editon edn., 2009, vol. 14, pp. 2466-2490. 138. F. Cuyckens and M. Claeys, Journal of Mass Spectrometry, 2004, 39, 1-15. 139. Y.L. Ma, Q.M. Li, H. Van den Heuvel, M. Claeys, Rapid Communications in Mass Spectrometry,
1997, 11, 1357-1364. 140. J. M. Harnly, R. F. Doherty, G. R. Beecher, J. M. Holden, D. B. Haytowitz, S. Bhagwat, S. Gebhardt,
Journal of Agricultural and Food Chemistry, 2006, 54, 9966-9977. 141. C. J. Schwartz, The mixed-model ANOVA: The truth, the computer packages, the books. Part I:
Balanced data, The American Statistician, 1993. 142. L. Zelles and Q. Y. Bai, Soil Biology & Biochemistry, 1993, 25, 495-507. 143. G. Gutnikov, Journal of Chromatography B-Biomedical Applications, 1995, 671, 71-89. 144. A. Martelli, C. Frattini and F. Chialva, Flavour and Fragrance Journal, 1985, 1, 3-7. 145. A. L. Thompson, A. Michalik, R. J. Nash, F. X. Wilson, R. van Well, P. Johnson, G. W. J. Fleet, C. Y.
Yu, X. G. Hu, R. I. Cooper and D. J. Watkin, Acta Crystallographica Section E-Structure Reports Online, 2009, 65, O2904-U2232.
146. G. B. A. Görg, O. Drews, A. Köpf, C. Lück, G. Reil, W. Weiss, A Laboratory Manual, Technische Universität München, Editon edn., 2003.
147. A. Shevchenko, H. Tomas, J. Havlis, J. V. Olsen and M. Mann, Nat. Protocols, 2007, 1, 2856-2860. 148. V. Egelhofer, K. Büssow, C. Luebbert, H. Lehrach and E. Nordhoff, Analytical Chemistry, 2000, 72,
2741-2750.
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 112
Jacobs University Bremen
149. B. Thiede, W. Höhenwarter, A. Krah, J. Mattow, M. Schmid, F. Schmidt and P. R. Jungblut, Methods, 2005, 35, 237-247.
150. J. L. Panero and V. A. Funk, Molecular Phylogenetics and Evolution, 2008, 47, 757-782.
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 113
Jacobs University Bremen
APPENDIX
A. Tandem mass spectra of steviol glycosides in negative ion mode.
Stevioside m/z 803 [M-H+]
-
Steviolbioside m/z 641 [M-H+]
-
317.0
479.1 -MS2(641.2)
317.0 -MS3(641.6->479.4)
-MS4(641.6->479.4->317.1)
0
1
2
7 x10 Intens.
0.0
0.5
1.0 7 x10
-1
0
1
2
200 400 600 800 1000 m/z
641.2 -MS2(803.4)
317.0
479.1 -MS3(803.8->641.2)
317.0 -MS4(803.8->641.5->479.3)
0.0 0.5 1.0
7 x10 Intens.
0 2 4 6
6 x10
0
1
2 6 x10
200 400 600 800 1000 m/z
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 114
Jacobs University Bremen
Rebaudioside F m/z 935 [M-H+]
-
Dulcoside A m/z 787 [M-H+]
- m/z 823[M+Cl
-]-
Rubusoside m/z 641[M-H+]
-
317.1
479.1 -MS2(641.2)
317.0
0
2
7 x10 Intens.
0.0
0.5
1.0
7 x10
-1 0 1 2
200 400 600 800 1000 m/z
-MS4(641.6->479.4->317.2)
-MS3(641.6->479.4)
625.3 -MS2(823.4)
317.0
479.1 -MS3(824.0->625.2)
317.0 -MS4(824.0->625.5->479.3)
0.0
0.5
1.0 7 x10
Intens.
0
2
4 6 x10
0 2 4 6
5 x10
200 400 600 800 1000 m/z
773.4 -MS2(935.5)
317.1 413.1 479.1
611.2 -MS3(936.0->773.3),
317.0
479.1 -MS4(936.0->773.7->613.4)
0.0
0.5
1.0 6 x10
Intens.
0 1 2 3 5 x10
0 2 4 6 4 x10
200 400 600 800 1000 m/z
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 115
Jacobs University Bremen
B. Quantification data of steviol glycosides
Bar plot showing minimum, maximum and average values obtained from all 166 stevia samples
within within +/- 3 σ
0.000
2.000
4.000
6.000
8.000
10.000
12.000
14.000
16.000
18.000
20.000
RebA stevioside DulcosideA Rubusoside RebC sum
min
average
max
allsamples
Stdev(σ) 0.741 3.077 0.152 0.080 0.154 3.200
g/100g
leaves RebA Stevioside DulcosideA Rubusoside RebC sum
min 0.079 0.252 0.005 0.005 0.026 0.554
average 1.017 7.314 0.264 0.122 0.319 9.036
max 5.336 17.509 0.680 0.459 0.820 18.067
Quantity of Steviol Glycosides
Hande Karaköse 116
Jacobs University Bremen
Sample No. Origin Variety Harvesting Year RebA
( g/100g)
Stevioside
(g/100g)
DulcosideA
( g/100g)
Rubusoside
( g/100g)
RebC
(g/100g)
Sum
( g/100g)
1 TCV 3 II 10.09.2010 0.596 6.168 0.140 0.128 0.196 7.229
2 TCV 4 II 14.09.2010 1.411 7.781 0.263 0.113 0.409 9.975
3 TCV 3 I 04.08.2010 1.203 7.980 0.299 0.086 0.383 9.952
4 TCV 1 I 04.08.2010 1.400 8.197 0.132 0.138 0.447 10.313
5 TCV 1 I 04.08.2010 1.592 7.608 0.103 0.101 0.518 9.921
6 TCV 2 I 04.08.2010 1.983 2.962 0.186 0.042 0.366 5.538
7 TCV 2 I 04.08.2010 1.959 2.863 0.104 0.049 0.476 5.450
8 TCV 4 I 04.08.2010 1.591 4.773 0.273 0.118 0.409 7.164
9 TCV 4 II 14.09.2010 1.007 5.390 0.300 0.128 0.297 7.122
10 TCV 2 II 28.09.2010 2.154 7.360 0.179 0.107 0.399 10.199
11 TCV 2 II 28.09.2010 1.607 6.803 0.210 0.101 0.308 9.030
12 Pojava 4 II 14.09.2010 0.565 6.316 0.235 0.083 0.211 7.411
13 TCV 3 I 04.08.2010 0.632 5.648 0.254 0.080 0.215 6.827
14 TCV 2 I 04.08.2010 2.058 3.585 0.073 0.038 0.345 6.100
15 TCV 1 I 01.09.2010 1.001 5.178 0.075 0.072 0.256 6.582
16 TCV 2 I 04.08.2010 1.569 3.314 0.073 0.035 0.503 5.494
17 TCV 3 I 04.08.2010 0.717 5.769 0.217 0.048 0.223 6.975
18 TCV 4 I 04.08.2010 0.800 5.322 0.169 0.047 0.248 6.586
19 TCV 3 II 10.09.2010 0.550 6.003 0.178 0.097 0.178 7.006
20 TCV 2 II 28.09.2010 1.622 4.889 0.078 0.051 0.316 6.956
21 TCV 1 II 28.09.2010 0.853 7.590 0.117 0.073 0.322 8.955
22 TCV 3 II 10.09.2010 0.362 7.808 0.308 0.136 0.154 8.768
23 TCV,Pojana 4 I 04.08.2010 0.640 6.297 0.165 0.048 0.254 7.404
24 TCV,pojana 3 I 04.08.2010 0.292 7.049 0.380 0.074 0.150 7.945
25 TCV 1 I 01.09.2010 0.923 5.300 0.065 0.075 0.256 6.619
26 TCV 4 II 14.09.2010 0.738 5.554 0.154 0.063 0.209 6.719
27 TCV 4 I 04.08.2010 0.657 5.138 0.158 0.041 0.196 6.190
Quantity of Steviol Glycosides
Hande Karaköse 117
Jacobs University Bremen
Sample No. Origin Variety Harvesting Year RebA Stevioside DulcosideA Rubusoside RebC sum
28 TCV 2 II 28.09.2010 1.669 4.870 0.266 0.058 0.820 7.684
29 TCV 3 II 10.09.2010 0.437 6.549 0.255 0.125 0.169 7.535
30 Uconor 4 II 07.05.2010 0.562 5.699 0.209 0.065 0.185 6.720
31 Uconor 3 I 11.08.2010 0.337 3.769 0.297 0.075 0.160 4.638
32 Uconor 2 I 11.08.2010 1.955 3.881 0.048 0.034 0.407 6.325
33 Uconor 2 II 07.05.2010 1.672 4.530 0.060 0.045 0.503 6.810
34 Uconor 4 II 07.09.2010 0.722 5.217 0.124 0.064 0.202 6.328
35 Uconor 4 I 11.08.2010 0.404 4.799 0.280 0.071 0.143 5.696
36 Uconor 4 I 11.08.2010 0.619 3.780 0.104 0.045 0.195 4.745
37 Uconor 4 I 11.08.2010 1.032 5.085 0.087 0.047 0.286 6.537
38 Uconor 3 I 11.08.2010 0.540 6.065 0.235 0.076 0.201 7.117
39 Uconor 4 I 11.08.2010 0.797 3.773 0.122 0.044 0.223 4.959
40 Uconor 2 I 11.08.2010 1.471 2.947 0.075 0.041 0.479 5.015
41 Uconor 3 I 11.08.2010 0.664 5.185 0.139 0.057 0.268 6.313
42 Uconor 3 II 07.09.2010 0.214 5.704 0.281 0.125 0.127 6.451
43 Agrinion 2 II 20.09.2010 1.561 4.908 0.173 0.080 0.325 7.047
44 Toumpa 1 I 30.07.2010 0.444 4.896 0.206 0.083 0.166 5.795
45 Portugal 1 I 07.07.2010 0.600 6.732 0.051 0.064 0.252 7.700
46 Amfilia 1 I 04.08.2010 0.266 5.301 0.240 0.078 0.133 6.018
47 Toumpa 3 II 10.09.2010 0.365 4.107 0.126 0.054 0.143 4.794
48 Agrinion 2 I 09.08.2010 1.048 4.093 0.133 0.053 0.242 5.568
49 Toumpa 3 I 30.07.2010 0.355 5.134 0.190 0.071 0.146 5.895
50 Agrinion 4 I 09.08.2010 0.789 3.533 0.064 0.026 0.251 4.664
51 Agrinion 4 II 20.09.2010 1.034 4.449 0.143 0.066 0.243 5.935
52 Portugal 4 I 26.06.2010 0.563 4.573 0.124 0.103 0.188 5.551
53 Amfilia 4 I 04.08.2010 0.434 4.741 0.200 0.071 0.155 5.601
54 Amfilia 4 II 15.09.2010 0.491 4.528 0.147 0.094 0.200 5.460
55 Argentinie 2009 - - 2009 0.514 5.264 0.289 0.220 0.216 6.504
Quantity of Steviol Glycosides
118
Sample No. Origin Variety Harvesting Year RebA Stevioside DulcosideA Rubusoside RebC sum
56 Paragan2009 - - 2009 0.373 3.750 0.178 0.162 0.200 4.663
57 Argentinien2010 - - 2010 0.552 7.212 0.337 0.240 0.252 8.593
58 Amfilia 2 II 15.09.2010 1.125 5.026 0.119 0.107 0.302 6.680
59 Agrinion 1 II 20.09.2010 1.044 7.864 0.348 0.116 0.362 9.735
60 Amfilia 2 I 04.08.2010 1.114 6.732 0.265 0.083 0.323 8.518
61 Amfilia 3 II 15.09.2010 0.770 6.073 0.184 0.127 0.328 7.481
62 Toumpa 1 I 30.07.2010 0.444 4.396 0.567 0.211 0.210 5.829
63 Amfilia 1 II 15.09.2010 0.695 3.428 0.366 0.187 0.257 4.933
64 Portugal 3 I 26.06.2010 1.768 6.869 0.546 0.339 0.578 10.100
65 Toumpa 1 II 10.09.2010 0.703 3.745 0.455 0.173 0.366 5.442
66 Agrinion 1 I 09.08.2010 0.673 3.569 0.357 0.107 0.327 5.033
67 Amfilia 3 I 04.08.2010 1.100 7.131 0.190 0.097 0.366 8.884
68 Toumpa 2 II 10.09.2010 0.419 8.115 0.458 0.384 0.190 9.566
69 Agrinio 3 II 20.09.2010 0.910 9.167 0.488 0.151 0.274 10.990
70 Toumpa 4 II 10.09.2010 0.733 7.944 0.342 0.123 0.268 9.410
71 Agrinion 3 I 09.08.2010 0.693 7.053 0.327 0.123 0.292 8.488
72 Toumpa 4 I 30.07.2010 0.806 5.995 0.222 0.060 0.243 7.326
73 Granada 2 I 15.09.2010 1.728 6.826 0.073 0.110 0.486 9.224
74 Granada 4 I 15.09.2010 0.779 6.875 0.261 0.080 0.332 8.327
75 Granada 3 I 09.09.2010 0.484 7.916 0.406 0.099 0.225 9.130
76 Equador - - - 0.647 7.599 0.005 0.005 0.067 8.323
77 DZ - - - 0.319 7.042 0.048 0.050 0.082 7.541
78 Kenia 2010 - - 2010 0.615 7.866 0.019 0.051 0.123 8.674
79 Paraguay 2009 - - 2009 0.193 10.950 0.078 0.049 0.094 11.364
80 Argentinien 2010 - - 2010 0.185 12.898 0.110 0.055 0.100 13.349
81 Indien 2010 - - 2010 0.556 14.084 0.054 0.053 0.162 14.907
82 Argenitinien 2009 - - 2009 0.245 12.551 0.091 0.048 0.026 12.962
83 Hohenheim 2010 - - 2010 0.443 17.509 0.074 0.016 0.026 18.067
Quantity of Steviol Glycosides
119
Sample No. Origin Variety Harvesting Year RebA Stevioside DulcosideA Rubusoside RebC Sum
84 Krim - - - 0.190 0.252 0.026 0.046 0.040 0.554
85 Turkei PS1 2010 - - 2010 0.500 12.696 0.028 0.021 0.214 13.282
86 Turkei PS2 2010 - - 2010 0.446 8.683 0.022 0.016 0.196 9.202
87 Uconor 7 I 13.07.2011 0.523 1.058 0.041 0.012 0.029 1.664
88 Agrinon 4 I 2011 0.721 2.416 0.049 0.011 0.031 3.228
89 Uconor 4 I 2011 0.079 4.269 0.299 0.107 0.188 4.943
90 Agrinion 3 I 2011 0.421 8.658 0.068 0.016 0.049 9.210
91 Agrinion 5 I 2011 0.396 3.335 0.177 0.061 0.272 4.241
92 Uconor 6 I 13.07.2011 0.455 2.968 0.124 0.066 0.191 3.804
93 Uconor 3 I 13.07.2011 0.151 3.060 0.306 0.056 0.075 3.648
94 Uconor 5 I 13.07.2011 0.639 2.590 0.137 0.041 0.170 3.577
95 Agrinion 6 I 2011 0.790 3.817 0.151 0.055 0.282 5.096
96 Toumpa 1 I 2011 0.576 5.840 0.372 0.182 0.263 7.234
97 Toumpa 2 I 2011 1.101 4.465 0.254 0.203 0.282 6.305
98 Toumpa 3 I 2011 0.501 5.638 0.356 0.171 0.168 6.833
99 Toumpa 4 I 2011 0.943 4.585 0.228 0.127 0.306 6.189
100 Amiflikeia 4 II 2011 1.603 6.729 0.563 0.103 0.530 9.529
101 Amiflikeia 5 II 2011 1.611 10.507 0.275 0.148 0.533 13.073
102 Amiflikeia 4 II 2011 1.753 10.559 0.403 0.141 0.452 13.309
103 Amiflikeia 3 II 2011 1.088 10.636 0.491 0.211 0.374 12.799
104 Amiflikeia - II 2011 0.802 11.020 0.306 0.146 0.305 12.578
105 Amiflikeia 6 II 2011 1.827 9.539 0.282 0.169 0.512 12.330
106 Amiflikeia 1 II 2011 1.053 10.517 0.457 0.233 0.337 12.597
107 APTTB 3 I 2011 1.029 10.030 0.420 0.149 0.307 11.935
108 APTTB 5 I 2011 1.101 9.921 0.350 0.154 0.333 11.859
109 APTTB 6 I - 1.282 8.761 0.268 0.164 0.358 10.833
110 APTTB 4 I - 1.078 8.955 0.286 0.140 0.398 10.856
111 Uconor 5 II 2011 1.133 8.765 0.337 0.108 0.444 10.786
Quantity of Steviol Glycosides
120
Sample No. Origin Variety Harvesting Year RebA Stevioside DulcosideA Rubusoside RebC sum
112 Uconor 4 II 2011 1.589 7.906 0.180 0.086 0.417 10.178
113 Uconor 3 II 2011 0.646 8.769 0.375 0.207 0.266 10.263
114 Uconor 6 II 2011 1.110 10.151 0.297 0.101 0.473 12.133
115 Conaga 3 I 2011 0.694 9.683 0.413 0.100 0.285 11.175
116 Conaga 5 I 2011 1.856 8.011 0.247 0.055 0.429 10.599
117 Conaga 4 I 2011 0.997 10.457 0.395 0.098 0.384 12.331
118 Agrinion 5 II 2011 1.601 9.733 0.349 0.149 0.467 12.298
119 Agrinion 4 II 2011 1.173 9.625 0.351 0.148 0.379 11.676
120 Conaga 6 I 2011 2.222 8.436 0.346 0.064 0.587 11.654
121 Conaga 7 I 2011 4.901 1.148 0.680 0.021 0.680 7.429
122 Agrinion 6 II 2011 1.397 9.888 0.306 0.147 0.547 12.285
123 Agrinion 3 II 2011 1.140 10.231 0.490 0.181 0.434 12.477
124 Toumpa 7 I 2011 5.336 0.966 0.050 0.018 0.711 7.082
125 Toumpa 4 III 2011 1.000 10.844 0.405 0.159 0.326 12.734
126 Toumpa 3 III 2011 0.897 11.795 0.596 0.271 0.340 13.898
127 Toumpa 3 II 2011 0.743 10.671 0.524 0.249 0.330 12.517
128 Agrinion 4 II 2011 1.447 8.980 0.337 0.155 0.461 11.380
129 Toumpa 4 II 2011 0.934 9.641 0.335 0.191 0.376 11.477
130 Toumpa 3 II 2011 0.722 10.616 0.547 0.190 0.303 12.377
131 Toumpa 4 II 2011 1.235 9.892 0.349 0.146 0.466 12.089
132 Agrinion 3 II 2011 0.956 9.886 0.383 0.222 0.407 11.855
133 Toumpa 5 II 2011 0.770 10.641 0.427 0.194 0.387 12.420
134 Toumpa 6 II 2011 1.030 9.752 0.290 0.159 0.464 11.695
135 Toumpa 7 II 2011 4.590 1.515 0.017 0.023 0.599 6.745
136 TCV 5 I 30.06.2011 0.836 7.947 0.139 0.165 0.366 9.454
137 TCV 3 I 30.06.2011 0.525 9.503 0.411 0.224 0.281 10.943
138 TCV 6 I 30.06.2011 0.824 7.083 0.222 0.142 0.377 8.648
139 TCV 3 II 11.08.2011 0.642 9.780 0.498 0.184 0.246 11.350
Quantity of Steviol Glycosides
Hande Karaköse 121
Jacobs University Bremen
Sample No. Origin Variety Harvesting Year RebA Stevioside DulcosideA Rubusoside RebC sum
140 TCV 6 II 17.08.2011 0.922 8.626 0.302 0.116 0.537 10.504
141 TCV 4 II 24.08.2011 1.090 8.332 0.285 0.125 0.405 10.236
142 TCV 7 I 18.08.2011 3.729 1.099 0.019 0.031 0.613 5.491
143 TCV 4 I 07.07.2011 0.670 8.299 0.291 0.129 0.440 9.828
144 TCV 5 II 17.08.2011 1.215 10.728 0.410 0.127 0.594 13.073
145 Amiflikeia 3 I 2011 0.574 10.429 0.490 0.207 0.269 11.970
146 Amiflikeia 4 I 2011 1.167 10.350 0.310 0.195 0.422 12.444
147 Amiflikeia 5 I 2011 1.440 10.736 0.297 0.230 0.456 13.159
148 Amiflikeia 6 I 2011 1.233 10.463 0.355 0.225 0.430 12.706
149 Amiflikeia 1 I 2011 0.771 12.365 0.529 0.383 0.287 14.335
150 Amiflikeia 2 I 2011 1.175 11.315 0.397 0.334 0.425 13.646
151 Amiflikeia 4 I 2011 0.867 11.958 0.501 0.312 0.299 13.937
152 Amiflikeia 3 I 2011 0.711 10.987 0.623 0.459 0.313 13.093
153 Toumpa 3 I 2011 0.704 11.494 0.580 0.354 0.249 13.381
154 Toumpa 5 I 2011 0.902 10.066 0.316 0.160 0.387 11.831
155 Toumpa 6 I 2011 1.033 7.601 0.185 0.093 0.381 9.293
156 Toumpa 4 I 2011 1.233 10.828 0.336 0.296 0.441 13.133
157 Toumpa 4 I 2011 1.060 9.909 0.352 0.132 0.368 11.821
158 Toumpa 3 I 2011 0.720 10.251 0.379 0.184 0.289 11.823
159 TCV 6 III 2011 1.222 9.232 0.326 0.169 0.464 11.413
160 TCV 5 III 2011 1.069 10.200 0.389 0.191 0.493 12.341
161 TCV 3 III 2011 0.819 10.650 0.478 0.176 0.361 12.485
162 Agrinion 3 III 2011 1.239 10.607 0.425 0.142 0.495 12.908
163 TCV 4 III 2011 0.608 8.321 0.342 0.165 0.296 9.731
164 Agrinion 6 III 2011 1.720 12.198 0.455 0.198 0.713 15.283
165 Agrinion 5 III 2011 1.365 12.007 0.393 0.170 0.635 14.570
166 Agrinion 4 III 2011 1.778 12.302 0.540 0.224 0.704 15.549
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 122
Jacobs University Bremen
Co-elution of steviol glycosides in the UV-chromatogram at 210 nm (LC-MS measurement
with Knauer amino column):
Co-elution of rebaudioside A with stevioside or rebaudioside B:
Co-elution of rebaudioside C with dulcoside A:
196.0300.2
401.2 518.2
803.4
965.5
1001.4
-MS, 11.2min #670
0.0
0.2
0.4
0.6
0.8
1.0
5x10
Intens.
200 400 600 800 1000 m/z
179.1
247.0
300.2359.1
406.1
480.2
787.4
949.5
985.4
-MS, 7.6min #452
0
1
2
3
4x10
Intens.
200 400 600 800 1000 m/z
1
1
2
2
3
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 123
Jacobs University Bremen
Co-elution of stevioside with steviolbioside:
Extracted ion chromatogram of rebaudioside A (HILIC)
641.3
803.4
-MS, 6.1min #366
0
2
4
6
8
5x10
Intens.
200 400 600 800 1000 m/z
389.3
457.2
528.2
965.4
1011.4
1079.4
-MS, 27.0min #1612
0
1
2
3
5x10
Intens.
200 400 600 800 1000 m/z
3
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 124
Jacobs University Bremen
C. Tandem mass spectra of CGAs and flavonoid glycosides in negative ion mode
3-caffeoylquinic acid m/z 353 [M-H+]
-
5-caffeoylquinic acid m/z 353 [M-H+]
-
190.7 -MS2(352.9)
93.0 172.7
126.8 -MS3(353.1->190.7)
83.4 -MS4(353.1->190.8->126.9)
0 2 4 6 7 x10
Intens.
0 1 2 3 5 x10
0.0
0.5
1.0 4 x10
200 400 600 800 1000 m/z
134.8
190.7 -MS2(352.9)
126.7 172.8 85.1
-MS3(353.1->190.6)
-MS4(353.1->190.7->85.4)
0.0
0.5
1.0 7 x10
Intens.
0 2 4 6 4 x10
-1
0
1
2
200 400 600 800 1000 m/z
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 125
Jacobs University Bremen
4-caffeoylquinic acid m/z 353 [M-H+]
-
Cis-5-caffeoylquinic acid m/z 353 [M-H+]
-
3,5-dicaffeoylquinic acid m/z 515
190.7
352.9 -MS2(515.0)
134.7
190.7 -MS3(515.3->352.9)
85.1
126.8
172.7
-MS4(515.3->353.1->190.7)
0.00 0.25 0.50
8 x10 Intens.
0
1
2
7 x10
0
1
2 5 x10
100 200 300 400 500 600 700 m/z
190.7 -MS2(352.9)
85.0 172.7
126.8 -MS3(353.1->190.7)
108.9 -MS4(353.1->190.7->126.8)
0 2 4
6 x10 Intens.
0
2
4 4 x10
0
500
1000
200 400 600 800 1000 m/z
172.7 -MS2(352.9)
71.3 154.7
93.0 -MS3(353.1->172.7)
-MS4(353.1->172.7->93.1)
0.0
0.5
7 x10 Intens.
0.0
0.5
1.0
5 x10
-1 0 1
2
200 400 600 800 1000 m/z
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 126
Jacobs University Bremen
4,5-dicaffeoylquinic acid m/z 515
Cis-4,5-dicaffeoylquinic acid m/z 515
Cis-4,5-dicaffeoylquinic acid m/z 515
172.7
352.9 -MS2(515.0)
134.8
172.7 -MS3(515.3->353.0)
71.3 93.1
110.9
154.7 -MS4(515.3->353.1->172.8)
0.0 0.5 1.0
7 x10 Intens.
0
1
2 6 x10
0.0
0.5
1.0
4 x10
100 200 300 400 500 600 700 m/z
172.7 202.8
352.9 -MS2(515.0)
134.8
172.7 -MS3(515.2->352.9)
71.3
93.0
110.9 154.7
-MS4(515.2->353.0->172.8)
0.00 0.25 0.50 0.75
7 x10 Intens.
0.0 0.5 1.0 1.5 2.0
6 x10
0
1
2
4 x10
100 200 300 400 500 600 700 m/z
172.7 202.7 254.8 298.9
352.9 -MS2(515.0)
134.8
172.7 -MS3(515.3->352.9)
71.3
93.0 110.8
154.7
-MS4(515.3->353.1->172.8)
0.00 0.25 0.50
8 x10 Intens.
0.0 0.5 1.0 1.5
7 x10
0
2
4 5 x10
100 200 300 400 500 600 700 m/z
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 127
Jacobs University Bremen
Rutin m/z 609 [M-H+]
-
Quercetin-galactoside m/z 463 [M-H+]
-
178.6
300.8
299.8
-MS2(462.9)
106.9
270.8 178.7
-MS3(463.2->302.6)
150.7
-MS4(463.2->300.7->178.7)
0.0
0.5
7 x10 Intens.
0.00 0.25 0.50 0.75
6 x10
0
2
4
4 x10
200 400 600 800 1000 m/z
178.6 270.8
300.8
342.9
299.8
-MS2(609.0)
106.9
178.7 270.7
-MS3(609.4->300.6)
184.7
243.7 -MS4(609.4->300.6->271.1)
0.0
0.5
1.0 7 x10
Intens.
0.0
0.5
1.0 6 x10
0
1
2
4 x10
200 400 600 800 1000 m/z
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 128
Jacobs University Bremen
Kaempferol-glucopyranoside; Quercetin-rhamnoside; Kaempferol-glucopyranoside;
Quercetin-rhamnoside m/z 447
178.7
300.8
299.9
-MS2(447.0)
106.8
150.7 270.7
178.6 -MS3(447.2->300.7)
106.8
150.7 -MS4(447.2->300.7->178.8)
0.00 0.25 0.50 0.75
7 x10 Intens.
0 2 4 6 5 x10
0
2
4
4 x10
200 400 600 800 1000 m/z
254.7 326.9
283.8 -MS2(446.9)
150.7 226.7
254.7 -MS3(447.1->285.8)
226.8 -MS4(447.1->284.3->255.2)
0.0 0.5
1.0 6 x10
Intens.
0
1
2
5 x10
0 1000 2000 3000 4000
200 400 600 800 1000 m/z
284.8 -MS2(446.9)
242.7
673.9
198.7 -MS3(447.1->284.9)
170.8 -MS4(447.1->284.9->198.8)
0
1
7 x10 Intens.
0.0 0.5 1.0 1.5
5 x10
0 500
1000 1500 2000
200 400 600 800 1000 m/z
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 129
Jacobs University Bremen
Kaempferol-rhamnopyranosyl-glucopyranoside(rutinoside) isomers; Quercetin-
dirhamnoside; Apigenin-diglucoside/galactoside m/z 593
Quercetin pentoside m/z 433
178.7
299.8 -MS2(432.9)
178.7
270.7 -MS3(433.2->300.5)
186.7
242.7 -MS4(433.2->300.5->271.0)
0.0
0.5
7 x10 Intens.
0
1
2
6 x10
0
1
2
4 x10
200 400 600 800 1000 m/z
300.8 -MS2(432.9)
106.9
150.7 270.7
178.7 -MS3(433.1->300.7)
107.2
150.7 -MS4(433.1->300.8->178.7)
0.0
0.5
7 x10 Intens.
0.00 0.25 0.50 0.75
6 x10
0
2
4 4 x10
200 400 600 800 1000 m/z
284.8 -MS2(446.9)
150.7
254.7 -MS3(593.3->284.5)
162.6 210.7
-MS4(593.3->284.5->255.3)
0 2 4 6 6 x10
Intens.
0.00 0.25 0.50 0.75
6 x10
0.0 0.5 1.0 1.5
4 x10
200 400 600 800 1000 m/z
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 130
Jacobs University Bremen
Kaempferol-xylosyl-glucoside; Naringin m/z 579
Apigenin-galactoside m/z 431
268.7 -MS2(430.9)
148.8 267.6
224.8 -MS3(431.2->267.8)
168.7
196.7 -MS4(431.2->268.8->225.3)
0.0
0.5
7 x10 Intens.
0 2 4 6 4 x10
0 500
1000 1500
200 400 600 800 1000 m/z
178.6 342.9 414.9 489.0 560.9
299.8 -MS2(579.0)
150.7
270.7 -MS3(579.3->300.0)
198.6
226.7 -MS4(579.3->300.0->270.9)
0 1 2 3 6 x10
Intens.
0 2 4 6
5 x10
0.00 0.25 0.50 0.75
4 x10
200 400 600 800 1000 m/z
178.6 414.9
299.8 -MS2(579.0)
178.7
270.7 -MS3(579.3->300.1)
243.8 -MS4(579.3->300.1->270.9)
0
2
4 6 x10
Intens.
0 2 4 6 8 5 x10
0 500
1000 1500
200 400 600 800 1000 m/z
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 131
Jacobs University Bremen
Kaempferol-glucosylrhamnosyl-glucoside/galactoside m/z 755
Quercetin-trisaccharide m/z 741
Kaempferol 3-rhamnopyranosyl-rhamnopyranosyl-glucopyranoside m/z 739
254.7 283.8
326.9 393.0 443.0 473.0 642.0 692.0
575.1 -MS2(739.2)
162.8 212.7 256.7
282.9 308.9
339.0 393.0
428.9
547.1 338.1
-MS3(739.5->575.2)
262.8 295.8 -MS4(739.5->575.3->339.0)
0 2
4 4 x10
Intens.
0 1000 2000 3000
0
100
200
100 200 300 400 500 600 700 m/z
299.8 461.9
579.0 -MS2(741.2)
354.9 414.9
299.8 -MS3(741.5->578.5)
270.7 -MS4(741.5->579.2->300.0)
0
1
5 x10 Intens.
0 1 2 3
4 x10
0 250 500 750
200 400 600 800 1000 m/z
284.8 367.0 469.0
593.0 -MS2(755.2)
254.7 326.9
283.8
-MS3(755.7->593.3)
150.7
254.7 -MS4(755.7->593.3->284.5)
0 1 2 3 6 x10
Intens.
0 2 4 6 5 x10
0 2 4 6 4 x10
200 400 600 800 1000 m/z
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 132
Jacobs University Bremen
Quercetin-diglucoside-rhamnoside m/z 771
300.8 469.0
609.1 -MS2(771.2)
270.8 342.9
299.8
-MS3(771.6->608.6)
106.9
150.7
210.7
270.7 -MS4(771.6->609.4->300.9)
0 2 4
6 x10 Intens.
0.0 0.5 1.0 1.5
6 x10
0.0
0.5
1.0 5 x10
200 400 600 800 1000 m/z
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 133
Jacobs University Bremen
D. Quantification data of polyhenols in stevia
Minimum, average and maximum amounts (g/100g leaves) of polyphenols in all stevia samples (average
values taken within +/- 3 σ)
Bar plot showing minimum, maximum and average values obtained from all 166 stevia samples
within +/- 3 σ
0.000
2.000
4.000
6.000
8.000
10.000
12.000
14.000
16.000
18.000
min
ave
max
3CQA 5CQA 4CQA totalmono 3,5 4,5 totaldiCQA k7g q3g totalflavones
min 0.002 0.193 0.010 0.205 0.173 0.210 0.311 0.092 0.001 0.329
ave 0.310 2.481 0.124 2.915 1.203 1.241 2.442 2.854 0.084 6.993
max 2.828 4.986 0.249 5.608 2.476 2.609 4.575 6.662 0.611 16.415
stddev 0.243 0.855 0.043 1.010 0.511 0.487 0.938 1.259 0.086 3.102
Quantity of CQAs and Flavonoid glycosides
Hande Karaköse 134
Jacobs University Bremen
Sample No. Origin Variety Harvesting Year 3cqa
( g/100g)
5cqa
( g/100g)
4cqa
( g/100g)
Totalmono
( g/100g)
3,5diCQA
( g/100g)
4,5diCQA
( g/100g)
Total diCQA
( g/100g)
k7g
( g/100g)
q3g
( g/100g)
Total flav.
( g/100g)
1 TCV 3 II 10.09.2010 0.139 1.847 0.092 2.078 0.804 0.624 1.428 0.787 0.087 1.641
2 TCV 4 II 14.09.2010 0.260 2.680 0.134 3.074 1.281 1.052 2.333 1.967 0.093 3.518
3 TCV 3 I 04.08.2010 0.325 2.900 0.145 3.369 0.512 1.155 1.667 1.541 0.085 3.041
4 TCV 1 I 04.08.2010 0.187 2.336 0.117 2.640 0.954 0.871 1.825 1.006 0.120 2.326
5 TCV 1 I 04.08.2010 0.232 2.197 0.110 2.538 0.735 0.814 1.549 0.609 0.136 1.529
6 TCV 2 I 04.08.2010 0.235 1.806 0.090 2.132 0.592 1.061 1.653 0.899 0.074 2.130
7 TCV 2 I 04.08.2010 0.198 1.961 0.098 2.256 0.670 0.959 1.630 0.987 0.107 2.208
8 TCV 4 I 04.08.2010 0.324 3.306 0.165 3.796 0.938 1.030 1.968 2.255 0.089 3.914
9 TCV 4 II 14.09.2010 0.223 2.911 0.146 3.280 1.413 0.997 2.410 2.290 0.073 4.009
10 TCV 2 II 28.09.2010 0.089 1.102 0.055 1.246 1.178 0.814 1.992 1.575 0.099 3.302
11 TCV 2 II 28.09.2010 0.051 0.593 0.030 0.673 0.514 0.488 1.002 0.582 0.110 1.180
12 Pojava 4 II 14.09.2010 0.221 2.103 0.105 2.429 0.778 0.993 1.771 1.735 0.080 2.696
13 TCV 3 I 04.08.2010 0.319 2.508 0.125 2.952 0.562 0.840 1.403 1.135 0.044 2.207
14 TCV 2 I 04.08.2010 0.253 1.876 0.094 2.223 0.537 0.899 1.437 0.767 0.110 1.507
15 TCV 1 I 01.09.2010 0.217 1.772 0.089 2.077 0.630 0.849 1.480 0.878 0.073 2.015
16 TCV 2 I 04.08.2010 0.232 2.058 0.103 2.393 0.557 0.791 1.348 0.772 0.145 1.627
17 TCV 3 I 04.08.2010 0.290 2.361 0.118 2.769 0.574 0.884 1.459 0.998 0.047 2.246
18 TCV 4 I 04.08.2010 0.317 2.057 0.103 2.477 0.519 0.853 1.372 1.416 0.059 2.481
19 TCV 3 II 10.09.2010 0.233 2.197 0.110 2.541 0.853 0.802 1.655 1.080 0.044 2.299
20 TCV 2 II 28.09.2010 0.136 0.729 0.036 0.901 0.907 1.046 1.953 1.125 0.053 2.522
21 TCV 1 II 28.09.2010 0.041 0.684 0.034 0.759 0.247 0.237 0.484 0.225 0.134 0.626
22 TCV 3 II 10.09.2010 0.276 2.361 0.118 2.755 0.922 1.130 2.052 0.471 0.090 1.103
23 TCV,Pojana 4 I 04.08.2010 0.279 2.615 0.131 3.024 0.492 0.795 1.287 1.210 0.142 2.022
24 TCV,pojana 3 I 04.08.2010 0.257 2.355 0.118 2.730 0.504 0.786 1.290 0.421 0.112 0.958
25 TCV 1 I 01.09.2010 0.198 1.615 0.081 1.894 0.473 0.653 1.126 0.450 0.107 1.097
26 TCV 4 II 14.09.2010 0.199 1.651 0.083 1.933 0.694 0.912 1.606 1.518 0.008 2.679
27 TCV 4 I 04.08.2010 0.282 2.223 0.111 2.617 0.435 0.694 1.128 1.303 0.079 2.252
28 TCV 2 II 28.09.2010 0.220 1.833 0.092 2.145 1.331 1.476 2.807 2.254 0.044 5.026
Quantity of CQAs and Flavonoid glycosides
Hande Karaköse 135
Jacobs University Bremen
Sample No. Origin Variety Harvesting Year 3cqa 5cqa 4cqa totalmono 3,5diCQA 4,5diCQA total diCQA k7g q3g Total flav.
29 TCV 3 II 10.09.2010 0.260 2.531 0.127 2.917 0.700 0.884 1.585 1.269 0.094 2.604
30 Uconor 4 II 07.05.2010 0.249 1.630 0.081 1.961 0.606 1.000 1.606 0.798 0.106 3.707
31 Uconor 3 I 11.08.2010 0.380 3.068 0.153 3.601 1.890 1.966 3.856 4.745 0.036 11.645
32 Uconor 2 I 11.08.2010 0.253 2.320 0.116 2.689 2.014 2.561 4.575 0.889 0.415 4.503
33 Uconor 2 II 07.05.2010 0.161 1.751 0.088 2.000 2.144 1.921 4.066 1.828 0.287 6.359
34 Uconor 4 II 07.09.2010 0.341 4.344 0.217 4.903 2.129 1.544 3.673 3.495 0.265 7.592
35 Uconor 4 I 11.08.2010 0.337 4.067 0.203 4.607 1.413 1.735 3.148 2.509 0.611 8.135
36 Uconor 4 I 11.08.2010 0.262 3.586 0.179 4.027 1.190 1.987 3.178 1.943 0.445 5.796
37 Uconor 4 I 11.08.2010 0.181 1.931 0.097 2.209 0.804 0.714 1.518 2.227 0.118 6.225
38 Uconor 3 I 11.08.2010 0.347 3.096 0.155 3.598 0.887 1.108 1.995 2.644 0.521 7.170
39 Uconor 4 I 11.08.2010 0.314 3.034 0.152 3.500 1.687 1.668 3.355 4.876 0.060 13.156
40 Uconor 2 I 11.08.2010 0.291 3.099 0.155 3.544 2.476 1.916 4.392 2.894 0.079 8.897
41 Uconor 3 I 11.08.2010 0.215 1.692 0.085 1.992 0.938 1.028 1.966 2.158 0.025 5.087
42 Uconor 3 II 07.09.2010 0.364 2.681 0.134 3.179 1.803 1.342 3.145 3.136 0.064 7.190
43 Agrinion 2 II 20.09.2010 0.386 2.634 0.132 3.152 1.695 1.681 3.376 3.204 0.067 6.682
44 Toumpa 1 I 30.07.2010 0.348 3.086 0.154 3.589 2.276 2.251 4.526 6.662 0.079 16.415
45 Portugal 1 I 07.07.2010 0.117 0.908 0.045 1.071 0.273 0.569 0.842 2.883 0.096 9.525
46 Amfilia 1 I 04.08.2010 0.719 4.135 0.207 5.061 1.333 1.391 2.725 3.026 0.047 7.179
47 Toumpa 3 II 10.09.2010 0.423 2.908 0.145 3.477 1.498 1.362 2.861 3.991 0.010 11.230
48 Agrinion 2 I 09.08.2010 0.456 2.808 0.140 3.404 0.406 1.328 1.734 2.551 0.025 6.536
49 Toumpa 3 I 30.07.2010 0.413 3.810 0.191 4.414 1.311 1.583 2.893 3.599 0.037 9.241
50 Agrinion 4 I 09.08.2010 0.542 3.346 0.167 4.056 1.122 1.551 2.673 4.589 0.019 8.169
51 Agrinion 4 II 20.09.2010 0.118 1.355 0.068 1.541 0.726 0.844 1.570 4.043 0.079 7.890
52 Portugal 4 I 26.06.2010 0.267 1.706 0.085 2.059 0.637 0.460 1.098 3.828 0.023 8.718
53 Amfilia 4 I 04.08.2010 0.284 3.049 0.152 3.485 0.452 0.704 1.157 3.423 0.146 5.716
54 Amfilia 4 II 15.09.2010 0.260 2.315 0.116 2.691 0.647 0.842 1.489 4.584 0.086 7.711
55 Argentinie 2009 - - 2009 0.284 2.403 0.120 2.808 1.442 1.641 3.083 3.250 0.008 7.656
56 Paragan2009 - - 2009 0.380 3.021 0.151 3.551 1.842 1.574 3.417 4.226 0.041 8.672
Quantity of CQAs and Flavonoid glycosides
Hande Karaköse 136
Jacobs University Bremen
Sample No. Origin Variety Harvesting Year 3cqa 5cqa 4cqa totalmono 3,5diCQA 4,5diCQA total diCQA k7g q3g Total flav.
57 Argentinien2010 - - 2010 0.288 2.138 0.107 2.533 1.357 1.427 2.784 3.420 0.009 8.073
58 Amfilia 2 II 15.09.2010 0.439 2.440 0.122 3.001 0.655 0.755 1.410 4.145 0.040 8.847
59 Agrinion 1 II 20.09.2010 0.123 1.341 0.067 1.531 1.035 1.245 2.280 3.638 0.043 7.423
60 Amfilia 2 I 04.08.2010 0.619 4.740 0.237 5.597 1.505 1.493 2.997 3.839 0.012 9.389
61 Amfilia 3 II 15.09.2010 0.407 2.143 0.107 2.658 0.914 1.110 2.024 3.792 0.113 8.905
62 Toumpa 1 I 30.07.2010 0.444 3.439 0.172 4.055 1.262 1.781 3.043 4.508 0.008 9.830
63 Amfilia 1 II 15.09.2010 0.317 1.932 0.097 2.346 0.642 0.714 1.356 4.884 0.112 15.769
64 Portugal 3 I 26.06.2010 0.223 2.125 0.106 2.454 0.754 0.738 1.492 4.666 0.035 13.528
65 Toumpa 1 II 10.09.2010 0.289 2.705 0.135 3.129 1.443 1.019 2.461 3.851 0.002 9.324
66 Agrinion 1 I 09.08.2010 0.374 3.483 0.174 4.031 1.249 1.487 2.737 4.356 0.078 9.066
67 Amfilia 3 I 04.08.2010 0.659 3.931 0.197 4.787 1.037 1.337 2.374 4.316 0.027 9.249
68 Toumpa 2 II 10.09.2010 0.366 3.010 0.151 3.527 1.053 0.914 1.967 2.577 0.058 6.943
69 Agrinio 3 II 20.09.2010 0.186 1.465 0.073 1.724 0.821 1.047 1.868 2.284 0.001 6.544
70 Toumpa 4 II 10.09.2010 0.442 3.975 0.199 4.615 1.331 0.868 2.200 3.638 0.088 7.843
71 Agrinion 3 I 09.08.2010 0.413 3.266 0.163 3.843 1.008 1.282 2.290 3.586 0.011 9.980
72 Toumpa 4 I 30.07.2010 0.591 3.768 0.188 4.547 1.522 2.381 3.903 2.278 0.087 4.869
73 Granada 2 I 15.09.2010 0.154 1.350 0.068 1.572 0.692 0.630 1.322 2.190 0.034 7.131
74 Granada 4 I 15.09.2010 0.160 1.962 0.098 2.220 0.369 0.559 0.928 4.286 0.063 8.816
75 Granada 3 I 09.09.2010 0.214 1.671 0.084 1.969 0.173 0.511 0.684 4.275 0.068 9.516
76 Equador - - - 0.015 0.308 0.015 0.338 0.325 0.210 0.535 0.099 0.135 0.350
77 DZ - - - 0.002 0.193 0.010 0.205 0.344 0.263 0.607 1.352 0.140 3.531
78 Kenia 2010 - - 2010 0.261 2.436 0.122 2.818 1.509 1.114 2.622 1.625 0.082 6.527
79 Paraguay 2009 - - 2009 0.450 2.772 0.139 3.361 2.033 1.895 3.928 4.301 0.036 10.225
80 Argentinien 2010 - - 2010 0.209 2.762 0.138 3.109 1.889 2.059 3.948 4.459 0.096 10.975
81 Indien 2010 - - 2010 0.122 1.322 0.066 1.510 0.790 1.757 2.547 2.175 0.122 8.091
82 Argenitinien 2009 - - 2009 0.354 2.488 0.124 2.966 2.019 2.526 4.546 6.294 0.005 13.673
83 Hohenheim 2010 - - 2010 0.123 1.133 0.057 1.313 0.266 0.450 0.716 3.298 0.030
7.328
Quantity of CQAs and Flavonoid glycosides
Hande Karaköse 137
Jacobs University Bremen
Sample No. Origin Variety Harvesting Year 3cqa 5cqa 4cqa totalmono 3,5diCQA 4,5diCQA total diCQA k7g q3g Total flav.
84 Krim - - - 0.033 1.183 0.059 1.275 0.549 0.238 0.311 2.969 0.191 7.441
85 Turkei PS1 2010 - - 2010 0.089 1.626 0.081 1.796 0.774 1.066 1.840 3.177 0.174 8.623
86 Turkei PS2 2010 - - 2010 0.120 1.917 0.096 2.132 0.496 0.900 1.395 2.246 0.294 5.762
87 Uconor 7 I 13.07.2011 0.365 3.645 0.182 4.193 1.786 2.045 3.831 5.159 0.026 9.928
88 Agrinon 4 I 2011 0.431 3.399 0.170 4.000 1.203 1.461 2.664 1.512 0.067 3.939
89 Uconor 4 I 2011 0.473 3.279 0.164 3.916 1.720 1.900 3.620 3.915 0.054 9.788
90 Agrinion 3 I 2011 0.398 3.040 0.152 3.590 1.245 1.436 2.681 3.913 0.056 10.088
91 Agrinion 5 I 2011 0.463 3.208 0.160 3.832 1.174 1.400 2.574 3.030 0.047 8.126
92 Uconor 6 I 13.07.2011 0.561 4.671 0.234 5.465 1.947 2.290 4.237 0.092 0.048 0.329
93 Uconor 3 I 13.07.2011 0.372 4.986 0.249 5.608 1.664 2.022 3.686 3.102 0.088 7.721
94 Uconor 5 I 13.07.2011 0.520 3.282 0.164 3.967 1.425 1.588 3.013 1.764 0.141 5.377
95 Agrinion 6 I 2011 2.828 1.225 0.061 4.114 1.592 1.956 3.548 2.220 0.111 6.236
96 Toumpa 1 I 2011 0.479 3.600 0.180 4.260 1.321 1.578 2.899 3.210 0.059 8.628
97 Toumpa 2 I 2011 0.495 3.203 0.160 3.858 0.909 1.137 2.046 3.659 0.061 10.518
98 Toumpa 3 I 2011 0.491 4.099 0.205 4.795 1.208 1.535 2.743 3.972 0.002 10.938
99 Toumpa 4 I 2011 0.461 2.879 0.144 3.484 1.883 2.052 3.935 3.709 0.091 7.981
100 Amiflikeia 4 II 2011 0.595 3.762 0.188 4.544 1.193 1.435 2.628 3.610 0.058 7.079
101 Amiflikeia 5 II 2011 0.434 2.756 0.138 3.328 0.968 1.410 2.378 3.248 0.163 7.859
102 Amiflikeia 4 II 2011 0.371 2.363 0.118 2.852 1.528 1.556 3.084 3.852 0.092 7.591
103 Amiflikeia 3 II 2011 0.479 2.814 0.141 3.433 1.623 1.790 3.413 3.248 0.022 7.671
104 Amiflikeia - II 2011 0.458 2.723 0.136 3.317 1.125 1.161 2.286 3.309 0.123 7.305
105 Amiflikeia 6 II 2011 0.451 2.574 0.129 3.154 1.099 1.166 2.264 2.884 0.073 7.167
106 Amiflikeia 1 II 2011 0.535 2.200 0.110 2.845 1.349 1.416 2.765 2.641 0.019 6.327
107 APTTB 3 I 2011 0.246 1.663 0.083 1.992 0.721 0.603 1.324 3.675 0.041 9.730
108 APTTB 5 I 2011 0.192 1.249 0.062 1.503 0.705 0.703 1.409 3.821 0.096 11.274
109 APTTB 6 I - 0.180 1.177 0.059 1.416 0.756 0.701 1.457 4.118 0.018 10.068
110 APTTB 4 I - 0.206 1.362 0.068 1.637 0.800 0.803 1.603 3.516 0.055 10.562
111 Uconor 5 II 2011 0.237 2.303 0.115 2.656 1.837 1.190 3.027 3.066 0.009 8.264
Quantity of CQAs and Flavonoid glycosides
Hande Karaköse 138
Jacobs University Bremen
Sample No. Origin Variety Harvesting Year 3cqa 5cqa 4cqa totalmono 3,5diCQA 4,5diCQA total diCQA k7g q3g Total flav.
112 Uconor 4 II 2011 0.293 2.063 0.103 2.459 1.406 1.275 2.681 3.578 0.035 7.598
113 Uconor 3 II 2011 0.300 3.108 0.155 3.564 1.553 1.525 3.077 2.840 0.054 6.933
114 Uconor 6 II 2011 0.301 2.170 0.109 2.580 1.482 1.228 2.711 3.110 0.025 8.264
115 Conaga 3 I 2011 0.356 2.621 0.131 3.107 0.901 1.506 2.408 3.041 0.030 7.221
116 Conaga 5 I 2011 0.250 2.422 0.121 2.792 1.216 1.422 2.638 2.195 0.097 6.593
117 Conaga 4 I 2011 0.354 2.309 0.115 2.778 1.045 1.524 2.569 3.227 0.067 7.027
118 Agrinion 5 II 2011 0.368 2.410 0.121 2.899 1.495 1.379 2.874 3.227 0.051 9.412
119 Agrinion 4 II 2011 0.358 3.144 0.157 3.659 1.433 1.514 2.947 3.564 0.068 7.308
120 Conaga 6 I 2011 0.326 3.301 0.165 3.792 1.295 2.068 3.362 3.428 0.088 10.269
121 Conaga 7 I 2011 0.262 2.747 0.137 3.147 0.507 1.189 1.696 0.637 0.225 2.064
122 Agrinion 6 II 2011 0.296 2.461 0.123 2.880 1.434 1.364 2.798 2.788 0.009 7.335
123 Agrinion 3 II 2011 0.247 2.626 0.131 3.004 1.570 1.342 2.913 4.336 0.019 9.887
124 Toumpa 7 I 2011 0.061 2.240 0.112 2.414 1.528 0.575 2.103 1.873 0.140 5.930
125 Toumpa 4 III 2011 0.148 1.884 0.094 2.126 1.311 1.310 2.621 3.680 0.060 8.463
126 Toumpa 3 III 2011 0.187 1.961 0.098 2.246 1.684 1.567 3.251 3.510 0.018 9.388
127 Toumpa 3 II 2011 0.175 2.490 0.124 2.789 1.650 1.271 2.921 3.769 0.013 9.366
128 Agrinion 4 II 2011 0.246 3.222 0.161 3.630 1.500 1.460 2.960 3.386 0.005 6.759
129 Toumpa 4 II 2011 0.155 2.863 0.143 3.161 1.725 1.426 3.151 4.079 0.142 8.424
130 Toumpa 3 II 2011 0.140 2.078 0.104 2.322 1.573 1.464 3.037 3.937 0.025 10.501
131 Toumpa 4 II 2011 0.112 2.289 0.114 2.515 1.321 1.288 2.609 4.684 0.057 9.631
132 Agrinion 3 II 2011 0.258 2.593 0.130 2.980 1.925 1.922 3.846 3.585 0.019 8.969
133 Toumpa 5 II 2011 0.134 2.375 0.119 2.628 1.732 1.454 3.186 3.827 0.092 11.186
134 Toumpa 6 II 2011 0.130 2.536 0.127 2.792 1.445 0.953 2.398 3.424 0.008 9.629
135 Toumpa 7 II 2011 0.116 2.285 0.114 2.516 0.850 0.657 1.507 1.442 0.179 4.449
136 TCV 5 I 30.06.2011 0.531 3.343 0.167 4.042 1.328 1.185 2.513 2.618 0.076 8.134
137 TCV 3 I 30.06.2011 0.441 3.153 0.158 3.752 1.578 1.266 2.844 2.870 0.034 7.671
138 TCV 6 I 30.06.2011 0.526 4.032 0.202 4.760 1.807 1.394 3.201 3.718 0.031 10.049
139 TCV 3 II 11.08.2011 0.267 2.692 0.135 3.094 1.854 1.155 3.010 3.503 0.068 8.909
Quantity of CQAs and Flavonoid glycosides
Hande Karaköse 139
Jacobs University Bremen
Sample No. Origin Variety Harvesting Year 3cqa 5cqa 4cqa totalmono 3,5diCQA 4,5diCQA total diCQA k7g q3g Total flav.
140 TCV 6 II 17.08.2011 0.289 2.399 0.120 2.808 1.679 1.520 3.200 3.105 0.062 8.506
141 TCV 4 II 24.08.2011 0.376 2.375 0.119 2.869 1.563 2.609 4.171 2.493 0.192 5.320
142 TCV 7 I 18.08.2011 0.084 1.777 0.089 1.950 1.715 1.412 3.127 1.711 0.208 4.790
143 TCV 4 I 07.07.2011 0.340 1.843 0.092 2.275 1.070 1.232 2.302 2.974 0.009 6.305
144 TCV 5 II 17.08.2011 0.385 2.323 0.116 2.824 1.933 2.474 4.407 2.781 0.031 7.833
145 Amiflikeia 3 I 2011 0.501 2.934 0.147 3.582 0.433 0.928 1.360 2.114 0.075 5.301
146 Amiflikeia 4 I 2011 0.554 3.595 0.180 4.329 1.182 0.989 2.170 2.013 0.128 4.423
147 Amiflikeia 5 I 2011 0.628 3.086 0.154 3.869 1.326 1.124 2.450 2.036 0.001 5.264
148 Amiflikeia 6 I 2011 0.524 3.398 0.170 4.092 1.353 0.735 2.089 2.035 0.044 5.307
149 Amiflikeia 1 I 2011 0.485 3.690 0.184 4.359 1.836 1.674 3.510 2.939 0.052 8.505
150 Amiflikeia 2 I 2011 0.399 2.798 0.140 3.337 2.183 1.078 3.262 2.966 0.048 7.316
151 Amiflikeia 4 I 2011 0.358 2.465 0.123 2.946 1.903 1.234 3.138 4.101 0.011 8.871
152 Amiflikeia 3 I 2011 0.335 2.154 0.108 2.596 1.842 1.119 2.961 3.351 0.086 8.778
153 Toumpa 3 I 2011 0.307 2.097 0.105 2.508 1.939 1.677 3.616 3.622 0.001 10.033
154 Toumpa 5 I 2011 0.242 3.266 0.163 3.672 1.694 1.281 2.975 2.500 0.099 6.702
155 Toumpa 6 I 2011 0.127 2.546 0.127 2.800 1.054 0.770 1.824 1.953 0.068 5.836
156 Toumpa 4 I 2011 0.242 2.434 0.122 2.797 2.178 1.794 3.972 4.547 0.092 9.592
157 Toumpa 4 I 2011 0.307 2.551 0.128 2.985 1.560 1.357 2.918 3.952 0.065 8.032
158 Toumpa 3 I 2011 0.171 2.763 0.138 3.072 1.485 0.871 2.356 3.479 0.095 8.160
159 TCV 6 III 2011 0.063 1.872 0.094 2.029 1.716 1.789 3.505 3.010 0.036 7.856
160 TCV 5 III 2011 0.147 1.552 0.078 1.777 1.622 1.176 2.798 2.464 0.091 6.536
161 TCV 3 III 2011 0.161 1.777 0.089 2.026 0.987 1.077 2.064 2.300 0.157 5.554
162 Agrinion 3 III 2011 0.168 1.372 0.069 1.608 0.974 0.938 1.912 3.132 0.041 8.298
163 TCV 4 III 2011 0.120 2.087 0.104 2.311 1.110 0.883 1.993 2.346 0.194 4.654
164 Agrinion 6 III 2011 0.148 2.041 0.102 2.291 1.489 1.147 2.636 4.011 0.006 10.699
165 Agrinion 5 III 2011 0.124 1.897 0.095 2.116 1.269 0.794 2.063 3.501 0.064 9.157
166 Agrinion 4 III 2011 0.179 1.837 0.092 2.108 0.982 0.865 1.847 4.156 0.118 8.261
Quantity of Trans & Cis-CQAs
Hande Karaköse 140
Jacobs University Bremen
Sample No. Origin Variety Harvesting 3cqa
( g/100g)
5cqa
( g/100g)
4cqa
( g/100g)
Cis-5CQA
(g/100g)
3,5diCQA
( g/100g)
4,5diCQA
( g/100g)
Cis-4,5diCQA
(g/100g)
Cis-4,5diCQA
(g/100g)
1 TCV 3 II 0.139 1.847 0.092 0.313 0.804 0.624 0.002 0.022
2 TCV 4 II 0.260 2.680 0.134 0.516 1.281 1.052 0.011 0.037
3 TCV 3 I 0.325 2.900 0.145 1.153 0.512 1.155 0.009 0.035
4 TCV 1 I 0.187 2.336 0.117 0.060 0.954 0.871 0.009 0.045
5 TCV 1 I 0.232 2.197 0.110 0.518 0.735 0.814 0.006 0.024
6 TCV 2 I 0.235 1.806 0.090 0.590 0.592 1.061 0.009 0.030
7 TCV 2 I 0.198 1.961 0.098 0.593 0.670 0.959 0.007 0.028
8 TCV 4 I 0.324 3.306 0.165 1.123 0.938 1.030 0.016 0.031
9 TCV 4 II 0.223 2.911 0.146 0.657 1.413 0.997 0.006 0.026
10 TCV 2 II 0.089 1.102 0.055 0.121 1.178 0.814 0.006 0.018
11 TCV 2 II 0.051 0.593 0.030 0.023 0.514 0.488 0.023 0.057
12 Pojava 4 II 0.221 2.103 0.105 0.112 0.778 0.993 0.006 0.009
13 TCV 3 I 0.319 2.508 0.125 0.102 0.562 0.840 0.035 0.061
14 TCV 2 I 0.253 1.876 0.094 0.110 0.537 0.899 0.010 0.013
15 TCV 1 I 0.217 1.772 0.089 0.081 0.630 0.849 0.009 0.013
16 TCV 2 I 0.232 2.058 0.103 0.076 0.557 0.791 0.059 0.089
17 TCV 3 I 0.290 2.361 0.118 0.126 0.574 0.884 0.033 0.043
18 TCV 4 I 0.317 2.057 0.103 0.134 0.519 0.853 0.008 0.018
19 TCV 3 II 0.233 2.197 0.110 0.056 0.853 0.802 0.034 0.048
20 TCV 2 II 0.136 0.729 0.036 0.042 0.907 1.046 0.005 0.012
21 TCV 1 II 0.041 0.684 0.034 0.001 0.247 0.237 0.020 0.028
22 TCV 3 II 0.276 2.361 0.118 0.110 0.922 1.130 0.010 0.005
23 TCV,Pojana 4 I 0.279 2.615 0.131 0.100 0.492 0.795 0.051 0.078
24 TCV,pojana 3 I 0.257 2.355 0.118 0.101 0.504 0.786 0.013 0.024
25 TCV 1 I 0.198 1.615 0.081 0.065 0.473 0.653 0.037 0.051
26 TCV 4 II 0.199 1.651 0.083 0.073 0.694 0.912 0.007 0.018
27 TCV 4 I 0.282 2.223 0.111 0.113 0.435 0.694 0.014 0.013
28 TCV 2 II 0.220 1.833 0.092 0.070 1.331 1.476 0.010 0.014
Quantity of Trans & Cis-CQAs
Hande Karaköse 141
Jacobs University Bremen
Sample No. Origin Variety Harvesting 3cqa 5cqa 4cqa cis 5CQA 3,5diCQA 4,5diCQA Cis 4,5-diCQA
(g/100g)
Cis 4,5-diCQA
(g/100g)
29 TCV 3 II 0.260 2.531 0.127 0.090 0.700 0.884 0.018 0.018
30 Uconor 4 II 0.249 1.630 0.081 0.075 0.606 1.000 0.004 0.004
31 Uconor 3 I 0.380 3.068 0.153 0.021 1.890 1.966 0.011 0.011
32 Uconor 2 I 0.253 2.320 0.116 0.037 2.014 2.561 0.023 0.023
33 Uconor 2 II 0.161 1.751 0.088 0.015 2.144 1.921 0.026 0.026
34 Uconor 4 II 0.341 4.344 0.217 0.043 2.129 1.544 0.035 0.035
35 Uconor 4 I 0.337 4.067 0.203 0.067 1.413 1.735 0.006 0.006
36 Uconor 4 I 0.262 3.586 0.179 0.019 1.190 1.987 0.008 0.008
37 Uconor 4 I 0.181 1.931 0.097 0.017 0.804 0.714 0.009 0.009
38 Uconor 3 I 0.347 3.096 0.155 0.023 0.887 1.108 0.016 0.016
39 Uconor 4 I 0.314 3.034 0.152 0.013 1.687 1.668 0.007 0.007
40 Uconor 2 I 0.291 3.099 0.155 0.018 2.476 1.916 0.016 0.016
41 Uconor 3 I 0.215 1.692 0.085 0.029 0.938 1.028 0.013 0.013
42 Uconor 3 II 0.364 2.681 0.134 0.198 1.803 1.342 0.017 0.017
43 Agrinion 2 II 0.386 2.634 0.132 0.138 1.695 1.681 0.045 0.045
44 Toumpa 1 I 0.348 3.086 0.154 0.006 2.276 2.251 0.026 0.026
45 Portugal 1 I 0.117 0.908 0.045 0.028 0.273 0.569 0.004 0.004
46 Amfilia 1 I 0.719 4.135 0.207 0.242 1.333 1.391 0.044 0.044
47 Toumpa 3 II 0.423 2.908 0.145 0.188 1.498 1.362 0.017 0.017
48 Agrinion 2 I 0.456 2.808 0.140 0.128 0.406 1.328 0.029 0.029
49 Toumpa 3 I 0.413 3.810 0.191 0.006 1.311 1.583 0.025 0.025
50 Agrinion 4 I 0.542 3.346 0.167 0.204 1.122 1.551 0.029 0.029
51 Agrinion 4 II 0.118 1.355 0.068 0.040 0.726 0.844 0.018 0.009
52 Portugal 4 I 0.267 1.706 0.085 0.048 0.637 0.460 0.009 0.008
53 Amfilia 4 I 0.284 3.049 0.152 0.113 0.452 0.704 0.009 0.009
54 Amfilia 4 II 0.260 2.315 0.116 0.085 0.647 0.842 0.040 0.018
55 Argentinie 2009 - - 0.284 2.403 0.120 0.038 1.442 1.641 0.005 0.017
56 Paragan2009 - - 0.380 3.021 0.151 0.108 1.842 1.574 0.022 0.024
Quantity of Trans & Cis-CQAs
Hande Karaköse 142
Jacobs University Bremen
Sample No. Origin Variety Harvesting 3cqa 5cqa 4cqa cis 5CQA 3,5diCQA 4,5diCQA Cis 4,5-diCQA
(g/100g)
Cis 4,5-diCQA
(g/100g)
57 Argentinien2010 - - 0.288 2.138 0.107 0.035 1.357 1.427 0.013 0.042
58 Amfilia 2 II 0.439 2.440 0.122 0.107 0.655 0.755 0.012 0.014
59 Agrinion 1 II 0.123 1.341 0.067 0.002 1.035 1.245 0.005 0.025
60 Amfilia 2 I 0.619 4.740 0.237 0.229 1.505 1.493 0.016 0.016
61 Amfilia 3 II 0.407 2.143 0.107 0.087 0.914 1.110 0.017 0.051
62 Toumpa 1 I 0.444 3.439 0.172 0.104 1.262 1.781 0.022 0.025
63 Amfilia 1 II 0.317 1.932 0.097 0.041 0.642 0.714 0.007 0.021
64 Portugal 3 I 0.223 2.125 0.106 0.042 0.754 0.738 0.008 0.012
65 Toumpa 1 II 0.289 2.705 0.135 0.127 1.443 1.019 0.006 0.012
66 Agrinion 1 I 0.374 3.483 0.174 0.155 1.249 1.487 0.038 0.032
67 Amfilia 3 I 0.659 3.931 0.197 0.078 1.037 1.337 0.013 0.004
68 Toumpa 2 II 0.366 3.010 0.151 0.093 1.053 0.914 0.010 0.003
69 Agrinio 3 II 0.186 1.465 0.073 0.036 0.821 1.047 0.008 0.020
70 Toumpa 4 II 0.442 3.975 0.199 0.008 1.331 0.868 0.015 0.016
71 Agrinion 3 I 0.413 3.266 0.163 0.174 1.008 1.282 0.021 0.056
72 Toumpa 4 I 0.591 3.768 0.188 0.093 1.522 2.381 0.042 0.114
73 Granada 2 I 0.154 1.350 0.068 0.043 0.692 0.630 0.009 0.017
74 Granada 4 I 0.160 1.962 0.098 0.058 0.369 0.559 0.001 0.004
75 Granada 3 I 0.214 1.671 0.084 0.024 0.173 0.511 0.007 0.013
76 Equador - - 0.015 0.308 0.015 0.005 0.325 0.210 #DIV/0! #DIV/0!
77 DZ - - 0.002 0.193 0.010 0.001 0.344 0.263 #DIV/0! #DIV/0!
78 Kenia 2010 - - 0.261 2.436 0.122 0.030 1.509 1.114 0.009 0.029
79 Paraguay 2009 - - 0.450 2.772 0.139 0.082 2.033 1.895 0.060 0.092
80 Argentinien 2010 - - 0.209 2.762 0.138 0.004 1.889 2.059 0.025 0.059
81 Indien 2010 - - 0.122 1.322 0.066 0.022 0.790 1.757 0.032 0.063
82 Argenitinien 2009 - - 0.354 2.488 0.124 0.022 2.019 2.526 0.011 0.030
83 Hohenheim 2010 - - 0.123 1.133 0.057 0.027 0.266 0.450 0.000 0.001
Quantity of Trans & Cis-CQAs
Hande Karaköse 143
Jacobs University Bremen
Sample No. Origin Variety Harvesting 3cqa 5cqa 4cqa cis 5CQA 3,5diCQA 4,5diCQA Cis 4,5-diCQA
(g/100g)
Cis 4,5-diCQA
(g/100g)
84 Krim - - 0.033 1.183 0.059 0.020 0.549 0.238 -0.015 0.000
85 Turkei PS1 2010 - - 0.089 1.626 0.081 0.023 0.774 1.066 0.008 0.040
86 Turkei PS2 2010 - - 0.120 1.917 0.096 0.054 0.496 0.900 0.019 0.018
87 Uconor 7 I 0.365 3.645 0.182 0.182 1.786 2.045 0.010 0.014
88 Agrinon 4 I 0.431 3.399 0.170 0.129 1.203 1.461 0.023 0.013
89 Uconor 4 I 0.473 3.279 0.164 0.177 1.720 1.900 0.011 0.018
90 Agrinion 3 I 0.398 3.040 0.152 0.009 1.245 1.436 0.007 0.012
91 Agrinion 5 I 0.463 3.208 0.160 0.158 1.174 1.400 0.015 0.017
92 Uconor 6 I 0.561 4.671 0.234 0.251 1.947 2.290 0.022 0.007
93 Uconor 3 I 0.372 4.986 0.249 0.208 1.664 2.022 0.038 0.013
94 Uconor 5 I 0.520 3.282 0.164 0.196 1.425 1.588 0.050 0.065
95 Agrinion 6 I 2.828 1.225 0.061 0.132 1.592 1.956 0.014 0.002
96 Toumpa 1 I 0.479 3.600 0.180 0.163 1.321 1.578 0.034 0.041
97 Toumpa 2 I 0.495 3.203 0.160 0.148 0.909 1.137 0.048 0.146
98 Toumpa 3 I 0.491 4.099 0.205 0.175 1.208 1.535 0.029 0.075
99 Toumpa 4 I 0.461 2.879 0.144 0.183 1.883 2.052 0.038 0.094
100 Amiflikeia 4 II 0.595 3.762 0.188 0.008 1.193 1.435 0.024 0.033
101 Amiflikeia 5 II 0.434 2.756 0.138 0.149 0.968 1.410 0.005 0.013
102 Amiflikeia 4 II 0.371 2.363 0.118 0.128 1.528 1.556 0.022 0.059
103 Amiflikeia 3 II 0.479 2.814 0.141 0.145 1.623 1.790 0.045 0.112
104 Amiflikeia - II 0.458 2.723 0.136 0.140 1.125 1.161 0.033 0.027
105 Amiflikeia 6 II 0.451 2.574 0.129 0.138 1.099 1.166 0.007 0.010
106 Amiflikeia 1 II 0.535 2.200 0.110 0.128 1.349 1.416 0.030 0.105
107 APTTB 3 I 0.246 1.663 0.083 0.064 0.721 0.603 0.003 0.001
108 APTTB 5 I 0.192 1.249 0.062 0.058 0.705 0.703 0.008 0.018
109 APTTB 6 I 0.180 1.177 0.059 0.053 0.756 0.701 0.010 0.034
110 APTTB 4 I 0.206 1.362 0.068 0.057 0.800 0.803 0.014 0.011
111 Uconor 5 II 0.237 2.303 0.115 0.122 1.837 1.190 0.022 0.040
Quantity of Trans & Cis-CQAs
Hande Karaköse 144
Jacobs University Bremen
Sample No. Origin Variety Harvesting 3cqa 5cqa 4cqa cis 5CQA 3,5diCQA 4,5diCQA Cis 4,5-diCQA
(g/100g)
Cis 4,5-diCQA
(g/100g)
112 Uconor 4 II 0.293 2.063 0.103 0.136 1.406 1.275 0.029 0.058
113 Uconor 3 II 0.300 3.108 0.155 0.146 1.553 1.525 0.029 0.026
114 Uconor 6 II 0.301 2.170 0.109 0.113 1.482 1.228 0.020 0.043
115 Conaga 3 I 0.356 2.621 0.131 0.138 0.901 1.506 0.028 0.018
116 Conaga 5 I 0.250 2.422 0.121 0.157 1.216 1.422 0.018 0.059
117 Conaga 4 I 0.354 2.309 0.115 0.088 1.045 1.524 0.043 0.042
118 Agrinion 5 II 0.368 2.410 0.121 0.135 1.495 1.379 0.021 0.043
119 Agrinion 4 II 0.358 3.144 0.157 0.125 1.433 1.514 0.033 0.074
120 Conaga 6 I 0.326 3.301 0.165 0.004 1.295 2.068 0.104 0.166
121 Conaga 7 I 0.262 2.747 0.137 0.177 0.507 1.189 0.023 0.021
122 Agrinion 6 II 0.296 2.461 0.123 0.103 1.434 1.364 0.025 0.032
123 Agrinion 3 II 0.247 2.626 0.131 0.126 1.570 1.342 0.012 0.011
124 Toumpa 7 I 0.061 2.240 0.112 0.099 1.528 0.575 0.013 0.007
125 Toumpa 4 III 0.148 1.884 0.094 0.094 1.311 1.310 0.055 0.048
126 Toumpa 3 III 0.187 1.961 0.098 0.082 1.684 1.567 0.019 0.064
127 Toumpa 3 II 0.175 2.490 0.124 0.108 1.650 1.271 0.018 0.007
128 Agrinion 4 II 0.246 3.222 0.161 0.197 1.500 1.460 0.048 0.095
129 Toumpa 4 II 0.155 2.863 0.143 0.113 1.725 1.426 0.028 0.024
130 Toumpa 3 II 0.140 2.078 0.104 0.112 1.573 1.464 0.036 0.033
131 Toumpa 4 II 0.112 2.289 0.114 0.100 1.321 1.288 0.014 0.022
132 Agrinion 3 II 0.258 2.593 0.130 0.106 1.925 1.922 0.032 0.080
133 Toumpa 5 II 0.134 2.375 0.119 0.006 1.732 1.454 0.011 0.009
134 Toumpa 6 II 0.130 2.536 0.127 0.145 1.445 0.953 0.025 0.015
135 Toumpa 7 II 0.116 2.285 0.114 0.155 0.850 0.657 0.020 0.004
136 TCV 5 I 0.531 3.343 0.167 0.121 1.328 1.185 0.027 0.027
137 TCV 3 I 0.441 3.153 0.158 0.135 1.578 1.266 0.033 0.029
138 TCV 6 I 0.526 4.032 0.202 0.164 1.807 1.394 0.033 0.029
139 TCV 3 II 0.267 2.692 0.135 0.147 1.854 1.155 0.008 0.015
Quantity of Trans & Cis-CQAs
Hande Karaköse 145
Jacobs University Bremen
Sample No. Origin Variety Harvesting 3cqa 5cqa 4cqa cis 5CQA 3,5diCQA 4,5diCQA Cis 4,5-diCQA
(g/100g)
Cis 4,5-diCQA
(g/100g)
140 TCV 6 II 0.289 2.399 0.120 0.131 1.679 1.520 0.029 0.091
141 TCV 4 II 0.376 2.375 0.119 0.084 1.563 2.609 0.048 0.116
142 TCV 7 I 0.084 1.777 0.089 0.112 1.715 1.412 0.023 0.038
143 TCV 4 I 0.340 1.843 0.092 0.120 1.070 1.232 0.022 0.028
144 TCV 5 II 0.385 2.323 0.116 0.004 1.933 2.474 0.024 0.039
145 Amiflikeia 3 I 0.501 2.934 0.147 0.192 0.433 0.928 0.001 0.024
146 Amiflikeia 4 I 0.554 3.595 0.180 0.206 1.182 0.989 0.021 0.034
147 Amiflikeia 5 I 0.628 3.086 0.154 0.199 1.326 1.124 0.017 0.007
148 Amiflikeia 6 I 0.524 3.398 0.170 0.189 1.353 0.735 0.029 0.012
149 Amiflikeia 1 I 0.485 3.690 0.184 0.093 1.836 1.674 0.032 0.040
150 Amiflikeia 2 I 0.399 2.798 0.140 0.115 2.183 1.078 0.012 0.000
151 Amiflikeia 4 I 0.358 2.465 0.123 0.174 1.903 1.234 0.008 0.017
152 Amiflikeia 3 I 0.335 2.154 0.108 0.115 1.842 1.119 0.014 0.014
153 Toumpa 3 I 0.307 2.097 0.105 0.094 1.939 1.677 0.027 0.034
154 Toumpa 5 I 0.242 3.266 0.163 0.138 1.694 1.281 0.007 0.007
155 Toumpa 6 I 0.127 2.546 0.127 0.127 1.054 0.770 0.008 0.002
156 Toumpa 4 I 0.242 2.434 0.122 0.122 2.178 1.794 0.008 0.008
157 Toumpa 4 I 0.307 2.551 0.128 0.166 1.560 1.357 0.021 0.021
158 Toumpa 3 I 0.171 2.763 0.138 0.117 1.485 0.871 0.007 0.007
159 TCV 6 III 0.063 1.872 0.094 0.006 1.716 1.789 0.030 0.030
160 TCV 5 III 0.147 1.552 0.078 0.008 1.622 1.176 0.024 0.024
161 TCV 3 III 0.161 1.777 0.089 0.084 0.987 1.077 0.008 0.008
162 Agrinion 3 III 0.168 1.372 0.069 0.099 0.974 0.938 0.017 0.017
163 TCV 4 III 0.120 2.087 0.104 0.084 1.110 0.883 0.011 0.011
164 Agrinion 6 III 0.148 2.041 0.102 0.099 1.489 1.147 0.028 0.028
165 Agrinion 5 III 0.124 1.897 0.095 0.107 1.269 0.794 0.015 0.015
166 Agrinion 4 III 0.179 1.837 0.092 0.093 0.982 0.865 0.011 0.011
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 146
Jacobs University Bremen
E. Correlation & data distribution of CQAs
Correlation table of cis and trans CQAs
cis5CQA cis45diCQa1 cis45diCQa2 cqa5 diCQA45
Spearman's
rho
cis5CQA
Correlation Coefficient 1.000 0.265** 0.183* -0.111 -0.075
Sig. (2-tailed) . 0.003 0.041 0.217 0.407
N 126 126 126 126 126
cis45diCQa1
Correlation Coefficient 0.265** 1.000 0.731** -0.012 -0.056
Sig. (2-tailed) 0.003 . 0.000 0.892 0.532
N 126 126 126 126 126
cis45diCQa2
Correlation Coefficient 0.183* 0.731** 1.000 0.035 -0.026
Sig. (2-tailed) 0.041 0.000 . 0.694 0.771
N 126 126 126 126 126
cqa5
Correlation Coefficient -0.111 -0.012 0.035 1.000 0.586**
Sig. (2-tailed) 0.217 0.892 0.694 . 0.000
N 126 126 126 126 126
diCQA45
Correlation Coefficient -0.075 -0.056 -0.026 0.586** 1.000
Sig. (2-tailed) 0.407 0.532 0.771 0.000 .
N 126 126 126 126 126
**. Correlation is significant at the 0.01 level (2-tailed).
*. Correlation is significant at the 0.05 level (2-tailed).
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 147
Jacobs University Bremen
F. Structures of literature reported terpenes in Stevia rebaudiana
OOH
myrtenol myrtenal
OH
pinocarveol alpha pinene beta pinene
sabinene terpinene
OH
terpinen-4-ol
OH
verbenol
O
cumin aldehyde cymene limonene
OH
linalool
OH
geraniol
MeO
anethole
OH
borneol
O
camphor
OH
carvacrol
O
1,8 cineol myrcene
3-carene
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 148
Jacobs University Bremen
nerolidol
OH
selinene
copaenecubebene beta elemene
germacrene D humulene (caryophyllene)
beta-trans-farnesene
bergamotenebisabolene
gamma cadinene delta cadinene
HO
alpha cadinolcalacorene calamenene
H
H
bourbonene
Phytochemical Characterization of Stevia rebaudiana
Hande Karaköse 149
Jacobs University Bremen
PUBLICATIONS
1. H. Karaköse, R. Jaiswal, N. Kuhnert “Characterisation and quantification of hydroxycinnamate
derivatives in Stevia rebaudiana leaves by LC-MSn” J. Agric. Food Chem., 2011, 59 (18),
10143–10150.
2. N. Kuhnert, F. Dairpoosh, R. Jaiswal, M. Matei, S.Deshpande, A.Golon, H. Nour, H.
Karaköse, N. Hourani “Hill coefficients of dietary polyphenolic enzyme inhibitiors: can
beneficial health effects of dietary polyphenols be explained by allosteric enzyme denaturing?” J
Chem Biol. 2011 July; 4(3): 109–116.
3. H. Karaköse, N. Kuhnert “Profiling the chlorogenic acids of Stevia rebaudiana by tandem LC-
MS” Polyphenol Commun. 2010, Vol 1, 544-546. 71.
4. N. Kuhnert, H. Karaköse, R. Jaiswal, “Analysis, characterization and pharmacokinetics of
dietary hydroxycinnamates” Invited review chapter in CRC Handbook of Food Analysis,
manuscript in press
5. G. Mikutis, H. Karaköse, R. Jaiswal, A. Le Gresley, T. Islam, M. Fernandez-Lahore, N.
Kuhnert “Phenolic promiscuity in the cell nucleus” Food and function, in Press, 2012
6. H. Karaköse, R. Jaiswal, S. Deshpande, N. Kuhnert “Investigating the photochemical changes
of chlorogenic acids induced by UV light in model systems and in agricultural practice with
Stevia rebaudiana cultivation as an example” (Manuscript)
7. H. Karaköse, N. Kuhnert “Development of a LC-ESI MS method for the identification and
quantification of steviol glycosides” (Manuscript)
8. H. Karaköse, A. Golon, N. Kuhnert “Gas chromatographic analysis of lipids and volatile
terpenes in Stevia rebaudiana” (In preparation)
9. H. Karaköse, R. Shah, N. Kuhnert “Identification of proteins by MALDI-TOF MS in Stevia
rebaudiana” (In preparation)
HANDE KARAKÖSE
Address: Clamersdorfer str. 21
28757 - Bremen/Germany
Personal Details
Date of Birth 01.11.1985
Place of Birth Ankara, Turkey
Email [email protected]
Mobile +4915208494817
Education
2009 – 2012 PhD in Chemistry Jacobs University Bremen, Germany/Bremen Title: Chemical Profiling of Stevia Rebaudiana Bertoni
• Projects: Identification and profiling of all primary and secondary metabolites of stevia using HPLC-MS, MALDI-TOF and GC-MS. Method development for analysis and quantification of selected compounds and the affect of the growth conditions to the metabolite profile. Extraction of secondary metabolites. Protein extraction and isolation of stevia and identification by MALDI-TOF Lipid and volatile terpene profiling by GC-MS. Statistical analysis (e.g. PCA, ANOVA) of the dataset obtained by LC-MS. Solid phase extraction.
2007 - 2009 Master in Nanomolecular Science Jacobs University Bremen, Germany/Bremen Title: Profiling and Characterisation of Chlorogenic Acids by LC-MSn
• Projects: Optimization of the extraction technique and determination of the main chlorogenic acids in various coffee, plum and potato samples by using an HPLC-MSn method. Isolation of selected chlorogenic acids by preparative LC and identification of novel chlorogenic acids Modelling of chlorogenic acids by computational chemistry methods Comparison of experimental spectroscopic data (NMR chemical shifts, Raman spectra, IR) with the calculated spectrical data obtained by quantum mechanical calculations.
2003 - 2007 Bachelor in Chemistry University of Dokuz Eylül Faculty of Science & Arts, İzmir Title: Precontration and Solid Extraction of Uranium (VI) from various water samples using N,N-Dibutyl-N`-Benzoylthiourea
Achievements & Awards
2007 Graduation with distinction; DEÜ, Faculty of Science & Arts, Chemistry Department
2007 Full Scholarship for Master Education in Jacobs University Bremen
2009 Fellowship for PhD in Jacobs University Bremen
Articles H. Karaköse, R. Jaiswal, N. Kuhnert “Characterisation and quantification of hydroxycinnamate derivatives in Stevia Rebaudiana leaves by LC-MSn” J. Agric. Food Chem., 2011, 59 (18), 10143–10150.
N. Kuhnert, F. Dairpoosh, R. Jaiswal, M. Matei, S.Deshpande, A.Golon, H. Nour, H.
Karaköse, N. Hourani “Hill coefficients of dietary polyphenolic enzyme inhibitiors: can
beneficial health effects of dietary polyphenols be explained by allosteric enzyme
denaturing?” J Chem Biol. 2011 July; 4(3): 109–116.
H. Karaköse, N. Kuhnert “Profiling the chlorogenic acids of Stevia Rebaudiana by tandem LC-MS” Polyphenol Commun. 2010, 1, 544-546. 71
G. Mikutis, H. Karaköse, R. Jaiswal, A. Le Gresley, T. Islam, M. Fernandez-Lahore, N. Kuhnert “Phenolic promiscuity in the cell nucleus” Food and function, in Press, 2012.
H. Karaköse, N. Kuhnert “Development of a LC-ESI MS method for the identification
and quantification of steviol glycosides” (Manuscript)
Books N. Kuhnert, H. Karaköse, R. Jaiswal, “Analysis, characterization and pharmacokinetics of dietary hydroxycinnamates” Invited review chapter in CRC Handbook of Food Analysis.
Conferences & Seminars
Münster/Germany ISC 2008 – 27th International Symposium on Chromatography 21 - 25 September 2008
Münster/Germany HPLC Masterclass Advanced Method Development (LC certification) 6 - 7 Mai 2010
Montpellier/France 25. International Conference on Polyphenols 23 - 27 August 2010
Poster Presentation: Profiling of Isomers of Chlorogenic Acids by LC-MSn
H. Karaköse, N. Kuhnert
Istanbul/Turkey Terpenist 2010 26 - 29 September 2010 Poster Presentation: Sweet Terpenes in Stevia Rebaudiana; H. Karaköse, N. Kuhnert Bremen/Germany GDCh – Wissenschaftsforum Chemie 4 - 7 September 2011
Poster Presentation: Profiling, PCA Analysis and Quantification of Chlorogenic acids in Stevia Rebaudiana; H. Karaköse, N. Kuhnert.
Sitges/Spain 5th International Conference on Polyphenol and Health 17-20 October 2011
Poster Presentation: Characterization of Chlorogenic Acids in Stevia Rebaudiana
Leaves by LC-MSn ; H. Karaköse, N. Kuhnert.
Internships 03.07 - 28.07.2006 Petkim Petrokimya Holding A.Ş. (Petroleum chemicals), Izmir/Turkey
07.02 - 04.03.2006 University of Leipzig, Faculty of Chemistry and Mineralogy Institute of Analytical Chemistry. Leipzig/Germany 27.06 - 05.08.2005 Petkim Petrokimya Holding A.Ş. (Petroleum chemicals), Izmir/Turkey
Languages English (very good) German (good) Skills
Computer Skills Microsoft Windows 7& XP Microsoft Office Programs Linux / Ubuntu Open Office
Teaching Skills During my PhD and Master education, I have led seminars, supervised
undergraduates in the laboratory Regularly supervise practicals for undergraduate students and have supervised the undergraduate research projects of two final year students. I gave several seminars for undergraduates in School of Engineering and Science
Interests Swimming, travelling, reading, photography.
Reference Prof. Nikolai Kuhnert Email: [email protected]
Telephone: +49 421 200-3120
Fax: +49 421 200-3229
Published: August 02, 2011
r 2011 American Chemical Society 10143 dx.doi.org/10.1021/jf202185m | J. Agric. Food Chem. 2011, 59, 10143–10150
ARTICLE
pubs.acs.org/JAFC
Characterization andQuantification of Hydroxycinnamate Derivativesin Stevia rebaudiana Leaves by LC-MSn†
Hande Karak€ose, Rakesh Jaiswal, and Nikolai Kuhnert*
School of Engineering and Science, Chemistry, Jacobs University Bremen, 28759 Bremen, Germany
bS Supporting Information
ABSTRACT: Stevia rebaudiana leaves are used as a zero-calorie natural sweetener in a variety of food products in Asian countries,especially in Japan. In this study, the hydroxycinnamate derivatives of S. rebaudiana have been investigated qualitatively andquantitatively by LC-MSn. Twenty-four hydroxycinnamic acid derivatives of quinic and shikimic acid were detected, and 19 of themwere successfully characterized to regioisomeric levels; 23 are reported for the first time from this source. These comprise threemonocaffeoylquinic acids (Mr 354), seven dicaffeoylquinic acids (Mr 516), one p-coumaroylquinic acid (Mr 338), one feruloylquinicacid (Mr 368), two caffeoyl-feruloylquinic acids (Mr 530), three caffeoylshikimic acids (Mr 336), and two tricaffeoylquinic acids(Mr 678). Cis isomers of di- and tricaffeoylquinic acids were observed as well. Three tricaffeoylquinic acids identified in stevia leavesare reported for the first time in nature. These phenolic compounds identified in stevia might affect the organoleptic properties andadd additional beneficial health effects to stevia-based products.
KEYWORDS: Stevia rebaudiana, chlorogenic acids, hydroxycinnamic acids, caffeoylquinic acids, caffeoylshikimic acids, tandemmass spectrometry
’ INTRODUCTION
Stevia rebaudiana is a plant belonging to the Asteraceae familyof plants, which is native to Brazil and Paraguay. Due to thenatural sweetness of its leaves, S. rebaudiana has caught attentionin scientific and industrial fields to act as a natural zero-caloriesweetener in many applications in the food industry. The leavescontain ent-kaurene glycosides, comprising stevioside, rebaudio-sides A, B, C, D, E, and F, and dulcoside A. All of these diterpeneglycosides comprise a steviol backbone structure; they differ onlyin the glucose moiety at positions C13 and C19 (Figure 1).Stevioside is the main sweet-tasting glycoside in stevia and wasreported to be 250�300 times sweeter than sucrose.1 Rebaudio-side A is the second most abundant ent-kaurene and sweetestcompound in stevia; its sweetness is 400 times greater than thatof sucrose, and it has more pleasant taste and is more water-soluble than stevioside.2 The amounts of diterpene glycosidesmay vary depending on the growth conditions of stevia; however,stevioside accounts for 4�13% (w/w) and rebaudioside Aaccounts for 2�4% (w/w),3 the other glycosides being presentin lower concentrations.
The principal advantage of stevia metabolites is that they arenatural, nonsynthetic products. Stevia leaves can be used in theirnatural state (fresh or dried form), due to their high sweeteningintensity. Only small quantities are needed in comparison to whitesugar to achieve comparable sweetness. The primary use of steviais as a commercial sweetener; it is used in a wide range of productssuch as soft drinks, ice cream, chocolate, yogurt, and baked andcooked foods. Stevia products also have beneficial uses in variousconsumer care products such as toothpaste or mouthwashes.4,5
Stevia may also be used for obesity, diabetics, dental caries, andtherapeutic effects such as hypoglycemic activity.6
The majority of the annual stevia production of an estimated4000 t is produced in China and South America. The stevia crop
has been shown to be highly adaptable to cultivation in manyother parts of the world. S. rebaudiana occurs naturally on acidsoils of pH 4�5 but will also grow on soils with pH levels of6.5�7.5, making it an interesting alternative to plants cultivatedon poor soils such as tobacco.7
In addition to diterpene glycosides, a number of secondaryplant metabolites have been identified from S. rebaudina includinglabdane-type diterpenes, triterpenoids and steroids, flavonoids,and oil components. From S. rebaudiana, 10 labdane-type diter-penoids were identified, including austroinulin, isoaustroinulin,6
and sterebins (A�H).8,9 A triterpenoid, lupeol 3-palmitate, wasalso separated from stevia.10 As plant sterols, β-sitosterol, stigmas-terol, and campesterol were identified from S. rebaudiana.11
Plant phenols are a large and diverse group of compoundsincluding hydroxycinnamates, tannins, flavonoids, stilbenes, cou-marins, lignans, and lignins.12 Chlorogenic acids (CGAs) are themost commonhydroxycinnamate derivatives observed in the plantkingdom. By definition, they are a large family of esters formedbetween quinic acid and one to four residues of certain trans-hydroxycinnamic acids, most commonly caffeic, p-coumaric, andferulic; sinapic and dimethoxycinnamic acids also occur, and insome plant species various aliphatic acids may replace one or moreof the trans-cinnamic acid residues.13 CGAs are involved inbiological functions in plants such as defense against pathogensand resistance to diseases. CGAs also participate in enzyme-catalyzed browning reactions that may adversely affect the color,flavor, and nutritional quality of dietary sources.14
Several pharmacological activities of CGAs including antiox-idant activity, the ability to increase hepatic glucose utilization,15,16
Received: June 1, 2011Revised: July 15, 2011Accepted: August 2, 2011
10144 dx.doi.org/10.1021/jf202185m |J. Agric. Food Chem. 2011, 59, 10143–10150
Journal of Agricultural and Food Chemistry ARTICLE
inhibition of theHIV-1 integrase,17,18 antispasmodic activity,19 andinhibition of the mutagenicity of carcinogenic compounds20
have been revealed by in vitro, in vivo, and human intervention
studies so far. CGAs and their metabolites display additionallyhighly favorable pharmacokinetic properties.21�23 Because thepolyphenols in stevia might affect the organoleptic properties of
Figure 1. Structures and numberings of caffeoylquinic acids.
Figure 2. Base peak chromatogram of Stevia rebaudiana extract using ion trap MS in negative ion mode. For numbering, see Table 1.
10145 dx.doi.org/10.1021/jf202185m |J. Agric. Food Chem. 2011, 59, 10143–10150
Journal of Agricultural and Food Chemistry ARTICLE
stevia-based product and could add additional health benefits tothe product, the objective of the present study was to profile thephenolic content of S. rebaudina leaves with a particular emphasison hydroxycinnamate derivatives.
’MATERIALS AND METHODS
The chlorogenic acids, 3-caffeoylquinic acid, 4-caffeoylquinic acid,5-caffeoylquinic acid (chlorogenic acid), 3,4-dicaffeoylquinic acid, 3,5-dicaffeoylquinic acid, and 4,5-dicaffeoylquinic acid, were purchased fromPhytoLab (Vestenbergsgreuth, Germany). All other chemicals werepurchased from Sigma-Aldrich (Bremen, Germany). Stevia leaves werepurchased from a market in Bremen, Germany.Sample Preparation. Two grams of S. rebaudiana leaves was
immersed in liquid nitrogen, ground in a hammermill, and extracted firstwith 150 mL of chloroform in a Soxhlet apparatus (Buchi B-811extraction system) for 2 h and then with 150 mL of methanol foranother 2 h. Solvents were removed from the methanolic extract invacuo, and extracts were stored at �20 �C until required.
UV Irradiation. The prepared sample of stevia leaf extract (1 mL)was placed in a photoreactor (LuzchemLZC -4 V, Ottawa, Canada)under a shortwave UV lamp and irradiated at 245 nm for 40 min.LC-MSn. The LC equipment (Agilent 1100 series, Bremen,
Germany) comprised a binary pump, an autosampler with a 100 μLloop, and a diode array detector with a light-pipe flow cell (recording at320 and 254 nm and scanning from 200 to 600 nm). This was interfacedwith an ion-trap mass spectrometer fitted with an ESI source (BrukerDaltonics HCT Ultra, Bremen, Germany) operating in Auto-MSnmodeto obtain fragment ions m/z. As necessary, MS2, MS3, and MS4
fragment-targeted experiments were performed to focus only on com-pounds producing a parent ion at m/z 335.1, 337.1, 367.1, 529.2, or677.3. Tandem mass spectra were acquired in Auto-MSn mode (smartfragmentation) using a ramping of the collision energy. Maximumfragmentation amplitude was set to 1 V, starting at 30% and ending at200%. MS operating conditions (negative mode) had been optimizedusing 5-caffeoylquinic acid28 with a capillary temperature of 365 �C, adry gas flow rate of 10 L/min, and a nebulizer pressure of 50 psi.
Figure 3. Extracted ion chromatograms (EIC) of m/z 515 in negative ion mode (A) before and (B) after UV irradiation.
Figure 4. Structures and numbering of tricaffeoylquinic acids.
10146 dx.doi.org/10.1021/jf202185m |J. Agric. Food Chem. 2011, 59, 10143–10150
Journal of Agricultural and Food Chemistry ARTICLE
High-resolution LC-MS was carried out using the same HPLCequipped with a MicroTOF Focus mass spectrometer (BrukerDaltonics) fitted with an ESI source, and internal calibration wasachieved with 10 mL of 0.1 mol/L sodium formate solution injectedthrough a six-port valve prior to each chromatographic run. Calibrationwas carried out using the enhanced quadratic calibration mode.HPLC. Separation was achieved on a 150 � 3 mm i.d. column
containing diphenyl 5μmwith a 4� 3mm i.d. guard column of the samematerial (Varian, Darmstadt, Germany). Solvent A was water/formicacid (1000 + 0.05 v/v), and solvent B was methanol. Solvents weredelivered at a total flow rate of 0.5 mL/min. The gradient profile wasfrom 10 to 70% B linearly in 60 min followed by 10 min isocratic and areturn to 10% B at 80 and 10 min isocratic to re-equilibrate.Calibration Curve of Standard Compounds. Stock solutions
of the standard compounds were prepared in methanol. A series ofstandard solutions was injected (5 μL) into the LC-MS system. Theareas of the peaks of each standard from UV chromatograms were usedto make the respective standard curves.Synthesis of the Mixture of Regioisomers of Tricaffeoyl-
quinic Acids. To a solution of quinic acid (96 mg, 0.5 mmol) andDMAP (16 mg, 0.12 mmol) in CH2Cl2 (10 mL) were added triethy-lamine (4 mL) and 3,4-diacetylcaffeic acid chloride (423 mg, 1.5 mmol)
at room temperature. The reaction mixture was stirred for 6 h andacidified with 2 mol/L HCl (pH ≈1) and then stirred for an additional3 h to remove the acetyl protecting groups. The layers were separated,and the aqueous phase was re-extracted with CH2Cl2 (1� 20 mL) andEtOAc (2 � 20 mL). The combined organic layers were dried overNa2SO4 and filtered, and the solvents were removed in vacuo. Theresulting esters were analyzed by HPLC-MS.
’RESULTS AND DISCUSSION
Methanol extracts of stevia dry leaves were directly used forLC-MS analysis. Efficient separation and resolution wereachieved with diphenyl packing and acetonitrile/water as solventin the HPLC method. Negative ion mode was used for all MSmeasurements. The HPLC method used here constitutes avariation of methods employed previously,24 with variationsrequired to achieve sufficient separation of triacyl chlorogenicacids and ent-kaurene glycosides. In comparison to isolation ofCGAs from green coffee beans, no removal of proteins/peptidesby Carrez reagent was necessary.13,24
All data for chlorogenic acids and diterpene glycosides pre-sented in this paper use the IUPAC numbering system,32 and
Figure 5. Tandem mass spectra of 1,3,5-triCQA in negative ion mode.
Figure 6. Tandem mass spectra of 3,4,5-triCQA in negative ion mode.
10147 dx.doi.org/10.1021/jf202185m |J. Agric. Food Chem. 2011, 59, 10143–10150
Journal of Agricultural and Food Chemistry ARTICLE
structures are presented in Figure 1. Peak assignments of CGAshave been made on the basis of structure diagnostic hierarchicalkeys previously developed,24�26 supported by means of theirparent ion, UV spectra, and retention times relative to 5-CQAusing validated methods in our laboratory.24,27 More sensitiveand more selective fragment-targeted MSn experiments wereused for quantitatively minor components. The base peakchromatogram of stevia extract is shown in Figure 2. Abbrevia-tions and numbering are given in Figure 1. Stevia extract wasanalyzed by LC-MSn in the negative ion mode using an ESI ion-trap mass spectrometer, allowing assignments of compounds toregioisomeric level, and also by high-resolution mass spectro-metry using ESI-TOF in negative ion mode connected to LC.With the guidance of previous studies from Clifford,24�29 threeCQAs (1�3), seven di-CQAs (4�10), three FQAs (18�20),one p-CoQA (11), three CFQAs (12�14), three CSAs(15�17), and four tri-CQAs (21�24) were located in thechromatogram. For all compounds the high-resolution mass datawere in good agreement with the theoretical molecular formulas,with a mass error of below 5 ppm, confirming the elementalcompositions of all compounds investigated.
Reliable characterization of diterpene glycosides content instevia is crucial. Because the structures of single glycosides arevery similar, they have very similar retention times in LC andtherefore result in overlapping of peaks in the chromatogram.In this paper, the general profile of diterpene glycosides in
S. rebaudiana is given. Characterization of the compounds wasachieved by ion-trapmass spectrometrywith SIM, and confirmationof elemental composition was provided by ESI-TOF measure-ments (see the Supporting Information).
Table 1. Tandem Mass Spectral Data of Hydroxycinnamates in Stevia rebaudiana Leaf Extract
MS2 MS3 MS4
secondary peaks secondary peaks secondary peaks
no. compd
m/z
(neg)
base
peak m/z int m/z int m/z int
base
peak m/z int m/z int m/z int
base
peak m/z int m/z int m/z int
1 3-CQA 353.0 190.7 178.8 49 134.9 8 126.8 172.8 37 85.2 55 110.8 65 188.8 174.6 64 134.4 84
2 5-CQA 353.0 190.7 126.8 172.7 49 85.1 61 110.8 21 108.8
3 4-CQA 353.0 172.7 178.8 60 190.6 14 134.8 8 93.0 110.8 62 154.7 24
4 3,5-diCQA 515.1 353.0 190.8 8 190.7 178.8 49 134.9 7 126.8 93.0 98 85.2 62 172.6 48
5 3,4-diCQA 515.1 353.0 335.0 12 172.8 17 172.8 178.8 67 190.8 57 134.8 10 93.0 110.8 41 83.2 6
6 4,5-diCQA 515.1 353.0 299.0 3 254.9 7 172.8 17 172.8 178.6 52 190.8 28 135.0 7 93.0 110.8 89 83.0 18
7 a cis-3,5-diCQA 515.1 353.0 190.8 10 190.7 178.8 50 172.8 11 134.9 10 85.0 126.8 85 93.0 53 172.7 36
8 a cis-4,5-diCQA 515.1 353.0 172.8 13 172.8 178.6 66 190.6 35 134.9 11 93.0 110.9 39 83.0 10
9 cis-4,5-diCQA 515.1 353.0 172.7 7 172.7 178.8 66 190.8 59 134.9 12 93.0 110.9 23 83.0 19
10 a cis-4,5-diCQA 515.1 353.0 172.8 12 172.7 178.6 76 190.6 70 134.8 18 93.0 110.8 39 83.0 6
11 5-p-CoQA 337.1 190.7 162.8 6 126.8 172.7 42 108.8 44 92.8 32
12 3F,5CQA 529.1 367.0 353.0 12 192.7 7 178.6 2 192.7 172.6 13 133.7 14 133.7 126.6 16
13 C,FQA 529.1 367.1 349.0 7 178.7 10 178.7 160.8 73 134.8 85 134.7
14 4C,5FQA 529.1 353.0 254.8 5 172.7 17 172.7 178.6 65 190.6 24 134.7 11 93.0 110.8 32 59.4 52
15 5-CSA 335.1 178.7 172.8 11 134.8 20 134.7
16 4-CSA 335.1 178.7 160.6 82 134.8 51 134.7
17 3-CSA 335.1 178.7 160.8 4 134.8 42 134.8
18 5-FQA 367.1 190.7 172.8 3 85.0 126.8 85 172.6 28
19 FQA 367.1 178.7 190.8 33 160.8 12 134.8 72 134.7 106.8 5
20 FQA 367.2 176.8 161.6 130.8 57
21 3,4,5-triCQA 677.1 515.1 353.0 20 353.0 335.0 16 299.0 4 172.8 30 172.7 178.6 58 190.6 42 134.8 16
22 1,3,5-triCQA 677.1 515.1 353.0 16 353.0 335.0 15 254.9 5 172.7 28 190.8 178.6 72 172.6 90 136.7 10
23 triCQA 677.1 515.0 353.0 15 353.0 172.7 38 254.8 4 172.7 178.8 60 190.6 33 134.7 14
24 triCQA 677.1 515.1 353.0 20 353.0 172.7 22 254.8 4 172.8 178.6 87 190.8 90 134.7 15
Figure 7. Extracted ion chromatograms (EIC) of m/z 677 in negativeion mode (A) before and (B) after UV irradiation.
10148 dx.doi.org/10.1021/jf202185m |J. Agric. Food Chem. 2011, 59, 10143–10150
Journal of Agricultural and Food Chemistry ARTICLE
Characterization of Caffeoylquinic Acids (Mr 354) andDicaffeoylquinic acids (Mr 516). Three peaks were detectedat m/z 353.1 and assigned using the hierarchial keys previouslydeveloped24 as well-known 3-CQA, 5-CQA, and 4-CQA. Threedicaffeoylquinic acid isomers were identified by their parent ionm/z 515.2 and were assigned as 3,5-diCQA, 3,4-diCQA, and 4,5-diCQA using the hierarchial keys.24,26 Three further peakspresent as minor components showed fragmentation patternssimilar to that of 4,5-diCQA. We have recently reported on cisisomers of chlorogenic acids present in plant tissue exposed toUV light, which have formed in a photochemical trans�cisisomerization reaction.30 To confirm if the remaining three peakscorrespond to cis isomers, the extract was irradiated with UVlight at 245 nm for 40 min. After irradiation, a significant increasein the intensities of two peaks (9 and 10 in Figure 3) wasobserved, if compared to their corresponding trans isomers fromthe original plant extract. In addition, a significant increase wasobserved in the intensity of cis-3,5-diCQA (7 in Figure 3) peakaccompanied by a decrease of the 3,4-diCQA (5 in Figure 3)peak. This finding suggests that under the chromatographicconditions employed the cis isomer is coeluting with 3,4-diCQA(Figure 3).On the basis of increased intensity after UV irradiation and
fragmentation pattern, three additional cis isomers were ob-served for 4,5-diCQA. One of these isomers was assigned ascis-4,5-diCQA (9), and two of them were assigned as cis�trans(a cis) isomer, but the distinction between 4-cis,5-trans-diCQAand 4-trans-5-cis-diCQA was not possible (8 and 10).Characterization of Feruloylquinic Acid (Mr 368), p-Cou-
maroylquinic Acid (Mr 338), and Caffeoylferuloylquinic Acid(Mr 530). Only one peak was detected at m/z 337.1, which wasidentified as 5-p-CoQA according to its fragmentation pattern.Three peaks were detected at m/z 367, and one of them wasidentified as 5-FQA; the other two peaks could not be assigneddue to their uncommon fragmentation pattern.A targeted MS3 experiment at m/z 529.2 ([M � H+]�)
applied to the extract located three peaks, and two of them wereidentified as 3F,5CQA and 4C,5FQA on the basis of theircharacteristic fragmentations in MS2 and MS3 spectra. Theassigments are achieved using the hierarchial keys previouslydeveloped, and mass spectra published previously are not pre-sented here.24,31
Characterization of Caffeoylshikimic Acids (Mr 336). Caf-feoylshikimic acids (CSA) have been reported in date palms,sweet basil, and carrot,32�35 and they have been characterized toregioisomeric level in yerba mat�e leaves by tandem mass spectrapreviously.36 This class of compounds is reported here for thefirst time from the Asteraceae family of plants. A targeted MS3
experiment at m/z 335.1 ([M � H+]�) applied to the extractlocated three peaks, and they were identified by their fragmenta-tion patterns as 5-CSA, 4-CSA, and 3-CSA (15�17).36 All three
regioisomers show m/z 178 (caffeic acid fragment) in their MS2
spectra. 4-CQA shows an intense characteristic fragment ion atm/z 160, which is absent in theMS2 spectra of 3-CSA and 5-CSA.
Characterization of Tricaffeoylquinic Acid (Mr 678). Fourtriacyl CQA isomers (Figure 4) were detected in the steviaextract at 677 for tricaffeoyls in neg. mode and confirmed astricaffeoyl derivatives by targeted MS4 experiments. Assignmentof regiochemistry was assisted by an independent synthesis of amixture of all four possible regioisomers of tricaffeoylquinicacids. The chromatogram of the mixture of all theoretically pos-sible four regioisomers of tricaffeoylquinic acid obtained throughsynthesis showed two well-resolved peaks with retention timesand MS data identical to those present in the stevia extract alongwith an intense broad peak in a retention time range where thetwo remaining isomers in the stevia extract were observed (seethe Supporting Information). Detailed studies of the tandemmass spectra at various retention times within the broad peaksuggest that this broad peak must correspond to two distinctunresolved regioisomers of tricaffeoylquinic acid. Comparisonof the chromatogram of the synthetic mixture with the extractallowed unambiguous assignment of the two regioisomers in theextract by identity of the fragmentation pattern compared to thesynthetic mixture. Identification of 1,3,5-triCQA in the extractwas followed automatically due to the absence of an MS4 basepeak atm/z 173 corresponding to a dehydratedMS2 base peak ofthe quinic moiety characteristic of 4-acylated isomers. The MS4
base peak atm/z∼191 and a secondary peak atm/z 178 (72% ofbase peak) suggest the 3,5-disubstitution pattern (Figure 5).3,4,5-triCQA was identified by comparison to material describedpreviously.36,37 (Figure 6)The two remaining peaks might be cis isomers of 3,4,5-triCQA
and 1,3,5-triCQA, or they can correspond to either 1,4,5-tri-CQA and 1,3,4-tri-CQA or any of their cis isomers (see Table 1).However, current information does not allow us to discriminateunambiguously between these regioisomers at the moment. Toprobe whether cis isomers were present, the extract was againirradiated with UV light, and after chromatographic analysis, asignificant increase in the intensity of the peaks of 4-acylatedisomers was observed (Figure 7). Otherwise, the experiment wasinconclusive. It is worth noting that in theory for each tricaffeoylderivative eight stereoisomers with various trans�cis stereoche-mistries are possible, thus increasing the total number of isomerictricaffeoylquinic acids to 32. Given the identity of MS data andthe absence of characteristic shoulders in the UV spectracharacteristic for cis-caffeoyl derivatives, we tentatively assignthe two remaining isomers as 1,4,5-tri-CQA and 1,3,4-tri-CQA. Only 3,4,5-tri-CQA has been previously reported innature, whereas the remaining isomers are reported here forthe first time.Quantification of Caffeoylquinic Acids. Following the qua-
litative profiling of chlorogenic acids in S. rebaudiana, we decided
Table 2. Quantities of Mono- and Di-CQAs in S. rebaudiana Leaves
compd concn range calibration curve correl coeff calcd amount (μg/g)
3-CQA 1 μg/mL�1 mg/mL Y = 4.457x � 460.04 0.99 35.5
5-CQA 1 μg/mL�3 mg/mL Y = 17.719x � 2361.90 0.99 44.3
4-CQA 0.07 μg/mL�1 mg/mL Y = 13.288x � 1223.00 0.99 70.3
3,5-diCQA 0.09 μg/mL�1 mg/mL Y = 5.3176x � 529.76 0.99 145.6
3,4-diCQA 0.07 μg/mL�1 mg/mL Y = 14.789x � 1401.40 0.99 28.6
4,5-diCQA 0.03 μg/mL�04 mg/mL Y = 16.251x � 697.77 0.99 37.2
10149 dx.doi.org/10.1021/jf202185m |J. Agric. Food Chem. 2011, 59, 10143–10150
Journal of Agricultural and Food Chemistry ARTICLE
to quantify the levels of selected compounds. Chlorogenic acidstandard solutions were analyzed by LC-MS using the samechromatographic method as used for stevia leaf extracts. For sixselected monoacyl- and diacylquinic acids, calibration curveswere obtained using six-point calibration from the UV chroma-togram recorded at 320 nm. The individual amounts calculatedfor mono- and dicaffeoylquinic acids are listed in Table 2, whichalso lists the correlation coefficient of linear regression for eachstandard sample and the concentration range.Among the monocaffeoylquinic acids, 4-CQA was found to be
the most abundant compound, and among all CQAs 3,5-diCQAwas found to be the most abundant compound. The totalchlorogenic acid amount determined here is around 370 μg/gof dry leaf.In this study we profiled the chlorogenic acids in S. rebaudiana
employing LC-MSn and LC-TOF techniques. A total of 24chlorogenic acids were detected in S. rebaudiana leaves, with23 compounds described for the first time from this source. Tri-CQAs were reported for the first time from S. rebaudiana withthree regioisomers found for the first time in nature. CSAs werecharacterized for the first time from a plant belonging to theAstareceae family by using tandem mass spectrometry. Quanti-fication of selected mono- and di-CQAs was achieved by usingthe UV chromatogram with total chlorogenic acid levels found tobe 370 μg/g of dry leaf.
’ASSOCIATED CONTENT
bS Supporting Information. Additional EIC of triacylCGAs, MS2 + MS3 data of all compounds mentioned in the text,table of high-resolution MS-TOF data for compounds identified,and structures of ent-kaurene terpenes. This material is free ofcharge via the Internet at http://pubs.acs.org
’AUTHOR INFORMATION
Corresponding Author*Phone: 49 421 200 3120. Fax: 49 421 200 3229. E-mail:[email protected].
Funding SourcesFinancial support from the European Union (project DIVAS) isgratefully acknowledged.
’ACKNOWLEDGMENT
We acknowledge the technical assistance of Anja M€uller.
’DEDICATION†This paper is dedicated to Prof.M. N. Clifford on the occasion ofhis 65th birthday.
’REFERENCES
(1) Kolb, N.; Herrera, J. L.; Ferreyra, D. J.; Uliana, R. F. Analysis ofsweet diterpene glycosides from Stevia rebaudiana: improved HPLCmethod. J. Agric. Food Chem. 2001, 49, 4538–4541.(2) Crammer, B.; Ikan, R. Sweet glycosides from the stevia plant.
Chem. Br. 1986, 22, 915.(3) Pol, J.; Hohnova, B.; Hyotylainen, T. Characterisation of Stevia
rebaudiana by comprehensive two-dimensional liquid chromatographytime-of-flight mass spectrometry. J. Chromatogr., A 2007, 1150, 85–92.
(4) Chatsudthipong, V.; Muanprasat, C. Stevioside and relatedcompounds: therapeutic benefits beyond sweetness. Pharmacol. Ther.2009, 121, 41–54.
(5) Brandle, J. E.; Rosa, N. Heritability for yield, leaf-stem ratio andstevioside content estimated from a landrace cultivar of Stevia-rebaudi-ana. Can. J. Plant Sci. 1992, 72, 1263–1266.
(6) Ibrahim, N. A.; El-Gengaihi, S.; Motawe, H.; Riad, S. A. Phyto-chemical and biological investigation of Stevia rebaudiana Bertoni;1-labdane-type diterpene. Eur. Food Res. Technol. 2007, 224, 483–488.
(7) Shock, C. C. Rebaudi’s stevia: natural noncaloric sweeteners.Calif. Agric. 1982, 4–5.
(8) Oshima, Y.; Saito, J.; Hikino, H. Sterebins A, B, C and D,bisnorditerpenoids of Stevia-rebaudiana leaves. Tetrahedron 1986, 42,6443–6446.
(9) Oshima, Y.; Saito, J.-I.; Hikino, H. Sterebins E, F, G and H,diterpenoids of Stevia rebaudiana leaves. Phytochemistry 1988, 27,624–626.
(10) Yasukawa, K.; Yamaguchi, A.; Arita, J.; Sakurai, S.; Ikeda, A.;Takido, M. Inhibitory effect of edible plant extracts on 12-O-tetradeca-noylphorbol-13-acetate-induced ear edema inmice. Phytother. Res. 1993,7, 185–189.
(11) Kinghorn, A. D. Stevia The Genus Stevia; CRC Press: BocaRaton, FL, 2001.
(12) Rice Evans, C. A.; Miller, N. J.; Paganga, G. Structure�antioxidant activity relationships of flavonoids and phenolic acids.Free Radical Biol. Med. 1996, 20, 933–956.
(13) Clifford, M. N.; Marks, S.; Knight, S.; Kuhnert, N. Character-ization by LC-MSn of four new classes of p-coumaric acid-containingdiacyl chlorogenic acids in green coffee beans. J. Agric. Food Chem. 2006,54, 4095–4101.
(14) Im, H. W.; Suh, B. S.; Lee, S. U.; Kozukue, N.; Ohnisi-Kameyama, M.; Levin, C. E.; Friedman, M. Analysis of phenoliccompounds by high-performance liquid chromatography and liquidchromatography/mass spectrometry in potato plant flowers, leaves,stems, and tubers and in home-processed potatoes. J. Agric. Food Chem.2008, 56, 3341–3349.
(15) Kono, Y.; Kobayashi, K.; Tagawa, S.; Adachi, K.; Ueda, A.;Sawa, Y.; Shibata, H. Antioxidant activity of polyphenolics in diets� rateconstants of reactions of chlorogenic acid and caffeic acid with reactivespecies of oxygen and nitrogen. Biochim. Biophys. Acta: Gen. Subj. 1997,1335, 335–342.
(16) Hemmerle, H.; Burger, H. J.; Below, P.; Schubert, G.;Rippel, R.; Schindler, P. W.; Paulus, E.; Herling, A. W. Chlorogenicacid and synthetic chlorogenic acid derivatives: novel inhibitors ofhepatic glucose-6-phosphate translocase. J. Med. Chem. 1997, 40,137–145.
(17) Robinson, W. E.; Cordeiro, M.; AbdelMalek, S.; Jia, Q.; Chow,S. A.; Reinecke, M. G.; Mitchell, W. M. Dicaffeoylquinic acid inhibitorsof human immunodeficiency virus integrase: Inhibition of the corecatalytic domain of human immunodeficiency virus integrase. Mol.Pharmacol. 1996, 50, 846–855.
(18) Robinson, W. E.; Reinecke, M. G.; AbdelMalek, S.; Jia, Q.;Chow, S. A. Inhibitors of HIV-1 replication that inhibit HPV integrase.Proc. Natl. Acad. Sci. U.S.A. 1996, 93, 6326–6331.
(19) Trute, A.; Gross, J.; Mutschler, E.; Nahrstedt, A. In vitroantispasmodic compounds of the dry extract obtained from Hederahelix. Planta Med. 1997, 63, 125–129.
(20) Stich, H. F.; Rosin, M. P.; Bryson, L. Inhibition of mutagenicityof a model nitrosation reaction by naturally-occuring phenolics, coffeeand tea. Mutat. Res. 1982, 95, 119–128.
(21) Hanlon, N.; Coldham, N.; Gielbert, A.; Kuhnert, N.; Sauer,M. J.; Kingi, L. J.; Ioannides, C. Absolute bioavailability and dose-dependent pharmacokinetic behaviour of dietary doses of the chemo-preventive isothiocyanate sulforaphane in rat. Br. J. Nutr. 2008, 99,559–564.
(22) Stalmach, A.; Mullen, W.; Barron, D.; Uchida, K.; Yokota, T.;Cavin, C.; Steiling, H.; Williamson, G.; Crozier, A. Metabolite profilingof hydroxycinnamate derivatives in plasma and urine after the ingestion
10150 dx.doi.org/10.1021/jf202185m |J. Agric. Food Chem. 2011, 59, 10143–10150
Journal of Agricultural and Food Chemistry ARTICLE
of coffee by humans: identification of biomarkers of coffee consumption.Drug Metab. Dispos. 2009, 37, 1749–1758.(23) Farah, A.; Monteiro, M.; Donangelo, C. M.; Lafay, S. Chloro-
genic acids from green coffee extract are highly bioavailable in humans.J. Nutr. 2008, 138, 2309–2315.(24) Clifford, M. N.; Johnston, K. L.; Knight, S.; Kuhnert, N.
Hierarchical scheme for LC-MSn identification of chlorogenic acids.J. Agric. Food Chem. 2003, 51, 2900–2911.(25) IUPAC Commission on the Nomenclature of Organic Chem-
istry (CNOC) and IUPAC-IUB Commission on Biochemical Nomen-clature (CBN). Nomenclature of cyclitols. Recommendations, 1973.Biochem. J. 1976, 153, 23�31.(26) Clifford,M.N.; Knight, S.; Kuhnert, N. Discriminating between
the six isomers of dicaffeoylquinic acid by LC-MSn. J. Agric. Food Chem.2005, 53, 3821–3832.(27) Clifford, M. N. Chlorogenic acids and other cinnamates �
nature, occurrence, dietary burden, absorption and metabolism. J. Sci.Food Agric. 2000, 80, 1033–1043.(28) Clifford, M. N. Chlorogenic acids and other cinnamates �
nature, occurrence and dietary burden. J. Sci. Food Agric. 1999, 79, 362–372.(29) Clifford, M. N. The Analysis and Characterization of Chlorogenic
Acids and Other Cinnamates; Royal Society of Chemistry: Cambridge,U.K., 2003.(30) Clifford, M. N.; Kirkpatrick, J.; Kuhnert, N.; Roozendaal, H.;
Salgado, P. R. LC-MSn analysis of the cis isomers of chlorogenic acids.Food Chem. 2008, 106, 379–385.(31) Clifford, M. N.; Wu, W. G.; Kuhnert, N. The chlorogenic acids
of Hemerocallis. Food Chem. 2006, 95, 574–578.(32) Mansouri, A.; Embarek, G.; Kokkalou, E.; Kefalas, P. Phenolic
profile and antioxidant activity of the Algerian ripe date palm fruit(Phoenix dactylifera). Food Chem. 2005, 89, 411–420.(33) Ziouti, A.; ElModafar, C.; ElMandili, A.; ElBoustani, E.;
Macheix, J. J. Identification of the caffeoylshikimic acids in the roots ofthe date palm, principle fungitoxic compounds vis-a-vis Fusarium-oxysporum f sp albedinis. J. Phytopathol. 1996, 144, 197–202.(34) Gang, D. R.; Beuerle, T.; Ullmann, P.; Werck-Reichhart, D.;
Pichersky, E. Differential production of meta hydroxylated phenylpro-panoids in sweet basil peltate glandular trichomes and leaves is con-trolled by the activities of specific acyltransferases and hydroxylases.Plant Physiol. 2002, 130, 1536–1544.(35) Kuhnl, T.; Koch, U.; Heller,W.;Wellmann, E. Chlorogenic acid
biosynthesis � characterization of a light-induced microsomal 5-O-(4-coumaroyl)-D-quinate shikimate 30-hydroxylase from carrot (Daucus-carota L) cell suspension cultures. Arch. Biochem. Biophys. 1987, 258,226–232.(36) Jaiswal, R.; Sovdat, T.; Vivan, F.; Kuhnert, N. Profiling and
characterization by LC-MSn of the chlorogenic acids and hydroxycinna-moylshikimate esters in mate (Ilex paraguariensis). J. Agric. Food Chem.2010, 58, 5471–5484.(37) Jaiswal, R.; Kuhnert, N. Hierarchical scheme for liquid chro-
matography/multi-stage spectrometric identification of 3,4,5-triacylchlorogenic acids in green Robusta coffee beans. Rapid Commun. MassSpectrom. 2010, 24, 2283–2294.
Supplemantary Information
`
TRICQAs
EIC of targeted MS for m/z 677
3,4,5‐triCQA (21)
1,3,5‐triCQA (22)
STEVIA_CGA_TAR7.D: EIC 677.0 -All MS
0.00
0.25
0.50
0.75
1.00
1.25
7x10Intens.
44 46 48 50 52 54 Time [min]
228.6293.0 338.0 448.8 515.0
579.1
754.9677.1365.1
-MS,
353.0 629.3
515.1-MS2(677.1),
172.7254.9
353.0-MS3(677.3->515.0)
136.7
190.8 -MS4(677.3->515.3->352.8
0 100
Intens. [%]
0 100
[%]
0 100
[%]
0 100
[%]
100 200 300 400 500 600 700 m/z
160.7 338.0 419.9 515.1
677.1
298.9 353.0
515.1-MS2(677.1)
172.7 254.8
353.0 -MS3(677.5->515.1),
134.8
172.7 -MS4(677.5->515.3->353.0)
0
100
Intens. [%]
0
100
[%]
0
100
[%]
0
100
[%]
100 200 300 400 500 600 700 m/z
-MS
21
glycoside
Supplemantary Information
`
4‐acylated‐triCQA (23/24)
4‐acylated‐triCQA (23/24)
112.8 186.6 238.7 390.8 448.9 487.1 593.1791.3
677.1
497.0
353.0
515.1-MS2(677.1)
172.7254.8
353.0 -MS3(677.5->515.1),
134.7
172.8
-MS4(677.5->515.3->353.1)
0
100
Intens. [%]
0
100
[%]
0
100
[%]
0
100
[%]
100 200 300 400 500 600 700 800 m/z
160.7 238.6 284.9 390.8 466.1 515.1561.3
793.4677.1
658.7
-MS
353.0 467.2
515.0-MS2(677.1)
172.7 254.8
353.0 -MS3(677.2->515.1),
134.7 172.7
325.0
-MS4(677.2->515.2->353.0),
0
100 Intens.
[%]
0
100 [%]
0
100 [%]
0
100 [%]
100 200 300 400 500 600 700 800 m/z
Supplemantary Information
`
Synthesis:
EIC of 677
1,3,5‐triCQA (22)
4‐acyl‐triCQA
CQAMIXTURE_1112.D: EIC 677.0 -All MS
0.00
0.25
0.50
0.75
1.00
1.25
6x10Intens.
0 10 20 30 40 50 60 Time [min]
178.7 260.7 341.1
402.8460.7
515.1544.7 602.6 642.8
677.1
-MS
353.1
515.1-MS2(677.1)
172.7
353.0 -MS3(677.4->515.1)
134.8
172.7 -MS4(677.4->515.3->352.9)
0
100
Intens. [%]
0
100
[%]
0
100
[%]
0
100
[%]
100 200 300 400 500 600 m/z
186.7 260.7 341.0 402.7 460.7
503.1
544.7
677.1
-MS
353.0
515.1 -MS2(677.1)
172.7 298.9
353.0 -MS3(677.3->514.9),
134.8
190.7 -MS4(677.3->515.2->353.0)
0
100 Intens.
[%]
0
100 [%]
0
100 [%]
0
100 [%]
100 200 300 400 500 600 m/z
Supplemantary Information
`
4‐acyl‐triCQA
m/z 353
STEVIA_CGA_DP01.D: EIC 353.0 -All MS
0.0
0.5
1.0
1.5
7x10Intens.
-5 0 5 10 15 20 25 30 Time [min]
160.7 260.7 338.1
460.7515.1
582.7
677.1
353.1
515.1-MS2(677.1)
172.8 254.9 299.0
353.0 -MS3(677.4->515.1)
134.8
172.7 -MS4(677.4->515.2->353.0)
0
100
Intens.[%]
0
100
[%]
0
100
[%]
0
100
[%]
100 200 300 400 500 600 m/z
Supplemantary Information
`
3‐CQA (1):
5‐CQA(2):
4‐CQA(3):
438.0 529.4
353.0
375.0
-MS
134.8
172.7 -MS2(353.0)
154.7
93.0 -MS3(353.2->172.7),
0 50
100
Intens. [%]
0 50
100
[%]
0 50
100
[%]
200 400 600 800 1000 m/z
353.0
190.8
-MS
190.7 -MS2(353.0)
85.1 172.7
126.8 -MS3(353.1->190.8)
0 50
100
Intens. [%]
0 50
100
[%]
0 50
100
[%]
200 400 600 800 1000 m/z
529.4
353.0
375.0
-MS
134.9
190.7 -MS2(353.0)
126.8
172.7
-MS3(353.2->190.8),
0
50 100
Intens.[%]
0
50 100 [%]
0
50 100 [%]
200 400 600 800 1000 m/z
Supplemantary Information
`
m/z 515
3,5‐diCQA(4)
3,4‐diCQA (5)
4,5‐diCQA (6)
353.0
515.1
447.0
-MS
172.8 254.8 298.9
353.0-MS2(515.1)
134.8
172.7 -MS3(515.3->353.0)
71.3
93.0
154.7
-MS4(515.3->353.2->172.8),
0
100
Intens. [%]
0
100
[%]
0
100
[%]
0
100
[%]
100 200 300 400 500 600 700 m/z
256.9 635.0
515.1593.1
-MS
172.8
353.0 -MS2(515.1)
134.9
172.8-MS3(515.3->353.0
93.0
154.7
-MS4(515.3->353.1->172.8)
0
100
Intens. [%]
0
100
[%]
0
100
[%]
0
100
[%]
100 200 300 400 500 600 700 m/z
353.0 635.0
515.1
537.1
-MS
190.8
353.0 -MS2(515.1)
134.9
190.7 -MS3(515.3->353.0),
93.0 172.7
-MS4(515.3->353.1->190.8),
0
100 Intens.
[%]
0
100 [%]
0
100 [%]
0
100 [%]
100 200 300 400 500 600 700 m/z
Supplemantary Information
`
A cis‐3,5‐diCQA (7)
cis 4,5‐diCQA (9)
A cis‐4,5‐diCQA (8)
515.1
353.0
-MS
172.8
353.0 -MS2(515.1)
134.9
172.8 -MS3(515.2->353.0)
59.4
93.0
154.7
-MS4(515.2->353.1->172.7),
0
100 Intens.
[%]
0
100 [%]
0
100 [%]
0
100 [%]
100 200 300 400 500 600 700 m/z
260.7 353.0 431.1 635.1 771.2
515.1
537.1
-MS
172.7
353.0 -MS2(515.1)
134.9
172.7 -MS3(515.1->353.0)
71.3
92.9
154.7
-MS4(515.1->353.0->172.8)
0
100
Intens. [%]
0
100
[%]
0
100
[%]
0
100
[%]
100 200 300 400 500 600 700 m/z
353.0 613.0
515.1
537.0
-MS
190.8
353.0 -MS2(515.1)
134.9
190.7 -MS3(515.2->353.0)
85.1 126.8 172.7
-MS4(515.2->353.1->190.7)
0
100
Intens. [%]
0
100
[%]
0
100
[%]
0
100
[%]
100 200 300 400 500 600 700 m/z
Supplemantary Information
`
A cis‐4,5‐diCQA (10)
5‐p‐CoQA (11)
5‐FQA (18)
96.9 186.7 260.7 595.1
367.1
528.6
-MS
296.7
190.7 -MS2(367.1)
85.1
172.7
-MS3(367.2->190.9)
0
50
100
Intens. [%]
0
50
100
[%]
0
50
100
[%]
200 400 600 800 1000 m/z
92.9
260.7 380.8 725.2
337.1
529.5
-MS
190.7 -MS2(337.1)
126.8
172.7
-MS3(337.2->190.8)
0
50 100
Intens. [%]
0
50 100
[%]
0
50 100
[%]
200 400 600 800 1000 m/z
353.0
515.1
549.2
-MS
172.8 353.0
-MS2(515.1)
134.8 172.7
-MS3(515.2->353.0)
71.2
93.0
154.7
-MS4(515.2->353.0->172.8)
0
100
Intens. [%]
0
100
[%]
0
100
[%]
0
100
[%]
100 200 300 400 500 600 700 m/z
Supplemantary Information
`
5‐CSA (15):
4‐CSA (16):
3‐CSA (17):
96.9260.7 380.8
733.1
335.1
528.7
-MS
134.8 260.6
178.7 -MS2(335.1)
134.8 -MS3(335.3->178.7))
0
50 100
Intens. [%]
0
50 100
[%]
0
50
100
[%]
200 400 600 800 1000 m/z
96.9260.7 529.5
353.1 335.1
-MS
290.9
178.7 -MS2(335.1)
134.7 -MS3(335.2->178.8),
0
50
100
Intens.[%]
0
50
100
[%]
0
50
100
[%]
200 400 600 800 1000 m/z
96.9 186.7 260.7 447.1 529.5 625.1 771.2
399.1 335.1
-MS
134.8 290.9
178.7 -MS2(335.1)
134.7 -MS3(335.3->178.8)
0
50 100
Intens. [%]
0
50 100
[%]
0
50 100
[%]
200 400 600 800 1000 m/z
Supplemantary Information
`
m/z 529
3F,5CQA (12)
(13)
4C,5FQA (14):
176.5 238.6 312.7 396.8 447.0705.3
529.1657.3
-MS
172.7 254.8 460.7 657.2
353.0 -MS2(529.1)
134.7
172.7 -MS3(529.2->352.1)
59.4 93.0
-MS4(529.2->353.0->172.8)
0
100 Intens.
[%]
0
100 [%]
0
100 [%]
0
100 [%]
100 200 300 400 500 600 700 m/z
186.5 254.6 322.8 377.0 445.1 657.3
755.2529.1
508.7
-MS
178.7 460.7
367.1 -MS2(529.1)
134.7 178.7 -MS3(529.4->367.7),
134.7 -MS4(529.4->367.3->178.5)
0
100 Intens.
[%]
0
100 [%]
0
100 [%]
0
100 [%]
100 200 300 400 500 600 700 m/z
260.6 316.8 385.1 431.1 577.2
771.1529.1
-MS
192.7 460.7
367.0 -MS2(529.1)
133.7
192.7 -MS3(529.2->367.1)
59.4 133.7 -MS4(529.2->367.1->192.5)
0 100
Intens. [%]
0 100
[%]
0 100
[%]
0 100
[%]
100 200 300 400 500 600 700 m/z
Supplemantary Information
`
TOF data
Compound No
Compound Molecular Formula
Experimentalm/z (M-H+)-
Theoretical m/z (M-H+)-
Relative Error (ppm)
1 3-CQA C16H17O9 353.0865 353.0878 3.6
2 5-CQA C16H17O9 353.0889 353.0878 3.2
3 4-CQA C16H17O9 353.0883 353.0878 1.3
4 3,5-diCQA C25H24O12 515.1199 515.1195 0.8
5 3,4-diCQA C25H24O12 515.1190 515.1195 0.9
6 4,5-diCQA C25H24O12 515.1183 515.1195 2.3
15 5-CSA C16H16O8 335.0777 335.0772 1.4
16 4-CSA C16H16O8 335.0777 335.0772 1.3
17 3-CSA C16H16O8 335.0778 335.0772 1.8
18 5-FQA C17H20O9 367.1051 367.1035 4.6
11 5-pCoQA C16H18O8 337.0940 337.0929 3.2
21 3,4,5-triCQA C34H30O15 677.1498 677.1512 2.1
- triCQA C34H30O15 677.1517 677.1512 0.8
- triCQA C34H30O15 677.1516 677.1512 0.6
22 1,3,5-triCQA C34H30O15 677.1531 677.1512 2.8
Supplemantary Information
`
Compound R R1 Molecular Formula
Experimental m/z (M-H+)-
Theoretical m/z (M-H+)-
Relative Error (ppm)
Steviol H H C20H30O3 317.2093 317.2122 9.4
Steviolbioside H glc2 - 1glc C32H50O13 641.3181 641.3179 0.4
Rubusoside glc glc C32H50O13 641.3166 641.3179 2.0
Stevioside glc glc2 - 1glc C38H60O18 803.3751 803.3707 5.5
Rebaudioside A glc glc32 -1glc
1glc
C44H70O23 965.425 965.4235 1.6
Rebaudioside B H glc32 -1glc
1glc
C38H60O18 803.368 803.3707 2.8
Rebaudioside C (Dulcoside B)
glc glc32 -1rham
1glc
C44H70O22 949.427 949.4286 1.7
Rebaudioside D glc2-1glc glc32 -1rham
1glc
C50H80O28 1127.4726 1127.4763 3.3
Rebaudioside E glc2-1glc glc2-1glc C44H70O23 965.4199 965.4235 3.7
Rebaudioside F glc glc32 -1xyl
1glc
C43H68O22 935.4097 935.4129 3.5
Dulcoside A glc glc2 - 1rham C38H60O17 787.3732 787.3758 3.3
OR120
18 19
31
5
CO2R
109
7
11 13
15
1617