MODERN TOOLS TO STUDY TRADITIONAL MEDICINAL PLANTS:GINGER AND TURMERIC
Item Type text; Electronic Dissertation
Authors Jiang, Hongliang
Publisher The University of Arizona.
Rights Copyright © is held by the author. Digital access to this material is made possible by the University Libraries, University of Arizona. Further transmission, reproduction or presentation (such as public display or performance) of protected items is prohibited except with permission of the author.
Download date 28/09/2021 01:50:16
Link to Item http://hdl.handle.net/10150/193562
MODERN TOOLS TO STUDY TRADITIONAL MEDICINAL PLANTS: GINGER AND TURMERIC
by
Hongliang Jiang
______
A Dissertation Submitted to the Faculty of the
DEPARTMENT OF PHARMACEUTICAL SCIENCES
In Partial Fulfillment of the Requirements For the Degree of
DOCTOR OF PHILOSOPHY
In the Graduate College
THE UNIVERSITY OF ARIZONA
2005
2
THE UNIVERSITY OF ARIZONA GRADUATE COLLEGE
As members of the Dissertation Committee, we certify that we have read the dissertation
prepared by Hongliang Jiang
entitled Modern Tools to Study Traditional Medicinal Plants: Ginger and Turmeric and recommend that it be accepted as fulfilling the dissertation requirement for the
Degree of Doctor of Philosophy
Barbara N. Timmermann, Ph.D. Date: 06/16/05
David R. Gang, Ph.D. Date: 06/16/05
Myron K. Jacobson, Ph.D. Date: 06/16/05
Victor J. Hruby, Ph.D. Date: 06/16/05
Danzhou Yang, Ph.D. Date: 06/16/05
Final approval and acceptance of this dissertation is contingent upon the candidate’s submission of the final copies of the dissertation to the Graduate College.
I hereby certify that I have read this dissertation prepared under my direction and recommend that it be accepted as fulfilling the dissertation requirement.
Dissertation Director: Barbara N. Timmermann Date: 06/16/05 David R. Gang Date: 06/16/05
3
STATEMENT BY AUTHOR
This dissertation has been submitted in partial fulfillment of requirements for an advanced degree at The University of Arizona and is deposited in the University Library to be made available to borrowers under rules of the Library.
Brief quotations from this dissertation are allowable without special permission, provided that accurate acknowledgment of source is made. Requests for permission for extended quotation from or reproduction of this manuscript in whole or in part may be granted by the head of the major department or the Dean of the Graduate College when in his or her judgment the proposed use of the material is in the interests of scholarship. In all other instances, however, permission must be obtained from the author.
SIGNED: Hongliang Jiang
4
ACKNOWLEDGEMENTS
I would like to thank Drs. Barbara N. Timmermann and David R. Gang for their support and guidance for my research and for the enormous help and effect they have had on my life and scientific attitude. In addition, I thank them for their encouragement, patience, and confidence right from the beginning of my research.
I would also like to thank Drs. Myron K. Jacobson, Victor J. Hruby, and Danzhou Yang for their teaching, advice, consultations, corrections, and generous help that have been critical to this work and my career planning.
My thanks go out to Dr. Arpad Somogyi for his extremely helpful consultations and discussions about mass spectrometry and for providing an excellent resource for part of the work in this study. I also want to thank Dr. Neil Jacobsen for his teaching about the theory and application of NMR and assistance in its operation.
My appreciation also goes out to Dr. Aniko Solyom and Veronica Rodriguez for their generous help in my use of their HPLC and LC-MS and discussion about the results obtained. Moreover, I want thank Dr. Guanjie Chen for his help with the anti- inflammatory bioassay, Dr. Yeisoo Yu for his help with DNA sequencing, Dr. Steven McLaughlin for plant material identification, as well as Jeffrey C. Oliver for consultation of phylogenetic analyses included in this study.
I would also like to thank the past and present members of Dr. Gang’s laboratory. My thanks to Zhengzhi Xie and Hyun Jo Koo for their help with the DNA and secondary metabolites extraction, to Brenda L. Jackson for her assistance in instrumental operation, to Eric McDowell for revision of manuscript, and to Drs. Xiaoqiang Ma, Marycarmen Ramirez, Jeremy Kapteyn, Jennifer Wing and fellow students Shirley Chai and Amanda Scholz.
Finally, I want to thank my parents Zonglu Jiang and Xiulian Li for their love, support, and encouragement throughout my growth. Moreover, I appreciate the help I received from my sister Hongmei Jiang in many aspects. Furthermore, I want to thank Yang Zhang for her love and care during this challenging time. Also, I want to thank all my friends for their general care, help, and confidence to me.
This project is supported by the National Institutes of Health NCCAM/ODS, grants #5 P50 AT 000474-05 and 3 P50 AT 000474-03 S1 to B.N.T., and the National Science Foundation Plant Genome Program, grant DBI-0227618 to D.R.G.
5
TABLE OF CONTENTS
LIST OF FIGURES…………………………………………………………………… 8
LIST OF TABLES……………………………………………………………………. 10
LIST OF SCHEMES………………………………………………………………….. 11
ABSTRACT...…………………………………………………………………………. 12
CHAPTER I − INTRODUCTION….………………………………………………… 14
1.1 Challenges and Opportunities for the Botanical Dietary Supplement Field…… 14
1.2 The Arizona Center for Phytomedicine Research (ACPRx)...………………… 16
1.3 A Brief Review of Ginger and Turmeric-Two Major Botanicals under
Investigation at the ACPRx...………………………………………………… 17
1.3.1 Introduction……...... 17
1.3.2 Ginger…...………………………………………………………………… 19
Historical and Current Uses and Clinical Studies of Ginger……...………… 19
Chemistry and Composition of Ginger……………………………………… 21
Analytical Techniques for Analysis of Ginger……………………………… 23
Phylogenetic Analysis of Ginger and Other Species in Genus Zingiber….… 24
Pharmacological Activities of Ginger……..………………………………… 24
Molecular Mechanisms of Anti-inflammatory Effects of Ginger…………… 25
Prospect of ginger as Anti-inflammatory Agents....………………………… 27
1.3.3 Turmeric…………………………………………………………………… 27
Historical and Current Uses and Clinical Studies of Turmeric……………… 27
Chemistry and Composition of Turmeric…………………………………… 29
6
TABLE OF CONTENTS − Continued
Analytical Techniques for Analysis of Turmeric….………………………… 30
Phylogenetic Analysis of Turmeric and Other Species in Genus Curcuma … 31
Pharmacological Activities of Turmeric…..………………………………… 32
Molecular Mechanisms of Anti-inflammatory Effects of Turmeric ………… 33
Prospect of Curcumin as an Anti-inflammatory Agent……………...……… 34
1.4 Explanation of Thesis Format..………………………………………………… 35
CHAPTER II − PRESENT STUDY..………………………………………………… 38
2.1 General Methods Used for These Studies……………………………………… 38
2.1.1 Sample Preparation……...………………………………………………… 38
2.1.2 Standards Preparation…...………………………………………………… 39
2.1.3 GC/MS Analysis...………………………………………………………… 39
2.1.4 Quantitative Analysis of Gingerols by HPLC-DAD……………………… 40
2.1.5 HPLC-DAD-MS/MS Analysis….………………………………………… 41
LC Separation of Diarylheptanoid...………………………………………… 41
Diode Array Detection…….………………………………………………… 41
MS and MS2 Parameters for ThermoFinnigan LCQ Advantage….………… 41
MS and MS2 Parameters for Agilent LC-MSD-Trap-SL…………………… 42
2.1.6 Other Instrumentation…...………………………………………………… 42
2.2 DNA Sequence- and Chemical Character- Based Phylogenetic Analyses…..… 43
2.2.1 Phylogenetic Analysis of Ginger Accessions and Related Species..……… 44
2.2.2 Phylogenetic Analysis of Turmeric Accessions …………………….……. 52
7
TABLE OF CONTENTS − Continued
2.3 Identification of Compounds in Ginger and Turmeric…………………...... … 58
2.3.1 Identification of Phenolics in Ginger..…………..………………………… 59
Identification of Gingerol Related Compounds in Ginger………………….. 60
Identification of Diarylheptanoids in Ginger……………………………….. 64
2.3.2 Identification of Compounds in Turmeric..….……………………………. 65
Identification of Diarylheptanoids from Turmeric…………………………. 66
Identification of Other Compounds from Turmeric………………………... 69
REFERENCES……..…………...……………………………………………………. 71
APPENDIX A……..…………...………………………………………….…………. 82
APPENDIX B……..…………...…………………………………………..………… 123
APPENDIX C……..…………...…………………………..…………………………. 164
APPENDIX D……..…………...…………………..…………………………………. 196
APPENDIX E……..…………...……………………..………………………………. 226
APPENDIX F……..…………...……………………..………………………………. 262
8
LIST OF FIGURES
Figure 1.1. Some anti-inflammatory constituents in ginger rhizome………………… 22
Figure 1.2. Proposed anti-inflammatory mechanism of [6]-gingerol………………… 27
Figure 1.3. Some anti-inflammatory constituents in turmeric rhizome………………. 30
Figure 1.4. Proposed anti-inflammatory mechanism of curcumin……………………. 33
Figure 2.1. The experimental procedures for DNA sequence- and chemical character-
based phylogenetic analysis……………………………………………… 44
Figure 2.2. Some sequence differences of Zingiber species and Alpinia galanga.…… 46
Figure 2.3. GC/MS profiles of different ginger accessions…………………………... 46
Figure 2.4. GC/MS profiles of different Zingiber species……………………………. 47
Figure 2.5. The content of gingerols in different ginger accessions………………….. 49
Figure 2.6. Comparison of phylogenetic trees of Zingiber species generated on the basis
of DNA sequence and chemical characters………………………………. 50
Figure 2.7. Some sequence differences of turmeric accessions and Alpinia galangal... 53
Figure 2.8. GC/MS profiles of different turmeric accessions………………………… 53
Figure 2.9. The content of curcuminoids in different turmeric accessions……..…….. 55
Figure 2.10. Comparison of phylogenetic trees of different turmeric accessions generated
on the basis of DNA sequence and chemical characters………………… 55
Figure 2.11. System biology investigation of ginger and turmeric…………………… 58
Figure 2.12. The experimental procedures for metabolic profiling of ginger and
Turmeric…………………………………………………………………... 59
Figure 2.13. LC-MS analysis of ginger rhizome extracts…………………………….. 60
9
LIST OF FIGURES-Continued
Figure 2.14. Fragmentation of [6]-, [8]-, [10]-gingerols in (−)ESI-MS/MS………….. 61
Figure 2.15. The chemical structures and molecular weights of gingerol-related
compounds identified in extracts from ginger rhizome....………………. 63
Figure 2.16. Chemical structures and molecular weights of diarylheptanoids identified in
ginger rhizome…………………………………………………………... 64
Figure 2.17. LC-MS analysis of turmeric rhizome extracts………………………….. 65
Figure 2.18. Fragmentation of some diarylheptanoids in ESI-MS/M……….………... 66
Figure 2.19. Chemical structures and molecular weights of diarylheptanoids identified in
turmeric rhizome……………………………………………………….. 69
Figure 2.20. Chemical structures and molecular weights of phenolic acids and terpenoids identified in turmeric rhizome……………………………………………………….. 70
10
LIST OF TABLES
Table 2.1. The sources of Zingiber and Alpinia Samples used in this analysis………. 45
Table 2.2. Chemical composition differences of Zingiber species and Alpinia
galanga…………………….…….………………………………………… 48
Table 2.3. Anti-inflammatory activities and cytotoxicity of Zingiber and Alpinia samples
MeOH extracts…………………………………………………………….. 51
Table 2.4. The sources of turmeric and Alpinia Samples used in this analysis………. 52
Table 2.5. Chemical composition differences of turmeric samples and Alpinia
galanga…………….….…………………………………………………... 54
Table 2.6. Anti-inflammatory activities and cytotoxicity of turmeric and Alpinia samples
MeOH extracts…………………………………………………………….. 56
11
LIST OF SCHEMES
Scheme 2.1. Proposed fragmentation pathway of [6]-, [8]-, [10]-gingerols in (−)ESI-
MS/MS…………………………………………………………………... 61
Scheme 2.2. Proposed fragmentation pathway of [6]-, [8]-, [10]-gingerols in (+)ESI-
MS/MS…………………………………………………………………... 62
Scheme 2.3. Proposed fragmentation pathway of three major curcuminoids in ESI-
MS/MS…………………………………………………………………... 67
Scheme 2.4. Proposed fragmentation pathway of three dihydrocurcuminoids in ESI-
MS/MS…………………………………………………………………... 68
12
ABSTRACT
This dissertation describes three related lines of research concerning the anti- inflammatory botanical dietary supplements ginger (Zingiber officinale Rosc.) and turmeric (Curcuma longa L.). The first part of this research sought to verify and distinguish relationships between ginger and other Zingiber species, entailing a phylogenetic analysis of 10 ginger (Zingiber officinale) accessions, four Zingiber species, and one Alpinia species, using a combination of unique DNA sequences and chemical signatures. The comparative DNA sequence/chemical signature-based phylogenetic analyses of the studied Zingiber species and accessions yielded nearly identical phylogenetic relationships. The ginger accessions exhibited both indistinguishable DNA sequence and similar chemical composition, when compared to each other, yet they were quite distinct from the other Zingiber species. This led us to conclude that either DNA sequence or chemical characters can be used effectively to authenticate ginger plant material.
The second part of this research investigated the utility of LC-ESI-MS/MS in the analysis of gingerols from ginger and curcuminoids from turmeric. All the three authentic curcuminoid standards demonstrated consistent ionization and fragmentation mechanisms in four different mass spectrometers. In contrast to the curcuminoids, the three authentic gingerol standards did show consistent ionization and fragmentation mechanisms in the same mass spectrometer, but instrument dependent ionization mechanisms were observed in the four different mass spectrometers used in this study. These observations establish a
13 foundation for the use of LC-ESI-MS/MS as a method for the identification of gingerols, curcuminoids, and related compounds via the comparison of MS/MS spectra with the appropriate authentic standards; or alternatively, against each other when analyzed in the same mass spectrometer.
In the third part of this research, LC-ESI-MS/MS was used to identify compounds related to the gingerols and curcuminoids in methanolic crude extracts of fresh ginger and turmeric rhizomes. From ginger rhizomes, 31 gingerol-related compounds and 26 diarylheptanoids (curcuminoid-related compounds) were identified, 18 of which are new compounds. From turmeric rhizomes, 19 diarylheptanoids were identified and 6 of these are new compounds. This type of investigation may therefore provide chemical evidence for authentication of fresh ginger and turmeric rhizome and for biosynthetic pathway study of diarylheptanoids and gingerol related compounds.
14
CHAPTER I
INTRODUCTION
1.1 Challenges and Opportunities for the Botanical Dietary Supplement Field
Complementary and alternative medicine (CAM) is defined by the NIH National
Center for Complementary and Alternative Medicine, as a group of diverse medical and health care systems, practices, and products, not considered to be an integral part of
conventional western medicine [1]. These CAM therapies sometimes focus on supposed
biologically active substances found in nature; materials such as botanicals, minerals, and
vitamins. In recent years, CAM’s popularity has grown substantially in the United States.
This trend is evident in data released from the National Health Interview Survey (NHIS)
that revealed that 36% of American adults, ages 18 years and older, used some form of
CAM during 2002 alone [2].
Another example of CAM’s growing acceptance and validity involves its
perception by the general population. The majority of individuals who use CAM have
begun to view it less as “alternative” and more as a “conventional” life style choice [3-5].
In fact, an estimated 15 million adults in the United States combined herbal remedies
with prescription medications, with a majority of these adults using botanicals on a
routine basis [4, 6]. The increased attention Americans have paid to botanicals has been
attributed to an increased dissatisfaction with conventional allopathic therapies and also
15 the desire to play a more active role in health care decisions [3, 7]. This attention has revealed significant economic opportunities in the botanical supplement field.
Along with interest, sales of botanical supplements grew substantially during the mid 1990s, however, from early 1999 to the present, this trend has seemingly reversed itself [7-9]. This decline has been largely attributed to issues of botanical product quality and safety [7, 10], indicating that scientific support for the efficacy of botanical supplements has not kept pace with the rapid expansion of the market and product lines.
Thus, the search for scientific evidence on quality, safety, and efficacy of botanical dietary supplements has presented the greatest challenge to the dietary supplements industry. Although research efforts by the botanical industry are indispensable in addressing this challenge, additional support from governmental agencies, academia and other interested nonprofit institutions are also critical.
Within this context, the National Institutes of Health (NIH), particularly the
Office of Dietary Supplements (ODS) and the National Center for Complementary and
Alternative Medicine (NCCAM), has served as the lead federal agency directing botanical dietary supplements research. Between 1999 and 2002, six national botanical centers were established and funded by ODS in collaboration with NCCAM to study different aspects of botanical supplement research. These six centers include the Arizona
Center for Phytomedicine Research, the UCLA Center for Dietary Supplements
Research: Botanicals, the UIC/NIH Center for Botanical Dietary Supplements for
Women's Health, the Purdue University Botanical Center for Age-Related Diseases, the
MU Center for Phytonutrient and Phytochemical Studies, and the Iowa Botanical
16
Supplements Research Center [7]. However, while each center has its own unique focus,
they all share the common objective of advancing the scientific knowledge base for
botanicals: including issues of quality, chemistry, safety, efficacy, biological activity and
mechanisms of action.
1.2 The Arizona Center for Phytomedicine Research (ACPRx)
The Arizona Center for Phytomedicine Research (ACPRx), one of the six national
botanical centers funded by NIH via the NIH Grant Number P50-AT00474 [11], focuses its research on two major botanicals reported to possess anti-inflammatory properties: ginger and turmeric [12, 13].
Headed by Dr. Barbara N. Timmermann, the ACPRx consists of two research cores and three research projects. The two research cores, the Botany, Agronomy and
Molecular Biology Core (Research Core 1) and the Analytical Chemistry Core (Research
Core 2), each conduct research activities supporting the work of the three related research projects. These three projects are each focused on a unique aspect of the anti- inflammatory botanicals from ginger and turmeric: chemistry and mechanism of action, assessment of bioavailability, and pharmacokinetics and pharmacodynamics. The different groups from these closely associated research cores and projects work collectively to address the central issues of safety and efficacy of anti-inflammatory botanicals. These questions include: the identity of living materials, rhizomes, processed products, rhizome bioactive components, in vitro and in vivo anti-inflammatory
17
mechanism, bioavailability, pharmacokinetics, and pharmacodynamics of bioactive
components.
It is within the authentication of the raw material context, that this dissertation research took place. The majority of my research described within this dissertation was conducted in the laboratory of Dr. David R. Gang, under the auspices of Research Core 1, and in Dr. Barbara N. Timmermann’s laboratory as part of Research Core 2. This research can be categorized under two general themes: the authenticatation of ginger materials by metabolic profiling and phylogenetic analysis, and also the detailed study on the chemistry of ginger and turmeric by HPLC-ESI-MS/MS. The results obtained from these two avenues of investigation could potentially be used for the establishment of
criteria for both the identification of plant material and also quality control assessments
of processed products, relating to the safety and efficacy of botanical dietary supplements.
1. 3 A Brief Review of Ginger and Turmeric-Two Major Botanicals under
Investigation at the ACPRx
1.3.1 Introduction
Inflammation related diseases are diverse and wide ranging. While acute
inflammations are more severe and short lived, chronic inflammations tend to be
prolonged and can often result in other medical complications. Large populations,
including those within the United States, have suffered with many different chronic
inflammations. Arthritis, dubbed ‘the nation’s primary crippler,’ affects more than 37
million Americans [13]. More specifically, debilitating rheumatoid arthritis occurs with a
18
prevalence of about 1% in the population [14]. Thus, broad-spectrum anti-inflammatory
drugs are needed. It has been estimated that the sales of anti-inflammatory and analgesic
drugs exceeded 223 million US dollars in 2000 alone. This number is expected to exceed
240 million US dollars by 2005 [15], revealing a huge market for anti-inflammatory
drugs. Chronic inflammations such as arthritis and rheumatism are typically treated with
nonsteroidal anti-inflammatory drugs (NSAIDs), accounting for 3% of the American prescription drug market [16, 17]. Clinically significant gastrointestinal (GI)
complications including ulcers, ulcer hemorrhages, and ulcer perforations, however, have
been associated with the use of NSAIDs. These GI complications result in approximately
17,000 deaths in the United States each year [16, 17]. Very recently, several anti-
inflammatory drugs such as Vioxx® and Bextra® were removed from the market due to
these safety issues [18]. Therefore, alternative therapies or more efficient drugs with
fewer associated side effects are required for the treatment of inflammation related
diseases.
Under these circumstances, it is no surprise that plant and animal derivatives are
being increasingly used as alternative treatments in the self-management of arthritis and
rheumatism. This trend is reflected in the estimated 5 billion US dollars North Americans
spend annually on herbal remedies for particular ‘anti-rheumatic’ treatments [19, 20]. In
order to prevent fraud and ensure public safety, it is critical to obtain scientific evidence
on the quality, safety, and efficacy of these anti-inflammatory botanicals. Thus, two
popular botanicals, ginger and turmeric, known to possess anti-inflammatory activity [13,
15] were chosen by ACPRx as targets for investigation.
19
In the following sections of this chapter, the uses, clinical studies, chemistry,
composition, analytical characterization techniques, phylogenetic analysis,
pharmacological activities, anti-inflammatory mechanisms, and prospects of ginger and
turmeric as anti-inflammatory botanicals will be briefly summarized. The complete
results will be presented in manuscripts that have been submitted for publication. Since this is only a brief summary, information from every available source will not be discussed. The following text will include summaries for ginger and turmeric in two separate parts.
1.3.2 Ginger
Historical and Current Uses and Clinical Studies of Ginger. Ginger, Zingiber
officinale Rosc. (Zingiberaceae), is a tropical and sub-tropical perennial plant, whose
rhizome has been regularly used for both culinary and medicinal purposes all over the
world for thousands of years. The annual world production of ginger is about 100,000
tons, 80% of which is grown in China [21]. Well known in products such as ginger ale,
ginger sticks, ginger confections, and as a spice [21], ginger is also a traditional Chinese
herbal remedy used to treat a number of diseases including arthritis, headache, colds,
motion sickness, vomiting, and nausea [22, 23]. In addition, Indian Ayuvedic medicine
uses ginger as an anti-inflammatory agent for the treatment of arthritis, rheumatic
disorders, and muscular discomfort [13]. Of these traditional applications, the use of
ginger for medical conditions such as osteoarthritis, motion sickness, and nausea and vomiting in pregnancy remain very popular today (ginger is among the top 20 selling
20 botanicals in United States) and therefore subject to numerous recent clinical studies [21,
24, 25].
One such study evaluated the effect of ginger extract on knee pain in two hundred sixty-one patients with Osteoarthritis (OA). This 6 week study was performed using a randomized, double-blind, placebo-controlled, multicenter, parallel-group and concluded that a highly purified and standardized ginger extract had a statistically significant effect on reducing symptoms of OA on knees with few GI-related side effects [26]. Another study tested the effect of ginger extracts on knee-related symptomatic OA on twenty-nine patients [26]. Conducted with a double blind placebo control and having a 6-month duration, the ginger extract group demonstrated a significant effect that was superior to the placebo group [27]. Due to the scientific evidence obtained from such recent clinical studies, official recommendations for the management of knee OA are periodically revised [28, 29].
Nausea and vomiting, affects 50-80% of pregnant women [24] and remains a significant public health problem with physiological, emotional, social, and economic consequences to women, their families, and society [30]. Consequently, a number of recent clinical studies testing the efficacy of ginger extract/ginger on nausea and vomiting have been conducted [23, 31-33]. All of these studies conclude that ginger extract/ginger is in fact an effective treatment for nausea and/or vomiting. In addition, the use of ginger to alleviate symptoms in early pregnancy was as effective as vitamin B6, a commonly acknowledged first line treatment for nausea and vomiting [32]. Severe adverse effects of ginger on pregnancy outcome were not reported for the above clinical studies. For this
21
reason, further systematic research concerning the risks and benefits of ginger during pregnancy were subsequently suggested due to the limited number of subjects studied [23,
32].
Motion sickness (MS) has also been the subject of recent clinical studies with ginger rhizome extract [25, 34]. However, conflicting conclusions were drawn on the efficacy of ginger extract for treatment of motion sickness [25, 34]. Therefore, further clinical studies including a larger population or a systematic study are necessary to
provide firmer scientific evidence on the efficacy of ginger extract/ginger as a treatment
of motion sickness.
Chemistry and Composition of Ginger. The chemistry of ginger has been well
studied due to its wide application in both medicinal and culinary uses. Moreover, several
reviews containing summaries of chemical compositions of ginger have been published
[21, 35-38]. All of these reviews focus on the oleoresin, the fraction containing most of
the essential oils and pungent compounds characteristic of ginger. Usually, the oleoresin
represents the total flavoring component of ginger.
The main constituents of these essential oils are mono- and sesquiterpenes, which
account for ginger’s distinct aroma and flavor. Despite the inconsistent quantitative data
that has been previously reported, zingiberene, farnesene, bisabolene, β-
sesquiphellandrene, camphene, cineole, β-phellandrene, and ar-curcumene appear to be
the most abundant essential oil chemical components [21, 39, 40]. Other pungent, non-
volatile compounds account for ginger’s medicinal properties. Consisting of gingerols
and gingerol-related compounds, these components belong to distinct groups of
22
homologs (Figure 1.1) that include: gingerols, gingerol derivatives, gingerdiols, their
derivatives, shogaols, paradols, gingerdiones, and dehydrogingerdiones [38, 41].
Differentiated by the length of an unbranched alkyl chain within the same group, these
distinct groups of gingerol homologs are biogenetically derived from phenylalkanes [42].
Among these compounds, the gingerols including [6]-, [8]-, and [10]-gingerols are the
major pungent constituents in fresh ginger, with [6]-gingerol [5-hydroxy-1-(4′-hydroxy-
′methoxyphenyl) decan-3-one] being the most abundant. In addition, the shogaols, a dehydrated form of gingerols, are also known to occur naturally and result from the
elimination of the C-5 OH group. This elimination of the C-5 OH leads to the consequent
formation of a double bond between C-4 and C-5, thereby producing some of the
predominant pungent constituents found in dried ginger [43, 44].
O OH O MeO MeO (CH2)4CH3 (CH2)4CH3
HO HO 6-Gingerol 6-Shogaol
O OH O MeO MeO (CH ) CH 2 6 3 (CH2)6CH3 HO 8-Gingerol HO 8-Shogaol
O OH O MeO MeO (CH ) CH 2 8 3 (CH2)8CH3
HO HO 10-Gingerol 10-Shogaol
Figure 1.1. Some anti-inflammatory constituents in ginger rhizome
23
Another group of ginger compounds possessing biological activity, the
diarylheptanoids, have also attracted much attention in recent years [38, 45]. As a distinct
group of homolog series, diarylheptanoids are also predominant in ginger rhizome [46,
47]. These groups contain the same 1,7-diarylheptane skeleton and are differentiated by subtle structural changes on the heptane skeletons, whereas homologs within each group differ by substitution patterns on the aromatic rings.
Analytical Techniques for Analysis of Ginger. Various analytical techniques
have been used to qualitatively and quantitatively analyze essential oils and pungent
compounds in ginger. Gas-chromatography coupled with mass spectrometry (GC/MS)
has often been used to analyze the composition of ginger essential oil [39, 48]. However,
pungent compounds, especially those with longer alkyl chains, are not easily detected by
GC/MS due to their chemically labile properties [40, 41]. Nevertheless, these compounds
can be identified by alternative methods such as derivatization or other chromatography
associated procedures. For instance, modification of ginger extracts and/or partially
purified fractions to their trimethylsilyl (TMS) derivatives has been shown to improve
their volatility, stability, and separation in the GC/MS [49, 50]. Other methods include
thin-layer chromatography (TLC), high performance liquid chromatography (HPLC),
HPLC coupled to mass spectrometry (LC-MS), and cyclodextrin-modified micellar
electrokinetic chromatography (MEKC) [51-54]. Among these methods, LC-MS provides
a fast and accurate on-line analysis of pungent compounds from complex matrices.
However, since only single dimensional MS analysis with only positive ionization has
been reported so far (providing only parent ion molecular weight information), no
24 detailed structural information of the analyzed compounds has been reported. This lack of information may potentially lead to ambiguous peak identifications. This issue is addressed in this dissertation. Regarding another group of bioactive components in ginger, the diarylheptanoids, no on-line analytical method has been previously developed as a means to characterize and measure these compounds. Thus, analytical methods using instruments such as LC-MS/MS would provide dependable means of characterization.
Such analyses would provide enough detailed structural data to permit the examination of difficult to detect bioactive components such as the diarylheptanoids, in either raw or processed ginger products.
Phylogenetic Analysis of Ginger and Other Species in the Genus Zingiber.
DNA sequence and/or chemically based phylogenetic investigation can be used to authenticate the identity of plant material and distinguish the relationships between different species/genera. A phylogenetic study of 104 species in 41 genera, representing all four tribes of the Zingiberaceae has been previously reported [55]. That study, which was based on DNA sequences of the nuclear internal transcribed spacer (ITS) and plastid matK regions, did not include Z. officinale (ginger) and its related species. As of yet, no other DNA sequence and/or chemically based phylogenetic studies on either ginger or its related species have been reported. Therefore, phylogenetic studies on ginger and its related species are necessary to provide reliable evidence for the authentification of plant materials for culinary and medicinal purposes.
Pharmacological Activities of Ginger. Ginger and its constituents demonstrate a variety of pharmacological activities [21, 35]. Two major groups of non-volatile
25
compounds, gingerol related compounds (pungent compounds) and diarylheptanoids, are
believe to be responsible for a range of diverse activities that include: anti-inflammatory,
analgesic, antipyretic, gastroprotective, cardiotonic, antihepatotoxic, antifungal, antioxidant, antiulcer, anti-tumor promoting, anti-platelet aggregation, molluscicidal, antiemetic, antithrombotic, antibacterial, and antischistosomal activities [21, 38, 56-66].
Although shogaols and paradols (gingerol-related compounds) have also been reported to be bioactive components in ginger [67, 68], most of the recent pharmacological studies of pungent compounds in ginger focus on the investigation of one of the most abundant pungent compounds in ginger: [6]-gingerol [69, 70]. Meanwhile, investigation of other pharmacological activities, such as the antihepatotoxic and antifungal activities of diarylheptanoids in ginger, has also attracted attention in recent years [38, 45, 61].
Although the details of these numerous pharmacological activities will not be discussed, other studies regarding the anti-inflammatory mechanisms of ginger and its constituents will be summarized in detail in the next subsection.
Molecular Mechanisms of Anti-inflammatory Effects of Ginger. Although ginger has long and widely been used in the treatment of inflammation related disease [13,
21], the molecular mechanisms of its anti-inflammatory effects are still not fully understood. Therefore, it is no great surprise that studies concerning the anti- inflammatory effects and mechanisms of actions of ginger and its constituents have gained increased attention. Subsequent studies with [6]-gingerol have reported this compound’s ability to suppress 12-O-tetradecanoylphorbol-13-acetate (TPA)-induced inflammation in mice [71], while other studies suggested that [6]-gingerol might possess
26
both analgesic and anti-inflammatory activity [70]. However, despite their efforts, both
groups were unable to elucidate the actual anti-inflammatory mechanisms of [6]-gingerol.
In 2002, Dedov et al. concluded that both [6]- and [8]-gingerols belong to a relatively potent and efficacious novel class of vanilloid receptor (VR1) agonists [72]. In this scenario, activation of the VR1 receptor at the site of pain generation during inflammation may serve to explain the use of ginger for pain relief in rheumatic and inflammatory conditions [13, 72]. [6]-Gingerol may have an effect on other pain related proteins as well. COX-2, a rate-limiting enzyme involved in prostaglandin (PG) biosynthesis and inflammation, has received attention recently as a molecular target of
many antiinflammatory as well as chemopreventive agents, due to its elevated expression
observed in multiple malignancies such as skin and prostate [73, 74]. In addition, the
transcription factor nuclear factor-kappaB (NF-κB) is known to regulate the induction of
inflammation [69]. Experiments studying the effect of [6]-gingerol on NF-κB and COX-2
have been conducted [69, 73]. These studies revealed that [6]-gingerol suppressed TPA-
induced DNA binding, as well as the transcriptional activities of NF-κB in mouse skin
cells [69, 73]. Moreover, [6]-gingerol proved effective in its ability to inhibit the TPA-
induced phosphorylation and consequent catalytic activity of p38 mitogen-activated
protein (MAP) kinase (Figure 1.2), the protein responsible for the regulation of COX-2
expression in mouse skin [69]. Based on the above studies, the anti-inflammatory
mechanism of [6]-gingerol may have been partially elucidated. Studies are ongoing in the
ACPRx to elucidate the mechanisms of action.
27
[6]-gingerol p38 MAP
NFκB IκBα Activated Expression of COX2 activation NFκB
Figure 1.2. Proposed anti-inflammatory mechanism of [6]-gingerol
Prospect of ginger as Anti-inflammatory Agents. Based on the data available
from the variety of sources summarized above, we believe that ginger (or its components)
makes an excellent candidate as an anti-inflammatory agent for the future. Moreover, the
use of ginger for medicinal purposes has yet to reveal any serious side-effects. Despite
this encouraging news, additional studies addressing the clinical efficacy and safety of
ginger should be performed to answer any potential questions future patients might have
concerning its use.
1.3.3 Turmeric
Historical and Current Uses and Clinical Studies of Turmeric. Like ginger,
turmeric (Curcuma longa L.), has been widely used as a food additive (spice), preservative, coloring agent, and a traditional medicine. Also a member of Zingiberaceae,
tumeric is thought to be indigenous to the Indian subcontinent. Due to its potent and
popular uses, turmeric’s cultivation spread from India to Southeast Asia, China, Northern
Australia, the West Indies, and South and central America. India, however, remains the
largest producer of turmeric in the world, producing about 487,000 metric tons annually
28
for domestic use and export [75]. Traditional Indian medicine utilizes tumeric as a
treatment of arthritis, anorexia, coryza, cough, diabetic wounds, rheumatism, sinusitis, as well as muscular, hepatic, and biliary disorders [76]. Alternatively, in traditional Chinese medicine, turmeric is used as a topical analgesic for conditions ranging from flatulence,
colic, arthralgia, psychataxia, dysmenorrhea, ringworm, hepatitis and chest pain [77, 78].
In the United States, turmeric is an approved food additive and is also used as an alternative treatment for disease and disease prevention [77]. Of all its numerous therapeutic applications, the anti-inflammatory and anticancer activities of turmeric have recently been selected and consequently subjected to clinical studies evaluating the efficacy and safety of its extracts [79, 80].
Four studies have been reported on anti-inflammatory effects of tumeric [81-84].
One trial included a double-blind crossover trial with 18 patients afflicted with rheumatoid arthritis [81]. This group was administered curcumin and phenylbutazone each for two weeks [81]. Patients reported some alleviation of morning stiffness and joint swelling, however, no placebo group was used in this study [81]. Therefore, the efficacy of curcumin can not be adequately evaluated in this instance. Two similar studies (also lacking a placebo group) concerning the anti-inflammatory effects of curcumin in a small number of patients have also been reported [82, 83]. Likewise, the results obtained from these two studies were also unconvincing. The fourth study, however, featured a randomized, double-blind, placebo-controlled trial conducted on 45 post-surgical patients and resulted in data that suggested an anti-inflammatory effect of curcumin [84].
29
Obviously, additional randomized, double-blind, placebo-controlled human studies are
required to evaluate the efficacy of curcumin as an anti-inflammatory agent.
Several other clinical studies have also been conducted evaluating the use of
curcumin as a treatment for a variety of cancers [85-87]. Phase I clinical studies of
turmeric extracts on colorectal cancer have been performed on 15 patients for a period of
4 months [86, 87]. The anticancer effect was assessed by measurements of three
biomarkers in patient blood cells. These measurements included glutathione S-transferase
(GST) activity and the levels of an oxidative DNA adduct (M1G) and prostaglandin E2
(PGE2) [86, 87]. This study suggested that turmeric extracts are safe for patients at levels up to 2.2 g/day despite low oral bioavailability. Another phase I study conducted by
Cheng et al. (2001) suggested that curcumin is an effective treatment of urinary bladder cancer and is not toxic to humans when consumed at no more than 8 g/day [85]. Based on the above summarized clinical studies, curcumin (turmeric) appears to be safe for medicinal use, however, its efficacy for different therapeutic applications still needs to be evaluated by clinical trial-based, scientific evidence. To this end, the ACPRx is conducting in vivo animal studies in an arthritic rat model (female Lewis rat).
Chemistry and Composition of Turmeric. Curcuminoids (Figure 1.3) and essential oils are the major active constituents in turmeric, crucial for the quality of both the plant and its processed products. Due to this fact, most chemical studies described below focused on these two groups of compounds.
The three major curcuminoids: curcumin, demethoxycurcumin, and bisdemethoxycurcumin (3-5% of raw plant), have been subjected to numerous qualitative
30
and quantitative studies [88-90]. In additional to the three major curcuminoids, other
diarylheptanoids (curcuminoid-related compounds) have been reported as minor constituents in turmeric [90]. The major essential oil constituents in turmeric are bisabolane sesquiterpenes that include ar-turmerone, curlone, α-turmerone, β-turmerone
as well as some other sesquiterpenes like zingiberene, curcumenone, curcumenol,
procurcumenol, dehydrocurdione, and germacrone-13-al [88, 91].
O O
HO Bisdemethoxycurcumin OH
O O MeO
HO Demethoxycurcumin OH
O O MeO OMe
Curcumin OH HO
Figure 1.3. Some anti-inflammatory constituents in turmeric rhizome
Analytical Techniques for Analysis of Turmeric. As indicated in the preceding
subsections, the three major curcuminoids have been subjected to various studies
regarding their separation and quantification. Numerous analytical techniques such as
31
GC/MS, HPLC, LC-MS, TLC, and CE have been applied in the analysis of the curcuminoids [88, 89, 92-96]. Among these methods, GC/MS and LC-MS are potentially useful as on-line identification tools; providing for both the separation and the required molecular weight for identification. However, the curcuminoids, due to their low volatility and to thermal degradation, are often difficult to detect by GC/MS without chemical derivatization [89, 97]. One dimensional LC-MS has proved to be a convenient tool in the detection of this group of compounds [88]. However, due to the lack of additional structural information, one dimensional LC-MS is unable to provide accurate identification of the known curcuminoids. Therefore, other hyphenated techniques such as LC-MS/MS and LC-NMR-MS should be applied for accurate characterization of curcuminoids in either raw turmeric or its processed products. In contrast, GC/MS has been proven as an ideal technique to analyze the other major group of bioactive components in turmeric, the essential oil constituents [89, 97-99].
Phylogenetic Analysis of Turmeric and Other Species in Genus Curcuma.
The morphology of the rhizome of different Curcuma species is very similar, making identification of correct turmeric samples difficult. Consequently, researchers have turned to phylogenetic analysis as a means to correctly identify members of the Curcuma genus.
One phylogenetic analysis of Chinese and Japanese Curcuma species used in traditional medicine has been conducted, utilizing the 18 rRNA and trnK gene sequences. Six
Curcuma species including C. longa, C. phaeocaulis, C. zedoaria, C. kwangsiensis, C. wenyujin, and C. aromatica were covered in this study [100]. The results obtained, therefore, might be useful in the identification of Curcuma plants. Genetic variation
32
within turmeric, however, has not been previously investigated. In addition, no chemical
based phylogenetic analysis (chemotaxonomy) has been conducted on Curcuma species.
For confident identification of turmeric samples, phylogenetic analysis of turmeric and related Curcuma species is therefore necessary.
Pharmacological Activities of Tumeric. Turmeric has been reported to possess anti-inflammatory, antiarthritic, antioxidant, antimutagenic, anticoagulant, antidiabetic, antifungal, antiviral, antiprotozoan, antifibrotic, antivenom, anticarcinogenic, antiallergic, antibacterial, immunomodulating, antiatherogenic, anticarminative, diuretic, antimicrobial, antigenotoxic, laxative, mosquitocidal, and anthelmintic activities [75, 91,
96, 101-105]. Generally, the curcuminoids, including curcumin, demethoxycurcumin, and bisdemethoxycurcumin, have been recognized as the components responsible for turmeric’s numerous biological activities. Curcumin, the most abundant curcuminoid, has been subjected to the most intensive laboratory and clinical research pertaining to anti- inflammatory action, cancer prevention and treatment, as well as the treatment of human immunodeficiency virus (HIV) infection [79-81, 84, 106-110]. Furthermore, sesquiterpenes such as ar-turmerone in turmeric have also been reported to have hepatoprotective, mosquitocidal, and apoptosis inducing properties [103, 104, 111].
Detailed information about the numerous pharmacological activities, available from the original cited sources and discussed in other reviews [75, 80, 91, 101], will not be discussed here. The anti-inflammatory effect and mechanisms of turmeric or curcumin, however, will be addressed in detail in the next subsection.
33
Molecular Mechanisms of Anti-inflammatory Effects of Turmeric. Numerous studies have been conducted to elucidate the anti-inflammatory mechanisms of curcumin.
Generally, it is believed that curcumin exerts its anti-inflammatory effect mainly via interactions with cytokines, lipid mediators, and eicosanoids [75]. A variety of proinflammatory cytokines such as tumor necrosis factor-alpha (TNFα) and interleukin-1
β (IL-1β) can be affected by curcumin. As a result, downstream events involving these proinflammatory cytokines are affected [112]. Moreover, many genes involved in inflammatory response initiation are regulated at the transcriptional level by the activated form of NF-κB. Expression of these inflammatory response initiation genes can therefore be inhibited by affecting the activation of NF-κB. Curcumin suppresses the activation of
NF-κB by inhibiting the activity of IκB-kinases (IKKs) thereby interfering with both
IKK’s activation, and IκB degradation, as indicated in Figure 1.4. [113-115].
Inactivated IKK
curcumin Activated IKK
NFκB IκBα Activated Expression of COX2 activation NFκB
Figure 1.4. Proposed anti-inflammatory mechanism of curcumin by affecting the IκBα degradation, IKK activation, and NFκB activation
34
Lipid mediators and eicosanoids also play crucial roles in inflammation.
Curcumin can inhibit the formation and utilization of cellular arachidonic acid, which is crucial for the generation and release of proinflammatory eicosanoids such as prostaglandins and leukotrienes [116]. In addition, curcumin also serves as an inhibitor of cyclooxygenases and lipoxygenases, thus inhibiting the production of the prostaglandin
E2 and the leukotrienes: B4 and C4 [117, 118].
Prospect of Curcumin as an Anti-inflammatory Agent. Based on this brief summary concerning the traditional uses, clinical studies, and anti-inflammatory mechanism studies of turmeric and curcumin, curcumin has great potential as a potent anti-inflammatory agent in the near future. Regarding the development of curcumin as a drug, safety appears to not be a problem due to its high dose tolerance in humans and long use in cooking. However, additional clinical studies should be performed to providing firmer evidence for its anti-inflammatory effects in humans.
35
1.4 Explanation of Thesis Format
In this dissertation, six manuscripts submitted for consideration for publication in different peer-reviewed journals of high impact were included as appendices in appendix
A. My contribution in the research and production of these papers are specified individually.
Appendix A represents the first manuscript included in this dissertation. Entitled
“Metabolic profiling and phylogenetic analysis of medicinal Zingiber species: tools for
authentication of ginger (Zingiber officinale Rosc.)”, this manuscript is currently in press
for publication in Phytochemistry. For this paper, I am the first author and the one responsible for writing the manuscript. Regarding the actual research included in this paper, I am, again, responsible for the majority of the research work described, including
all of the metabolic profiling and phylogenetic analysis of medicinal Zingiber species.
Other coauthors contributed to the plant material identification, collection, metabolites
extraction, DNA extraction, and manuscript revision. Moreover, for the anti- inflammatory assays, I am the one who prepared all the samples for these assays. The
assays, however, were performed by our collaborators: Drs. Lantz and Chen in the Dept.
of Pharmacology, U of A.
Appendix B represents the second manuscript included in this dissertation.
Entitled “Instrument dependence of ESI ionization and MS/MS fragmentation of the
gingerols: A cautionary tale for metabolomics investigations”, this manuscript has been
submitted to the Journal of the American Society for Mass Spectrometry. For this
manuscript, I am the first author as well as the writer of the manuscript. I am responsible
36
for the majority of the research work, including the experimental design and data analysis
too. Drs. Gang, Timmermann, and Somogyi also contributed to the experimental design.
In addition, Dr. Somogyi contributed to the interpretation of the data obtained from FT-
ICR-MS and to the revision this manuscript. The other coauthors also contributed to the revision of this manuscript too.
Appendix C represents the third manuscript contained in this dissertation. Entitled
“Analysis of curcuminoids by positive and negative electrospray ionization and tandem
mass spectrometry”, this manuscript has been submitted to Rapid Communication in
Mass Spectrometry. For this manuscript, I am also the first author and the composer of the manuscript. Experiment design and the majority of the research work and data analysis included for this paper are also my doing. Drs. Gang, Timmermann, and
Somogyi contributed to the analysis and explanation on the tautomerism of curcuminoids.
These coauthors revised this manuscript as well.
Appendix D represents the fourth manuscript and is entitled: “Characterization of
gingerol-related compounds in ginger rhizome (Zingiber officinale Rosc.) by high-
performance liquid chromatography/electrospray ionization mass spectrometry”. This
manuscript has been submitted to Rapid Communication in Mass Spectrometry. For this manuscript, I am also the first author and the one responsible for writing the manuscript.
Furthermore, I am also responsible for the majority of the research work and data analysis. Drs. Gang, Timmermann, and Solyom provided some help in experimental design and data analysis. Moreover, all these coauthors contributed to manuscript revision.
37
Appendix E represents the fifth manuscript and is entitled: “Use of LC-ESI-
MS/MS to identify diarylheptanoids in turmeric (Curcuma longa L.) rhizome”. This manuscript has been submitted to the Journal of Chromatography A. For this manuscript,
I am the first author and wrote the manuscript. I am responsible for the research work and data analysis included in this manuscript. The other two coauthors contributed to the revision of this manuscript.
Appendix F represents the sixth manuscript entitled: “Identification of
diarylheptanoids in ginger (Zingiber officinale Rosc.) by LC-ESI-MS/MS” and was
submitted to the Journal of Agricultural and Food Chemistry. For this manuscript, I am
the first author and wrote the manuscript. I am responsible for the research work and data
analysis included for this manuscript. The other two coauthors contributed to the revision
of this manuscript.
38
CHAPTER II
PRESENT STUDY
The methods, results, and conclusions of this study are presented in the manuscripts appended to this dissertation. The following is a summary of the general methods used for all of these studies as well as the most important findings in these manuscripts.
2.1 General Methods Used for These Studies
2.1.1 Sample Preparation
Frozen, fresh ginger rhizome was ground into a fine powder in the presence of liquid nitrogen with a mortar and pestle and subsequently divided into 1 g aliquots for extraction with methyl t-butyl ether (MTBE) or MeOH. Samples were then transferred to a 4 ml glass vial, covered with 2 ml of solvent (MTBE or MeOH) and capped with PTFE lined cap. After extraction overnight at room temperature with shaking at 200 rpm, each sample was centrifuged in the same glass vial in a SORVALL RC-5 Superspeed
Refrigerated Centrifuge, GSA Rotor (Du Pont instruments, Norwalk, CT, USA) at 1500 rpm for 25 min to pellet the ground plant material. Solvated samples were then filtered through an Acrodisc® CR 13 mm syringe filter with 0.20 µm PTFE membrane. The
MTBE filtrate was used for GC/MS analysis and for anti-inflammatory activity assays.
The MeOH filtrate was used for anti-inflammatory activity assays and for LC-MS
39
analysis. A similar extraction procedure produced MeOH extracts for gingerol
quantitation by HPLC, with the only difference being that the samples were extracted by
sonication for 30 min instead of overnight with shaking. After sonication, samples were
centrifuged and filtered as described above. Triplicate extracts were used for quantitation
analysis.
2.1.2 Standards Preparation
For LC-MS analysis, authentic gingerol standards and curcuminoids were
dissolved in MeOH at a concentration of 100 µg/ml. Each HPLC injection utilized 5 µl of these stock solutions.
For quantitative analysis of gingerols using HPLC, the three authentic gingerol standards were weighed and a stock solution with a concentration of 1000 µg/ml for each of the mixed three gingerol standards was prepared. Thereafter, 5 solutions with concentrations at 25 µg/ml, 50 µg/ml, 100 µg/ml, 150 µg/ml, and 200 µg/ml were prepared from the stock solution for generation of calibration curves. Each HPLC injection utilized 20 µl of these diluted solutions.
For 1H-NMR (500 MHz) experiments, all of the three authentic curcuminoid
standards were dissolved in CD3OD at a concentration of 100 µg/ml.
2.1.3 GC/MS Analysis
This research can be found in Appendix A. All GC/MS data were recorded with a
ThermoFinnigan Trace DSQ GC/MS (ThermoElectron, San Jose, CA). The gas
40 chromatograph was fitted with an Alltech ECONO-CAPTM-ECTM-5 (30m × 0.25mm ID ×
0.25µm) capillary column, with a 5 m guard column. Operating conditions: column oven temperature programmed at 40°C for 2 min, then to 100°C at 8°C/min, then to 280°C at
3°C/min, then to 300°C at 10°C/min and held for 3.5 min; injector/transfer line/ion source temperatures are 220/250/200°C, respectively; electron voltage, 70 eV. UHP helium was used as the carrier gas at a flow rate of 1.2 ml/min. Injection volume was 3 µl and the split ratio was 10. Eluted compounds were identified using the NIST Mass
Spectral library Version 2.0 (NIST/EPA/NIH, USA) and by referral to a publication from
Jolad et al. [41].
2.1.4 Quantitative Analysis of Gingerols by HPLC-DAD
This work is included in Appendix A. An Agilent HPLC system was used for gingerol quantitation. Detector: DAD; Column: Luna C18 (2), 5 µm, 25cm×4.6 mm
(Phenomenex); Guard column: Security Guard AJO-4287 (Phenomenex); Mobil phase: nanopure water (A) and HPLC grade acetonitrile (B); The gradient elution had the following profile: 0-8 min, 45-50% B; 8-17 min, 50-65% B; 17-32 min, 65-100% B; 32-
38 min, 100% B; flow rate: 1ml/min; temperature 48°C ; injection volume: 20 µl; detection: 210, 230, and 280 nm. Triplicate injections for each of the three replicate extracts were performed for each accession to ensure accuracy and reproducibility in this analysis.
41
2.1.5 HPLC-DAD-MS/MS Analysis
LC-MS/MS analyses were included in Appendix B, C, D, E, and F. In addition,
DAD detector was also coupled to the LC-MS/MS for the work included in Appendix D
and E.
LC-DAD-MS/MS analyses involved in this dissertation study were performed on
two separate ion trap mass spectrometer systems: i) a ThermoFinnigan Surveyor MS
HPLC coupled to an in-line PDA detector and a ThermoFinnigan LCQ Advantage ion
trap (San Jose, CA, USA), and ii) an Agilent 1100 HPLC system coupled to an in-line
DAD detector and an Agilent LC-MSD-Trap-SL ion trap (Palo Alto, CA, USA).
LC Separation of Diarylheptanoids. For both instruments, the same column and
elution parameters were used. Column: Discovery® HS C18, 3 µm, 15 cm × 2.1 mm
(Supelco, Bellefonte, PA, USA); guard column: Discovery® HS C18, 3 µm, 2 cm × 2.1 mm (Supelco); mobile phase: (A) buffer (5mM ammonium formate, 0.1% formic acid, in ddH2O) and (B) acetonitrile; gradient (in buffer A): 0-2 min, 5% B; 2-57 min, 5-100% B;
57-60 min, 100% B; 60-65 min, 100-5% B; 65-75 min, 5% B. flow rate: 0.25 ml·min-1;
temperature, 40 °C; injection volume, 5 µl.
Diode Array Detection. The DAD was set at 425 nm, 280 nm, and 230 nm, at 4
nm bandwidth individually, with 550 nm reference wavelength, at 50 nm bandwidth. Full
spectral scanning was also performed from 200 to 600 nm, with range step of 2 nm.
MS and MS2 Parameters for ThermoFinnigan LCQ Advantage. The
acquisition parameters for positive and negative mode were as follows: sheath gas flow
29 (positive), 36 (negative) (in arbitrary units); aux/sweep gas flow 6 (positive), 0
42
(negative); source voltage 5 kV (positive), 4.5 kV (negative); source current 80 µA
(positive and negative); capillary voltage 36 V (positive), -32 V (negative); capillary
temperature 270 °C; tube lens offset 45 V (positive), -25 V (negative); collision gas
pressure ca. 10-5 torr; Q-value 25; mass range measured: 100-1000 m/z.
MS and MS2 Parameters for Agilent LC-MSD-Trap-SL. The acquisition
parameters for positive and negative mode were: drying N2 temperature, 350°C, 10 l/min;
nebulizer pressure 60 psi; HV capillary 4500 V; HV end plate offset -500 V; capillary
current 65.9 nA (positive mode), 62.3 nA (negative mode); end plate current 1482.7 nA
(positive mode), 1378.7 nA (negative mode); capillary exit RF amplitude 99.3 V
(positive mode), -99.3 V (negative mode); skimmer 40.0 V (positive mode), -40.0 V
(negative mode); mass range measured: 50-900 m/z.
2.1.6 Other Instrumentation
A ThermoElectron LCQ Classic ESI-ion trap (San Jose, CA, USA) was used for
direct ESI-MS/MS studies of authentic gingerol and curcuminoid standards, both are included in Appendix B and C. Ionization conditions for the Thermoelectron (Finnigan)
LCQ classic instrument were as follows: needle voltage: 4.5 kV (positive), -4.0 V
(negative), sheath gas flow: 60 (positive), 30 (negative) (both in arbitrary units); other
parameters were optimized for the instrument and did not differ significantly from those
described for the LCQ Advantage as described above.
In addition, an IonSpec 4.7 Fourier transform ion cyclotron resonance (FT-ICR)
mass spectrometer (Lake Forest, CA) was used for high resolution and accurate mass MS
43
and MS/MS experiments of authentic gingerol and curcuminoid standards studies. This
information is included in Appendix B and C. Ions were generated using an Analytica
(Branford, CT) second generation electrospray (ESI) source using diluted solutions of
those analyzed in the ion trap instruments. Direct infusion with a flow rate of 2 µl/min
was applied to generate negatively or positively charged ions by ESI. Tandem MS/MS fragmentation was achieved using the sustained off-resonance irradiation (SORI) technique with N2 as the collision gas. The SORI excitation time was 1500 ms with a N2 pressure of ca. 5 × 10-6 torr in the ICR cell.
Moreover, a Bruker AVANCE DRX500 spectrometer was also used in the study
of the tautomerism of curcuminoids dissolved in CD3OD. This information is included in
Appendix A3.
2.2 DNA Sequence- and Chemical Character- Based Phylogenetic Analyses
DNA sequence- and chemical character- based phylogenetic analyses were used in authentication of ginger and turmeric samples. The experimental procedures of these analyses are outlined in Figure 2.1.
For DNA sequence based phylogenetic analysis, leaves were used for DNA extraction. Then, PCR amplification of selected gene fragments including trnL and rps16 was conducted, followed by isolation and sequencing of the resulting gene fragments.
Phylogenetic analysis was completed based on the DNA sequences from different samples.
44
Rhizome tissues LeafLeaf tissues tissues Rhizome tissues
AntiinflammatoryAntiinflammatory DNA extraction Metabolite extraction DNA extraction Metabolite extraction activityactivity (PGE(PGE2 andand TNF TNFαα)) 2 PCR amplification of gene PCR amplification of gene MetaboliteMetabolite analysis analysis fragments (trnL and rps16) using GC-MS fragments (trnL and rps16) using GC-MS
IsolationIsolation and and sequencing sequencing MetaboliteMetabolite profiling profiling ofof above above fragments fragments
PhylogeneticPhylogenetic ChemotaxonomicChemotaxonomic analysisanalysis analysisanalysis
ClassificationClassification of of Zingiberaceae Zingiberaceae
Figure 2.1. The experimental procedures for DNA sequence- and chemical character- based phylogenetic analysis
For chemical character based phylogenetic analysis, the medicinal part of the plant, the rhizome, was used for metabolite extraction. The metabolite analysis was carried out initially by GC/MS. Based on the metabolic profiles obtained from different samples, chemical character based phylogenetic analysis was performed. In addition, the extracts prepared for metabolite analysis were also used for in vitro anti-inflammatory assays.
2.2.1 Phylogenetic Analysis of Ginger Accessions and Related Species
Many ginger accessions and related Zingiber species and Alpinia galanga from different original sources (Table 2.1) were used in DNA sequence- and chemical character- based phylogenetic analyses, which were conducted on the basis of the experimental design depicted in Figure 2.1.
45
Table 2.1. The original sources of Zingiber and Alpinia Samples used in this analysis
Accession Species Original Source L1 Zingiber officinale Rosc. Alden Botanica, Moreno Valley, CA L2 Zingiber officinale Rosc. ABCO, Tucson, AZ L5 Zingiber officinale Rosc. 17th Street Farmers Market, Tucson, AZ L6 Zingiber officinale Rosc. Tucson Cooperative Warehouse, Tucson, AZ L18 Zingiber officinale Rosc. Stokes Tropicals, New Iberia, LA L32 Zingiber officinale Rosc. Pacific Botanicals, Grants Pass, OR L45 Zingiber officinale Rosc. Plantation Gardens, Clermont, FL L54 Zingiber officinale Rosc. Super K-Mart, Tucson, AZ L55 Zingiber officinale Rosc. Trader Joe's, Tucson, AZ L56 Zingiber officinale Rosc. Trader Joe's, Tucson, AZ L15 Zingiber zerumbet Smith Fairchild Tropical Garden, Miami, FL L37 Zingiber montanum (Koenig) Theilade Gingerwood Nursery, St. Gabriel, LA L46 Zingiber mioga (Thunberg) Roscoe Plantation Gardens, Clermont, FL L31 Zingiber spectabile Griff. Stokes Tropicals, New Iberia, LA L8 Alpinia galanga (L.) Sw. Fairchild Tropical Garden, Miami, FL
All ginger accessions demonstrated identical DNA sequences, which were different from the sequences from the other analyzed species (see Figure 2.2). Therefore, these unique DNA sequences in ginger (as indicated in the dashed box in Figure 2.2) can be used as a DNA fingerprints to distinguish ginger from its related Zingiber species.
Although slightly differences were observed, GC/MS profiles of different ginger accessions were very similar (Figure 2.3). As can readily be seen, terpenoids with relatively low polarity were detected by GC/MS as the predominant compounds for all ginger samples. Gingerol-related compounds with relatively high polarity were also detected by GC/MS. However, some gingerol-related compounds with longer side chains can not be detected by GC/MS without chemical derivatization due to their low volatility.
46
L* ------T T T T C G C C A- - Z. officinale L15 ------C C C C C A C T AA Z. zerumbet L37 GATAAACCTTAGT C C C C C G C C A- - Z. montanum L4 6 G ATA AA ACTA AG G C C C C C G C C AA Z. mioga L31 GATAAACCTTAGT C C C C C A C C A- - Z. spectabile L8 GCTAAACCTTAGT C C C C T G A C ---- A.galanga 64-76 432 497 683 704 74 111 196 336 419-420
The trnL intron and trnL-trnF The intron of rps16 (742 bp) intergeni c spacer (891 bp)
Figure 2.2. Some sequence differences in Zingiber species and Alpinia galanga. L* represents different ginger accessions including L1, L2, L5, L6, L18, L32, L45, L54, L55, and L56. DNA sequence in the dashed box is unique to ginger.
100 7.51 80 9.14 Z. officinale L2 22.82 60 15.21 23.83 Gingerol-related compounds 40 22.32 Terpenoids
14.26 51.12 20 10.58 26.85 60.02 18.5729.72 34.1135.12 39.25 46.68 48.37 59.07 0 100 9.18 7.52 23.05 80 23.40 Z. officinale L32 23.96 60 Terpenoids Gingerol-related compounds 40 15.24 22.26 51.25 14.25 54.83 60.10
Relative Abundance 20 18.57 26.88 50.04 55.17 27.58 35.79 39.25 44.30 0 100 80 7.52 9.16 Z. officinale L56 22.95 60 23.31 Gingerol-related compounds 40 Terpenoids 22.33 54.89 51.20 20 12.21 50.06 60.05 18.56 26.86 27.38 35.80 39.26 44.31 0 5 10 15 20 25 30 35 40 45 50 55 60 Time (min)
Figure 2.3. GC/MS profiles of different ginger accessions. Compounds in the blue box are terpenoids; compounds in the red box are gingerol-related compounds.
47
Although the metabolic profiles of different ginger accessions were similar,
significant differences were observed in the GC/MS profiles of different Zingiber species
(Figure 2.4). Gingerol-related compounds (in the red box in Figure 2.4) were only
detected from ginger, not from the other Zingiber species analyzed.
100 8.16
80 Zingiber montanum L37 9.75 12.67 60 27.93 7.20 40 34.18 39.88 25.42 20 14.20 23.75 31.32 17.64 40.71 45.05 47.97 55.40 63.46 0 100
80 9.18 Zingiber officinale L32 23.05 60 Gingerol-related compounds 23.96 40 15.24 51.25 22.26 54.83 60.10 Relative Abundance Relative 20 14.25 18.57 26.88 27.58 35.79 39.25 44.30 50.04 55.17 0 100 80 Zingiber zerumbet L15
60 7.54 21.52 40 31.50 8.41 11.66 20.10 20 26.41 15.63 19.72 22.12 27.93 33.17 35.68 40.00 46.49 50.08 55.82 63.40 0 5 10 15 20 25 30 35 40 45 50 55 60 Time (min)
Figure 2.4. GC/MS profiles of different Zingiber species. Compounds in the red box are gingerol-related compounds.
Qualitatively, different ginger accessions contained the same chemical set. When evaluated quantitatively, however, differences in the chemical composition of different ginger accessions were observable (Table 2.2). Moreover, the chemical compositions
48 between different species were very different (see Table 2.2). For example, the compound in the blue box was only detected in Zingiber zerumbet. On the other hand, the compounds in the red box were only detected in ginger. These unique compounds therefore could be used as chemical markers to distinguish ginger from related Zingiber species.
Table 2.2. Chemical composition differences of Zingiber species and Alpinia galanga
Z. montanum Z. spectabile Z. mioga A. galanga Z. officinale Z. zerumbet
Compound RT M.W. L1 L2 L5 L6 L18 L32 L45 L54 L55 L56 L15 L37 L46 L31 L8 Name
7.06 3-Thujene 136 ○ 0 0 0 0 0 0 0 0 0 0 0 2 0 1 0
7.20 1R-α-Pinene 136 2 2 2 2 2 2 2 2 2 2 2 2 3 2 2
7.51 Camphene 136 3 3 3 2 2 2 3 3 3 3 3 1 1 0 0
8.35 α-Pinene 136 2 2 2 2 2 2 2 2 2 2 2 2 1 1 0
8.74 3-Carene 136 ○ 0 0 0 0 0 0 0 0 0 0 2 0 0 0 0
9.03 Cymene 134 1 1 1 1 1 1 1 1 1 1 1 1 1 0 0
9.11 Limonene 136 ○ 0 0 0 0 0 0 0 0 0 0 2 0 0 0 0
46.69 [6]-Paradol 278 ● 1 1 1 1 1 1 1 1 1 1 0 0 0 1 0
48.37 [6]-Shogaol 276 ● 2 2 2 2 2 2 2 2 2 2 0 2 0 0 0
49.37 [7]-Paradol 292 ● 2 2 2 1 2 2 2 2 2 2 0 0 3 0 0
51.11 [6]-Gingerol 294 ● 3 3 2 2 2 2 2 2 3 3 0 1 0 0 3
53.95 [8]-Shogaol 304 ● 1 1 1 1 1 1 1 1 1 1 0 2 0 0 0
54.84 [9]-Paradol 320 ● 1 1 1 1 1 2 1 1 1 1 0 1 0 0 0
56.56 [8]-Gingerol 322 ● 2 2 2 1 1 2 1 1 2 2 0 1 0 0 0
59.08 [10]-Shogaol 332 ● 2 2 2 2 1 2 2 2 2 2 0 1 0 0 0
60.03 [11]-Paradol 348 ● 2 2 2 2 2 2 2 2 2 2 0 0 0 0 0
Note: 0 indicates absence, 1 indicates < 0.5%, 2 indicates 0.5% – 5%, and 3 indicates > 5% of total integrated peak area of TIC of a particular sample. Compound in the blue box was only detected in Z. zerumbet; compounds in the red box were only detected in ginger.
49
In addition, the content of [6]-, [8]-, [10]-gingerols, of which slight variation was observed between different ginger accessions (Figure 2.5), were also determined by
HPLC.
2.5
2
1.5 6-gingerol 8-gingerol
mg/g 1 10-gingerol 0.5
0 L55-1 L56-2 L54-2 L45-1 L32-5 L18-1 L6-1 L5-4 L2-2 L1-1 Figure 2.5. The content of gingerols in different ginger accessions
Comparisons of DNA sequence and chemotaxonomic derived phylogenetic trees showed that the chemical characters of the investigated Zingiber species and accessions were able to generate essentially the same phylogenetic relationships as the DNA sequences (Figure 2.6). This information revealed that either DNA sequence or chemical
characters can be used effectively to identify relationships between plant species, a
crucial requirement permitting the authentication of ginger plant material for diverse
culinary and medicinal uses.
50
Figure 2.6. Comparison of phylogenetic trees of Zingiber species generated on the basis of DNA sequence and chemical characters. A: an unrooted phylogenetic tree generated on the basis of trnL and rps16 sequences; B: an unrooted phylogenetic tree generated on the basis of chemical characters identified using GC/MS.
The anti-inflammatory activity and cytotoxicity (Table 2.3) of Zingiber and
Alpinia samples were also determined. These anti-inflammatory assays were conducted by measuring the IC50 of crude MeOH extracts on PGE2 production. The IC50 of ginger samples ranged from 0.058 to 0.629 mg/ml (Table 2.3) and the IC50 of most of these ginger samples were roughly comparable to that of the anti-inflammatory drug indomethacin in our assay system. However, the cytotoxic doses of ginger samples were
10-fold higher that that of indomethacin. These in vitro assays suggested an advantage of using botanical products. In addition, the anti-inflammatory activity of these Zingiber
51 species did not follow the DNA sequence- and chemical character- based phylogeney.
Moreover, this biological activity did not correlate with the content of gingerols in different ginger samples. These results revealed that gingerols are not the only anti- inflammatory compounds in ginger. Therefore, some minor compounds or the synergy of a set of compounds in ginger should also contribute to the anti-inflammatory activity.
Table 2.3. Anti-inflammatory activities and cytotoxicity of MeOH extracts from Zingiber and Alpinia samples -1 -1 Accession Species IC50(PGE2) (µg·ml ) Cytotoxic Dose (µg·ml ) L1 Z. officinale 0.058 >50 L2 Z. officinale 0.059 >50 L5 Z. officinale 0.167 >50 L6 Z. officinale 0.629 50 L18 Z. officinale 0.074 >50 L32 Z. officinale 0.146 >50 L45 Z. officinale 0.069 50 L54 Z. officinale 0.073 50 L55 Z. officinale 0.058 10 L56 Z. officinale 0.065 1-5 L15 Z. zerumbet 0.079 5-10 L37 Z. montanum 7.678 >50 L46 Z. mioga * >50 L31 Z. spectabile 1.171 1 L8 A. galanga 0.055 1-5 Ref. Compd. Indomethacin 0.055 >5 Note: * indicates no inhibitory activity.
In conclusion, ginger samples are relatively uniform at both the DNA sequence and chemical composition levels. Therefore, the identity of plant material could be authenticated by either selected DNA fingerprints or chemical markers. However, the anti-inflammatory activity of ginger may also depend on other compounds besides gingerols.
52
2.2.2 Phylogenetic Analysis of Turmeric Accessions
As was done for the ginger accessions and related species, turmeric accessions from different original sources (Table 2.4) were used in this study.
Table 2.4. The original sources of turmeric and Alpinia Samples used in this analysis
Accession Species Original Source L4 C. longa L. Aloha Tropical, Oceanside, CA L22 C. longa L. Stokes Tropicals, New Iberia, LA L27 C. longa L. Andy’s Organics, Pahoa, HI L33 C. longa L. Pacific Botanicals, Grants Pass, OR L9 C. longa L. Fairchild Tropical Garden, Miami, FL L50 C. longa L. Gingers-R-Us, Tallahassee, FL L52 C. longa L. Gingers-R-Us, Tallahassee, FL L43 C. longa L. Plantation Gardens, Clermont, FL L51 C. longa L. Gingers-R-Us, Tallahassee, FL L8 Alpinia galanga (L.) Sw. Fairchild Tropical Garden, Miami, FL
In contrast to ginger accessions, which possessed identical DNA sequences, turmeric accessions showed variable DNA sequences (Figure 2.7). Turmeric accessions in the red box of Figure 2.7 exhibited identical DNA sequences, which were different from other turmeric accessions (in the blue box).
Similarly, turmeric samples from different accessions demonstrated very different
GC/MS profiles (Figure 2.8). Turmerone and curlone, indicated in Figure 2.8, were only detected from accession L33 but not from accession L4. In addition, curcuminoids, the predominant components in turmeric, were not detected by GC/MS under the conditions evaulated due to their low volatility.
53
L4 AAAGT T A T (1591 bp) L22 AA--GT C A -- (1589 bp) L27 AA----G C -- T (1588 bp) L33 AA----G C -- T (1588 bp) Turmeric L9 A------G C -- T (1587 bp) L50 ------G C -- -- (1585 bp) L52 ------G C -- -- (1585 bp) L43 ------G C -- -- (1585 bp) L51 ------G C -- -- (1585 bp) L8 ------T C A -- (1582 bp) A. galanga
273-7 637 100 592
The intron of rps16 (717 bp) the trnL intron and trnL-trnF intergenic spacer (874 bp) Figure 2.7. Some sequence differences in the trnL intron, the trnL-trnF intergenic spacer and the intron of rps16 from turmeric accessions and Alpinia galanga. Red and blue boxes represent different subgroups with identical DNA sequence.
100 9.2 Turmeric Accession L4 80 Aloha Tropical 60 26.8 40 11.7 33.1 e 33.3 c 7.5 19.2 n 20 22.3 39.5 a 12.017.3 23.1 29.8 35.9 50.1 54.7 d n 0 u
b
100 turmerones 29.7 lative A 80 Turmeric Accession L33 30.6 curlone Re 60 23.0 Pacific Botanicals 24.0 40 8.610.4 20.1 22.4 20 7.7 39.7 26.9 33.9 66.169.5 0 5 10 15 20 25 30 35 40 45 50 55 60 65 70
Time (min) Figure 2.8. GC/MS profiles of different turmeric accessions. Compounds in the red box including turmerones and curlone were only detected from accession 3.
54
The chemical compositions determined by GC/MS of different turmeric
accessions were also compared. Some of the compounds identified are presented in Table
2.5. Significant differences were observed between these turmeric samples. For example,
the compound in the blue box in Table 2.5 was only detected from accession L22. On the
other hand, compounds in the green box were not detected from accession L4 (Table 2.5).
Table 2.5. Chemical composition differences of turmeric samples and Alpinia galangal
A. galanga Turmeric
RT Name of compounds L4 L22 L27 L33 L9 L50 L52 L43 L51 L8 12.85 r-Menth-1-en-8-ol, 1 2 0 0 0 1 1 1 0 1 15.86 2-Undecanone 0 2 1 1 0 0 0 0 0 0 17.28 δ-Elemene 2 0 0 0 0 0 0 1 1 0 17.53 4-Allylphenyl acetate 0 0 0 0 0 0 0 0 0 1 21.25 a-Caryophyllene 0 0 0 0 1 0 0 0 0 1 21.37 α-Farnesene 2 0 0 0 2 2 2 2 2 2 22.26 Germacrene D 2 2 0 0 0 0 0 0 0 0 22.41 a-Selinene 1 2 0 0 0 0 0 0 0 0 23.32 β-Bisabolene 0 0 0 0 2 2 2 2 2 0 23.69 Cadina-1(10), 4-diene 0 1 0 0 0 0 0 0 0 0 24 β-Sesquiphellandrene 0 0 0 0 3 3 3 3 3 2 24.89 β-Elemene 2 1 0 0 0 0 0 0 0 0 28.99 Ar-turmerone 0 0 3 3 3 3 2 3 3 0 29.19 turmerone 0 0 2 3 3 2 3 3 3 0 29.8 (E,E)-Germacrone 2 3 0 0 0 2 2 2 2 0 30.25 Curlone 0 0 3 3 3 3 2 3 3 0
Note: 0 indicates absence, 1 indicates < 0.5%, 2 indicates 0.5% – 5%, and 3 indicates > 5% of total integrated peak area of TIC of a particular sample. Compound in the blue box was only detected in turmeric accession L22; compounds in the green box were not detected in turmeric accession L22
In contrast to what was observed from ginger accessions, which were relatively
uniform in the content of gingerols, turmeric accessions demonstrated a highly variable
content of curcuminoids (Figure 2.9), as determined by HPLC.
55
7 Turmeric 6 5
4 Bisdemethoxycurcumin mg/g 3 Demethoxycurcumin 2 Curcumin 1 0
L4-1 L22-1 L27-4 L33-2 L9-2 L43-5 L50-2 L52-2 L51-2 Figure 2.9. The content of curcuminoids in different turmeric accessions
Figure 2.10. Comparison of phylogenetic trees generated on the basis of DNA sequence and chemical characters of different turmeric accessions. A: an unrooted phylogenetic tree generated on the basis of trnL and rps16 sequences; B: an unrooted phylogenetic tree generated on the basis of chemical characters identified using GC/MS. Accessions in the same box possessed very similar or identical DNA sequences and chemical profiles.
56
Phylogenetic trees (Figure 2.10) of different turmeric accessions were generated
on the basis of DNA sequence and chemical characters. Due to the highly variable DNA
sequence and chemical profiles of different turmeric samples, several subgroups indicated with different boxes were generated in the phylogenetic trees (Figure 2.10). For
samples in the same subgroup, they demonstrated similar or identical DNA sequence and
chemical profiles.
Table 2.6. Anti-inflammatory activities and cytotoxicity of turmeric and Alpinia samples MeOH extracts -1 -1 Accession Species IC50(PGE2) (µg·ml ) Cytotoxic Dose (µg·ml ) L4 C. longa L. 0.338 50 L22 C. longa L. 0.686 >50 L27 C. longa L. 0.308 10-50 L33 C. longa L. 0.077 10-50 L9 C. longa L. 0.386 5 L50 C. longa L. 0.063 10-50 L52 C. longa L. 0.074 >50 L43 C. longa L. 0.062 10-50 L51 C. longa L. 0.060 10-50 L8 A. galanga 0.055 1-5 Ref. Compd. Indomethacin 0.055 >5
Anti-inflammatory assays (Table 2.6), which determined the IC50 of turmeric
samples on PGE2 production, were also conducted. Interestingly, the anti-inflammatory
activity of turmeric samples neither followed the DNA sequence- and chemical character- based phylogeny nor correlated the content of curcuminoids. For example, accessions
L27 and L33, with identical DNA sequence and similar chemical profiles, belonged to
the same subgroup in the phylogenetic analysis (Figure 2.10) but exhibited very different
IC50 (Table 2.6). These results suggested that the curcuminoids are not the only anti- inflammatory compounds in turmeric. Therefore, other (minor) compounds or the
57 synergy of a set of compounds may also be responsible for or contribute to the anti- inflammatory activity of turmeric.
In contrast to ginger, turmeric samples are highly variable in chemistry, DNA sequence, and biological activity. This may be the result of multiple origins for turmeric samples. More evidence is required to authenticate turmeric samples.
58
2.3 Identification of Compounds in Ginger and Turmeric
This part of my research has been part of a collaborative project to study the biosynthesis of gingerols, curcuminoids, and other important metabolites in ginger and turmeric using systems biology approach (Figure 2.11). For this study, comparative analyses including microarray, proteomic, and metabolomic data were conducted. My research work focused on metabolite identification and metabolic profiling. Metabolic profiling of ginger and turmeric followed the experimental procedure outlined in Figure
2.12. Identification of compounds detected by GC/MS was conducted by comparing the
EI mass spectra of compounds with data in the mass spectra library that came with the instrument and in the literature. However, for identification of compounds detected by
LC-MS, no good commercial library is available. So the majority of my efforts here have been to focus on identification of compounds in ginger and turmeric using LC-MS/MS.
*
EST database to identify putative Proteomics Metabolic profiling genes for pathways; to identify (metabolomics) to identify DNA microarrays to identify expressed as many metabolites as differentially expressed genes protein posssible
comparative analysis of microarray, proteomic, and metabolomic data.
Elucidate the biosynthetic pathways of gingerols, curcuminoids, and other important metabolites in ginger and turmeric
Figure 2.11. Systems biology investigation of ginger and turmeric
59
Sampling: fresh frozen rhizome and leaves tissues
Metabolite extraction: different solvents; different extraction methods
Metabolic profiling: GC-MS (ThermoFinnigan Trace DSQ); LC-MS (ThermoFinnigan LCQ Advantage and Agilent LC-MSD-Trap-SL ion trap)
Metabolite identification: GC-MS (EI-MS spectra, RT, and RI): library search and literature comparison LC-MS/MS (negative and positive ion ESI-MS/MS spectra and RT): authentic standards’ spectra and literature comparison FT-ICR-MS (positive ion ESI-MS/MS spectra and accurate mass)
Figure 2.12. Experimental approach for metabolic profiling of ginger and
turmeric
2.3.1 Identification of Phenolics (Gingerol Related Compounds and
Diarylheptanoids) in Ginger
Both negative and positive ion mode ESI analyses (Figure 2.13) were performed on ginger rhizome extracts. Diarylheptanoids and gingerol-related compounds, indicated by different boxes in the chromatograms shown in Figure 2.13, were able to be detected by the two modes under optimized conditions.
60
Identification of Gingerol-Related Compounds in Ginger
Ion trap-based ESI-MS/MS proved to be a very sensitive and precise method for identifying specific gingerols with different lengths of side chains. This specificity is due to the consistent and predictable ionization and fragmentation behaviors (Figure 2.14) observed for all three authentic gingerol standards when analyzed in the same instrument under the same optimized condition. Their fragmentation behaviors in (−)ESI-MS/MS were rationalized in the fragmentation pathway proposed in Scheme 2.1. For example, the major product ion with a mass unit of 193 was formed by a McLafferty rearrangement and the loss of a neutral moiety. The other product ions were also rationalized in this proposed pathway. In this manner, the proposed structures of these compounds were confirmed.
Diarylheptanoids Gingerol-related compounds 6 X10 (Curcuminoid-related) 1.5 1.0
0.5
0.0 0 10203040506070 Total Ion Current (TIC) Chromatogram from negative ion (-)ESI-HPLC-MS
8 Diarylheptanoids Gingerol-related compounds X10 1.0 (Curcuminoid-related) Absolute Abundance 0.5
0.0 010203040506070 Total Ion Current (TIC) Chromatogram from positive ion (+)ESI-HPLC-MS
Time
Figure 2.13. LC-MS analysis of ginger rhizome extracts. The top and bottom chromatograms are, respectively, the total ion current (TIC) chromatograms from negative ion (–)ESI-HPLC-MS and positive ion (+)ESI-HPLC-MS.
61
100 (1) Q 193.2
% 6-Gingerol S R P O 99.7 178.7 275.3 293.5 0 50 100 150 200 250 300 350 400
100 (2) Q 193.2 8-Gingerol % S R O 127.7 P 178.6 303.3 321.7 0 50 100 150 200 250 300 350 400
100 (3) Q 193.2 10-Gingerol % S R 155.7 178.6 P O 331.3 349.7 0 50 100 150 200 250 300 350 400 Figure 2.14. Fragmentation of [6]-, [8]-, [10]-gingerols in (−)ESI-MS/MS.
O O OH O O 3' H MeO - CH (CH )n (CH )nCH 3 2 OMe + 1 3 5 2 3 OCH (CH )nCH 3 H 2 3 O O HO OH 4' 294 (1, n=4) H 293 (1, n=4) 193 (1) 100 (1, n=4) 322 (2, n=6) 321 (2, n=6) O 193 (2) Q 128 (2, n=6) Mr 350 (3, n=8) 349 (3, n=8) 193 (3) 156 (3, n=8)
H O - 2 CH - 3 - H+ CH C=O O O - 3 O CH (CH )n 3 2 CH (CH )n C 3 2 OMe O O O O OH 275 (1, n=4) 135 (1) 178 (1) 99 (1, n=4) 303 (2, n=6) P 135 (2) T 178 (2) R 127 (2, n=6) S 331 (3, n=8) 135 (3) 178 (3) 155 (3, n=8)
Scheme 2.1. Proposed fragmentation pathway of [6]-, [8]-, [10]-gingerols in (−)ESI-MS/MS. Product ion Q in the red box is the major product ion for these three gingerols.
In addition, the gingerols were also detected in the positive ion mode. In this mode, different adducts such as ammonium adducts and sodium adducts could be formed.
These adducts fragmented in different ways. For ammonium adducts, the major product
62
ion B was formed through a rearrangement as indicated in Scheme 2.2. In contrast, the
sodium adducts fragmented through a McLafferty rearrangement, leading to the
formation of their major product ion C (Scheme 2.2). The different fragmentation
behaviors of these adducts might depend on the position of the positive charge on the ion species.
Ammonium adduct (a) CH3(CH2)nCHO + O + O + [M+NH4] MeO MeO MeO (CH )nCH + (CH2)nCH3 + 2 3 [M+H-H2O] HO HO + A HO [M+H] 277 (1; n=4) B 177 (1, 2, 3) A 305 (2; n=6) 333 (3; n=8)
Sodium adduct + OH OH O OH +Na + CH3(CH2)nCHO + (b) 3' Na Na MeO CH (CH )n (CH )nCH 3 2 OMe OMe 1 3 5 2 3 OH HO O O 4' H C 217 (1, 2, 3) 294 (1; n=4) 317 (1; n=4) 322 (2; n=6) 345 (2; n=6) 350 (3; n=8) 373 (3; n=8)
[6]-gingerol (1) [8]-gingerol (2)
[10]-gingerol (3)
Scheme 2.2. Proposed fragmentation pathway of [6]-, [8]-, [10]-gingerols in (+)ESI-MS/MS. Product ions in the boxes are the major product ions for these three gingerols.
In addition, the complementary information obtained from the fragmentation of these various adducts could also be used to confirm the proposed structures of the compounds. For the three major gingerols, their structures were confirmed by comparison with authentic standards. In addition, FTMS was also used to measure the accurate mass
63 of both the precusor and product ions of these compounds. Therefore, the confirmed ionization and fragmentation mechanisms of the three gingerols could be used to identify other gingerols or their homologs with different lengths of the side chain.
O O OH OH O MeO MeO MeO (CH2)nCH3 (CH )nCH (CH2)nCH3 2 3 MeO HO (1) [6]-gingerol, n=4 HO (8) [6]-shogaol, n=4 (2) [8]-gingerol, n=6 (5) Methyl [6]-gingerol, n=4 (9) [8]-shogaol, n=6 (3) [10]-gingerol, n=8 (6) Methyl [8]-gingerol, n=6 (10) [10]-shogaol, n=8 (4) [12]-gingerol, n=10 (7) Methyl [10]-gingerol, n=8 (11) [12]-shogaol, n=10
O O O O OAc MeO MeO MeO (CH2)nCH3 (CH2)nCH3 (CH2)nCH3 HO (12) [6]-paradol, n=4 HO HO (13) 1-Dehydro- [6]-gingerdione, n=4 (17) Acetoxy-[6]-gingerol, n=4 (14) 1-Dehydro-[8]-gingerdione, n=6 (18) Acetoxy-[8]-gingerol, n=6 (15) 1-Dehydro-[10]-gingerdione, n=8 (19) Acetoxy-[10]-gingerol, n=8 (16) 1-Dehydro-[12]-gingerdione, n=10 O OAc OH OAc OH OAc MeO (CH )nCH MeO MeO 2 3 (CH2)nCH3 (CH )nCH 2 3 MeO HO MeO (21) Methyl acetoxy-[6]-gingerol, n=4 (20) 5-Acetoxy-[6]-gingerol, n=4 (30) Methyl 5-acetoxy-[6]-gingerdiol, n=4
OAc OAc OAc OAc MeO MeO (CH )nCH (CH2)nCH3 2 3
HO MeO (22) Diacetoxy-[4]-gingerol, n=2 (26) Methyl diacetoxy-[4]-gingerol, n=2 (23) Diacetoxy-[6]-gingerol, n=4 (27) Methyl diacetoxy-[6]-gingerol, n=4 (24) Diacetoxy-[8]-gingerol, n=6 (28) Methyl diacetoxy-[8]-gingerol, n=6 (25) Diacetoxy-[10]-gingerol, n=8 (29) Methyl diacetoxy-[10]-gingerol, n=8
Figure 15. The chemical structures and molecular weights of gingerol- related compounds identified in extracts from ginger rhizome. Note: compounds in different boxes are from different homologous series; compounds in red color are new compounds.
Some other gingerol-related compounds were also identified from ginger rhizome by LC-MS/MS. They were identified in the same way as was done for the three gingerols.
These compounds are from different homologous series as indicated with different boxes in Figure 2.15. Compounds from the same homologous series are differentiated by the
64 length of the side chain. But for compounds from different homologous series, their structural differences lie on both the aromatic ring and on the side chain. We identified several new compounds.
O OH OH OH
R1 R2 R1 R2 MW=390 (1, R1=Ar4, R2=Ar3) MW=436 (4, R1=R2=Ar5) MW=360 (2, R1=Ar2, R2=Ar3) MW=406 (5, R1=Ar5, R2=Ar3) MW=404 (9, R1=Ar5, R2=Ar3) MW=376 (8, R1= R2=Ar3) MW=374 (10, R1=R2=Ar3) OH OAc
OAc OAc R1 R2 MW=420 (3, R1=Ar4, R2=Ar2; or R1=Ar2, R2=Ar4) R1 R2 MW=450 (5, R1=R2=Ar4) MW=390 (6, R1=R2=Ar2) MW=492 (14, R1=R2=Ar4) MW=404 (11, R1=Ar2, R2=Ar3; or R1=Ar3, R2=Ar2) MW=462 (15, R1=Ar4, R2=Ar2) MW=374 (12, R1=Ar2, R2=Ar1; or R1=Ar1, R2=Ar2) MW=432 (16, R1= R2=Ar2) MW=434 (13, R1=Ar4, R2=Ar3) MW=506 (19, R1=Ar4, R2=Ar5) MW=448 (17, R1=Ar5, R2=Ar3) MW=416 (20, R1=Ar2, R2=Ar1) MW=418 (18, R1=R2=Ar3) MW=476 (21, R1=Ar4, R2=Ar3) O O MW=446 (22, R1=Ar2, R2=Ar3) MW=490 (23, R1=Ar5, R2=Ar3) MW=430 (24, R1=Ar3, R2=Ar1) MW=460 (25, R1=R2=Ar3) R1 R2 MW=370 26( , R1= R2=Ar3) MeO MeO
HO HO HO HO HO
OH OMe OH OMe Ar1 Ar2 Ar3 Ar4 Ar5
Figure 2.16. Chemical structures and molecular weights of diarylheptanoids identified in ginger rhizome. Note: compounds in different boxes are from different homologous series; compounds in red color are new compounds.
Identification of Diarylheptanoids in Ginger
Many diarylheptanoids (Figure 2.16), another major group of compounds in ginger, were also identified. These compounds also belong to different homologous series.
65
For compounds from the same homologous series, their structural differences lie on the aromatic ring. But for compounds from different homologous series, their structural differences lie on the heptane chain. A total of 15 new diarylheptanoids were identified from ginger rhizome by LC-MS/MS (see Figure 2.16).
2.3.2 Identification of Compounds in Turmeric
As was done for ginger, both negative and positive ion mode ESI analyses were performed on turmeric rhizome extracts (Figure 2.17). The diarylheptanoids, including the three major curcuminoids, were detected by both negative and positive mode under optimized conditions. Some other compounds such phenolic acids and terpenoids were also detected.
Diarylheptanoids Gingerol-related compounds 6 X10 (Curcuminoid-related) 1.5
1.0
0.5
0.0 0 10203040506070 Total Ion Current (TIC) Chromatogram from negative ion (-)ESI-HPLC-MS
8 Diarylheptanoids Gingerol-related compounds X10 1.0 (Curcuminoid-related) Absolute Abundance
0.5
0.0 010203040506070 Total Ion Current (TIC) Chromatogram from positive ion (+)ESI-HPLC-MS
Time Figure 2.17. LC-MS analysis of ginger rhizome extracts. The top chromatogram and the bottom one are total ion current (TIC) chromatograms from negative ion (–)ESI-HPLC-MS and positive ion (+)ESI-HPLC-MS, respectively.
66
Identification of Diarylheptanoids from Turmeric
Here I will take one example on how we identify diarylheptanoids from turmeric.
The ESI-MS/MS spectra of some diarylheptanoids are shown in Figure 2.18. The top three pairs of spectra are from the three major curcuminoids in turmeric. Their fragmentation patterns in both modes are consistent as indicated (Figure 2.18). Another three pairs of MS/MS spectra (bottom of Figure 2.18) were also observed in both modes during the analyses of turmeric samples.
O O 5 187.1 x10 5 147.3 x10 -Ms 2(307.2) 33.0 min 14 225.2 14 A 6.0 C E +Ms 2(309.4) 33.0 min 1.0 4.0 D B 215.2 0.5 HO OH 143.6 307.2 2.0 121.3 189.2 Bisdemethoxycurcumin 239.2 309.4 0.0 0.0 O O 4 217.2 6 +Ms (339.4) 33.8 min 16 x10 -Ms 2(337.1) 33.8 min 16 x10 2 245.2 A DE MeO 6.0 1.0 255.2 A 4.0 C B 187.6 0.5 175.3 C HO OH 2.0 173.9 337.1 147.4 177.4 269.2 339.4 Demethoxycurcumin 0.0 0.0 5 O O 5 -Ms (367.4) 34.5 min 217.2 245.2 x10 2 18 x10 +Ms (369.3) 34.5 min 18 2 D OMe A C MeO 6.0 1.0 177.4 E B 4.0 175.3 285.2 0.5 173.7 HO OH 367.4 2.0 299.2 369.3 Curcumin 0.0 0.0
6 1 O O 4 189.2 x10 147.3 7 x10 -Ms (309.1) 32.8 min 13 +Ms 2(311.4) 32.8 min 13 A 2 C 4.0 1.0 4 G F 2.0 119.6 0.5 HO OH 203.2 309.1 119.4 225.2 311.4 0.0 0.0 4 -Ms (339.2) 33.5 min 219.2 5 177.3 +Ms (341.4) 33.5 min 15 O O x10 2 15 x10 C 2 A 147.3 C MeO 0.8 2.0 G A 339.2 0.4 G 149.3 189.3 1.0 245.2 323.2 HO OH 119.5 F 341.4 0.0 0.0 6 O O 4 -Ms (369.3) 34.1 min 219.2 177.2 x10 2 17 x10 +Ms 2(371.2) 34.1 min 17 A C MeO OMe 0.8 2.0 G 149.5 F 0.4 1.0 233.1 245.2 HO OH 134.7 175.4 369.3 145.3 285.2 353.1 371.2 0.0 0.0 50 100 150 200 250 300 350 m/z 50 100 150 200 250 300 350 m/z Dihydrocurcuminoids
(−)ESI-MS/MS (+)ESI-MS/MS Figure 2.18. Fragmentation of some diarylheptanoids in ESI-MS/MS. Note: (−)ESI-MS/MS (left column); (+)ESI-MS/MS (right column). Boxes indicate major product ions; circles indicate precursor ions.
These spectra showed very similar fragmentation patterns when compared to the three major curcuminoids. Both their precursor and major product ions were
67 differentiated by 2 mass units when compared with the three major curcuminoids. Based on this information, the structures of the three compounds were suggested to be related to the curcuminoids, but with the absence of one double between carbon 6 and carbon 7.
They are called dihydrocurcuminoids.
a) Fragmentation pathway in (−)ESI-MS/MS
HO
HO O - O O O H H R1 H O C R2 O R2 R1 R2 R2 O C H H O O O O O
307 (14) R1=R2=H 187(14) 143(14) 337 (16) R1=OMe, R2=H; or R1=H, R2=OMe 217, 187 (16) 173 (16) R2 C A B 367 (18) R1=R2=OMe 217 (18) 173(18) H 323 (11) R1=OH, R2=H; or R1=H, R2=OH 187, 203 (11) 143 (11) O
b) Fragmentation pathway in (+)ESI-MS/MS O O HO O OH R1 1 7 H O R1 R1 + R2 2 6 + R2 + OH R2 Ar1 307 (14) R1=R2=Ar1 O 337 (16) R1=Ar3, R2=Ar1; or R1=Ar1, R2=Ar3 R1H 367 (18) R1=R2=Ar3 HO OH O OH MeO R2 O O R1 + R2 + Ar3 + 225 (14) R1 R2 E 255 (16) 285 (18) 147 (14) 215 (14) CD177, 147 (16) 245, 215 (16) 177 (18) 245 (18) Scheme 2.3. Proposed fragmentation pathway of three major curcuminoids in ESI-MS/MS.
The structures of three major curcuminoids were rationalized by their fragmentation behavior in both negative and positive mode (Scheme 2.3). For example, the base peak (product ion A) in the negative mode was obtained by a β-hydrogen shift rearrangement (Scheme 2.3). The formation of three major product ions in the positive
68 mode was also rationalized in this proposed pathway. Similarly, the structures of the three dihydrocurcuminoids were also rationalized by their fragmentation behavior
(Scheme 2.4) in both negative and positive mode.
a) Fragmentation pathway in (−)ESI-MS/MS
OH O HO O O OH H H H H H OR H R1 R2 R1 R2 R1 R2 O O O O O O 309 (13) R1=R2=H 339 (15A) R1=OMe, R2=H; or ( 15B) R1=H, R2=OMe OH HO H 369 (17) R1=R2=OMe R2 R1 O O C O R2 O HO O H H H C R2 R1 R1 O C O O O 189 (13) 203 (13) 119 (13) A 219 (15A), 189 ( 15B) F 203 (15) G 149 (15A), 119 ( 15B) 219 (17) 233 (17) 149 (17)
b) Fragmentation pathway in (+)ESI-MS/MS
OH HO HO O OH R2 O 1 7 R1 + R2 R1 + MeO 26 Ar1 Ar3 309 (13) R1=R2=Ar1 147 (13) 339 (15A) R1=Ar3, R2=Ar1; or ( 15B)R1=Ar1, R2=Ar3 C 177 (15A), 147 ( 15B) 369 (17) R1=R2=Ar3 177 (17)
Scheme 2.4. Proposed fragmentation pathway of three dihydrocurcuminoids in ESI-MS/MS.
Other diarylheptanoids (Figure 2.19) in turmeric were also identified by comparing their mass spectra with authentic compounds or against each other. These compounds are also from different homologous series, differentiated by the substituent
69 groups on the aromatic ring and/or on the heptane chain. Some new diarylheptanoids (see
Figure 2.19) were identified from turmeric.
OH OH O OH O O O OH R1 R2 R1 R2 R1 R2 R1 R2 (1):R1=R2=Ar1, MW=316 (2): R1=R2=Ar1, MW=314 (12): R1=R2=Ar1, MW=312 (7): R1=R2=Ar2, MW=344 *
O O O O O O OH R1 R2 R1 R2 R1 R2 (13): R1=R2=Ar1, MW=310 (14): R1=R2=Ar1, MW=308 9 (15); R1=Ar1, R2=Ar3, MW=340 (16): R1=Ar3, R2=Ar1, MW=338 ( ): R1=Ar3, R2=Ar1, MW=356 * 10 (17): R1=R2=Ar3, MW=370 (18): R1=R2=Ar3, MW=368 ( ): R1=Ar3, R2=Ar2, MW=372 * (11): R1=Ar2, R2=Ar1, MW=324 *
O O R1 R2 R1 R2 (3): R1=Ar3, R2=Ar1, MW=324; (4): R1=R3=Ar3,MW=354; (6): R1=Ar3, R2=Ar1, MW=322 *; (8): R1=R2=Ar3, MW=352 (5): R1=Ar4, R2=Ar3, MW=384 *
OMe HO HO HO HO
HO MeO MeO Ar1 Ar2 Ar3 Ar4
Figure 2.19. Chemical structures and molecular weights of diarylheptanoids identified in turmeric rhizome. Note: compounds in red color are new compounds.
Identification of Other Compounds from Turmeric
Some other compounds such phenolic compounds and terpenoids were also identified from turmeric by LC-MS/MS (Figure 2.20).
70
COOH COOH
R1 R1 R2 R2 Caffeic acid (1): R1=R2=OH, MW=180 p-Hydroxy-hydrocinnamic acid (2): R1=H, R2=OH, MW=166 p-Hydroxy-cinnamic acid (3): R1=H, R2=OH, MW=164 4-Hydroxy-3-methoxy-hydrocinnamic acid (4): Ferulic acid (5): R1=OMe, R2=OH, MW=194 R1=OMe, R2=OH, MW=196
O O O O
OH a b Turmeronol A or B (19): ar-Turmerone (20): Turmerone (a) or Curlone (b) (21): MW=232 MW=217 MW=218
Figure 2.20. Chemical structures and molecular weights of phenolic acids and terpenoids identified in turmeric rhizome.
In all, more than 300 compounds were detected from the ginger rhizome using
LC-MS/MS, and 70 of these were identified as diarylheptanoids and gingerol-related
compounds. On the other hand, more than 150 compounds were detected from turmeric rhizome by LC-MS/MS, and 27 of these were identified. The majority of these compounds are diarylheptanoids including the three major curcuminoids. Using this
technique, many compounds, including a number of new compounds, have been
identified.
71
REFERENCES
1. Sparreboom, A.; Cox, M. C.; Acharya, M. R.; Figg, W. D. Herbal remedies in the United States: potential adverse interactions with anticancer agents. J. Clin. Oncol. 2004, 22, 2489-2503.
2. Purcell, K. Survey shows 36% of U. S. adults use CAM. HerbalGram 2005, 66, 65-66.
3. Astin, J. A. Why patients use alternative medicine: results of a national study. J. Am. Med. Assoc. 1998, 279, 1548-1553.
4. Johnson, B. Prevention magazine assesses use of dietary supplements. HerbalGram 2000, 48, 65.
5. Jonas, W. B. Alternative medicine - Learning from the past, examining the present, advancing to the future. J. Am. Med. Assoc. 1998, 280, 1616-1618.
6. Kaufman, D. W.; Kelly, J. P.; Rosenberg, L.; Anderson, T. E.; Mitchell, A. A. Recent patterns of medication use in the ambulatory adult population of the United States: the Slone survey. J. Am. Med. Assoc. 2002, 287, 337-344.
7. Cardellina, J. H. Challenges and opportunities confronting the botanical dietary supplement industry. J. Nat. Prod. 2002, 65, 1073-1084.
8. Blumenthal, M. Herb market levels after five years of boom: 1999 sales in mainstream market up only 11% in first half of 1999 after 55% increase in 1998. HerbalGram 1999, 47, 64.
9. Blumenthal, M. Herb sales down 7.4 percent in mainsteam market. HerbalGram 2005, 66, 63.
10. Fugh-Berman, A. Herb-drug interactions. Lancet 2000, 355, 134-138.
11. The Arizona Center for Phytomedicine. [Online] Accessed at http://acprx.pharmacy.arizona.edu/about/sponsors.html May 5, 2005.
12. Ammon, H. P.; Safayhi, H.; Mack, T.; Sabieraj, J. Mechanism of antiinflammatory actions of curcumine and boswellic acids. J. Ethnopharmacol. 1993, 38, 113-119.
13. Srivastava, K. C.; Mustafa, T. Ginger (Zingiber officinale) in rheumatism and musculoskeletal disorders. Med. Hypotheses 1992, 39, 342-348.
72
14. Taylor, P. C.; Feldmann, M. Rheumatoid arthritis: pathogenic mechanisms and therapeutic targets. Drug Discovery Today: Disease Mechanisms 2004, 1, 289- 295.
15. Darshan, S.; Doreswamy, R. Patented antiinflammatory plant drug development from traditional medicine. Phytother. Res. 2004, 18, 343-357.
16. Spiegel, B. M.; Chiou, C. F.; Ofman, J. J. Minimizing complications from nonsteroidal antiinflammatory drugs: cost-effectiveness of competing strategies in varying risk groups. Arthritis Rheum. 2005, 53, 185-197.
17. Wolfe, M. M.; Lichtenstein, D. R.; Singh, G. Gastrointestinal toxicity of nonsteroidal antiinflammatory drugs. N. Engl. J. Med. 1999, 340, 1888-1899.
18. Romanovsky, A. A. Vioxx, Celebrex, Bextra. Do we have a new target for anti- inflammatory and antipyretic therapy? Am. J. Physiol-Reg. I 2005, 288, 1098- 1099.
19. Eisenberg, D. M.; Davis, R. B.; Ettner, S. L.; Appel, S.; Wilkey, S.; Van Rompay, M.; Kessler, R. C. Trends in alternative medicine use in the United States, 1990- 1997: results of a follow-up national survey. J. Am. Med. Assoc. 1998, 280, 1569- 1575.
20. Tindle, H. A.; Davis, R. B.; Phillips, R. S.; Eisenberg, D. M. Trends in use of complementary and alternative medicine by US adults: 1997-2002. Altern. Ther. Health Med. 2005, 11, 42-49.
21. Langner, E.; Greifenberg, S.; Gruenwald, O. Ginger: history and use. Adv. Ther. 1998, 15, 25-44.
22. Grant, K. L.; Lutz, R. Ginger. Am. J. Health Syst. Pharm. 2000, 57, 945-947.
23. Vutyavanich, T.; Kraisarin, T.; Ruangsri, R. A. Ginger for nausea and vomiting in pregnancy: Randomized, double-masked, placebo-controlled trial. Obstet. Gynecol. 2001, 97, 577-582.
24. Lacroix, R.; Eason, E.; Melzack, R. Nausea and vomiting during pregnancy: A prospective study of its frequency, intensity, and patterns of change. Am. J. Obstet. Gynecol. 2000, 182, 931-937.
25. Lien, H. C.; Sun, W. M.; Chen, Y. H.; Kim, H.; Hasler, W.; Owyang, C. Effects of ginger on motion sickness and gastric slow-wave dysrhythmias induced by circular vection. Am. J. Physiol. Gastrointest. Liver Physiol. 2003, 284, 481-489.
73
26. Altman, R. D.; Marcussen, K. C. Effects of a ginger extract on knee pain in patients with osteoarthritis. Arthritis Rheum. 2001, 44, 2531-2538.
27. Wigler, I.; Grotto, I.; Caspi, D.; Yaron, M. The effects of Zintona EC (a ginger extract) on symptomatic gonarthritis. Osteoarthritis Cartilage 2003, 11, 783-789.
28. Recommendations for the medical management of osteoarthritis of the hip and knee: 2000 update. American College of Rheumatology Subcommittee on Osteoarthritis Guidelines. Arthritis Rheum. 2000, 43, 1905-1915.
29. Schnitzer, T. J. Update of ACR guidelines for osteoarthritis: role of the coxibs. J. Pain Symptom Manage 2002, 23, S24-30; discussion S31-24.
30. Flake, Z. A.; Scalley, R. D.; Bailey, A. G. Practical selection of antiemetics. Am. Fam. Physician 2004, 69, 1169-1174.
31. Keating, A.; Chez, R. A. Ginger syrup as an antiemetic in early pregnancy. Altern. Ther. Health Med. 2002, 8, 89-91.
32. Smith, C.; Crowther, C.; Willson, K.; Hotham, N.; McMillian, V. A randomized controlled trial of ginger to treat nausea and vomiting in pregnancy. Obstet. Gynecol. 2004, 103, 639-645.
33. Willetts, K. E.; Ekangaki, A.; Eden, J. A. Effect of a ginger extract on pregnancy- induced nausea: a randomised controlled trial. Aust. N. Z. J. Obstet. Gynaecol. 2003, 43, 139-144.
34. Stewart, J. J.; Wood, M. J.; Wood, C. D.; Mims, M. E. Effects of ginger on motion sickness susceptibility and gastric function. Pharmacology 1991, 42, 111- 120.
35. Afzal, M.; Al-Hadidi, D.; Menon, M.; Pesek, J.; Dhami, M. S. Ginger: an ethnomedical, chemical and pharmacological review. Drug Metabol. Drug Interact. 2001, 18, 159-190.
36. Govindarajan, V. S. Ginger-chemistry, technology, and quality evaluation.2. CRC Cr. Rev. Food Sci. 1982, 17, 189-258.
37. Govindarajan, V. S. Ginger chemistry, technology, and quality evaluation.1. CRC Cr. Rev. Food Sci. 1982, 17, 1-96.
38. Ruedi, P.; Juch, M. Chemistry and biological activities of long-chain alkyloxy- catechols of the [n]-gingerol-type. Curr. Org. Chem. 1999, 3, 623-646.
74
39. Gong, F.; Fung, Y. S.; Liang, Y. Z. Determination of volatile components in ginger using gas chromatography-mass spectrometry with resolution improved by data processing techniques. J. Agr. Food Chem. 2004, 52, 6378-6383.
40. Singh, G.; Maurya, S.; Catalan, C.; de Lampasona, M. P. Studies on essential oils, Part 42: chemical, antifungal, antioxidant and sprout suppressant studies on ginger essential oil and its oleoresin. Flavour Frag. J. 2005, 20, 1-6.
41. Jolad, S. D.; Lantz, R. C.; Solyom, A. M.; Chen, G. J.; Bates, R. B.; Timmermann, B. N. Fresh organically grown ginger (Zingiber officinale): composition and effects on LPS-induced PGE2 production. Phytochemistry 2004, 65, 1937-1954.
42. Denniff, P.; Whiting, D. A. Biosynthesis of [6]-gingerol, pungent principle of Zingiber-officinale. J. Chem. Soc. Chem. Comm. 1976, 711-712.
43. Connell, D. W.; Sutherla, M. D. Reexamination of gingerol, shogaol, and zingerone, the pungent principles of ginger (Zingiber officinale Roscoe). Aust. J. Chem. 1969, 22, 1033-1043.
44. Mustafa, T.; Srivastava, K. C.; Jensen, K. B. Drug development report. 9. pharmacology of ginger, Zingiber-officinale. J. Drug Dev. 1993, 6, 25-39.
45. Endo, K.; Kanno, E.; Oshima, Y. Structures of antifungal diarylheptanoids, gingerenone-A, gingerenone-B, gingerenone-C and isogingerenone-B, isolated from the rhizomes of Zingiber-officinale. Phytochemistry 1990, 29, 797-799.
46. Kikuzaki, H.; Nakatani, N. Cyclic diarylheptanoids from rhizomes of Zingiber officinale. Phytochemistry 1996, 43, 273-277.
47. Ma, J. P.; Jin, X. L.; Yang, L.; Liu, Z. L. Diarylheptanoids from the rhizomes of Zingiber officinale. Phytochemistry 2004, 65, 1137-1143.
48. Chen, C. C.; Ho, C. T. Gas-chromatographic analysis of volatile components of ginger oil (Zingiber-officinale Roscoe) extracted with liquid carbon-dioxide. J. Agr. Food Chem. 1988, 36, 322-328.
49. Harvey, D. J. Gas-chromatographic and mass-spectrometric studies of ginger constituents-identification of gingerdiones and new hexahydrocurcumin analogs. J. Chromatogr. 1981, 212, 75-84.
50. Masada, Y.; Inoue, T.; Hashimot, K.; Fujioka, M.; Uchino, C. Studies on constituents of ginger (Zingiber-officinale-Roscoe) by GC-MS. Yakugaku Zasshi 1974, 94, 735-738.
75
51. Balladin, D. A.; Headley, O.; Chang-Yen, I.; McGaw, D. R. High pressure liquid chromatographic analysis of the main pungent principles of solar dried West Indian ginger (Zingiber officinale Roscoe). Renew. Energ. 1998, 13, 531-536.
52. Connell, D. W.; McLachlan, R. Natural pungent compounds. IV. Examination of gingerols, shogaols, paradols and related compounds by thin layer and gas chromatography. J. Chromatogr. 1972, 67, 29-35.
53. He, X. G.; Bernart, M. W.; Lian, L. Z.; Lin, L. Z. High-performance liquid chromatography electrospray mass spectrometric analysis of pungent constituents of ginger. Journal of Chromatography A 1998, 796, 327-334.
54. Huang, H. Y.; Kuo, K. L.; Hsieh, Y. Z. Determination of cinnamaldehyde, cinnamic acid, paeoniflorin, glycyrrhizin and [6]-gingerol in the traditional Chinese medicinal preparation Kuei-chih-tang by cyclodextrin-modified micellar electrokinetic chromatography. J. Chromatogr. A 1997, 771, 267-274.
55. Kress, W. J.; Prince, L. M.; Williams, K. J. The phylogeny and a new classification of the gingers (Zingiberaceae): evidence from molecular data. Am. J. Bot. 2002, 89, 1682-1696.
56. Adewunmi, C. O.; Oguntimein, B. O.; Furu, P. Molluscicidal and antischistosomal activities of Zingiber officinale. Planta Med. 1990, 56, 374-376.
57. Ali, M. S.; Banskota, A. H.; Tezuka, Y.; Saiki, I.; Kadota, S. Antiproliferative activity of diarylheptanoids from the seeds of Alpinia blepharocalyx. Biol. Pharm. Bull. 2001, 24, 525-528.
58. Bhattarai, S.; Tran, V. H.; Duke, C. C. The stability of gingerol and shogaol in aqueous solutions. J. Pharm. Sci. 2001, 90, 1658-1664.
59. Ficker, C.; Smith, M. L.; Akpagana, K.; Gbeassor, M.; Zhang, J.; Durst, T.; Assabgui, R.; Arnason, J. T. Bioassay-guided isolation and identification of antifungal compounds from ginger. Phytother. Res. 2003, 17, 897-902.
60. Flynn, D. L.; Rafferty, M. F.; Boctor, A. M. Inhibition of 5- hydroxyeicosatetraenoic acid (5-HETE) formation in intact human neutrophils by naturally occurring diarylheptanoids: inhibitory activities of curcuminoids and yakuchinones. Prostag. Leukotr. Med. 1986, 22, 357-360.
61. Hikino, H.; Kiso, Y.; Kato, N.; Hamada, Y.; Shioiri, T.; Aiyama, R.; Itokawa, H.; Kiuchi, F.; Sankawa, U. Antihepatotoxic actions of gingerols and diarylheptanoids. J. Ethnopharmacol. 1985, 14, 31-39.
76
62. Ishida, J.; Kozuka, M.; Tokuda, H.; Nishino, H.; Nagumo, S.; Lee, K. H.; Nagai, M. Chemopreventive potential of cyclic diarylheptanoids. Bioorgan. Med. Chem. 2002, 10, 3361-3365.
63. Ishida, J.; Kozuka, M.; Wang, H.; Konoshima, T.; Tokuda, H.; Okuda, M.; Yang Mou, X.; Nishino, H.; Sakurai, N.; Lee, K. H.; Nagai, M. Antitumor-promoting effects of cyclic diarylheptanoids on Epstein-Barr virus activation and two-stage mouse skin carcinogenesis. Cancer Lett. 2000, 159, 135-140.
64. Koo, K. L.; Ammit, A. J.; Tran, V. H.; Duke, C. C.; Roufogalis, B. D. Gingerols and related analogues inhibit arachidonic acid-induced human platelet serotonin release and aggregation. Thromb. Res. 2001, 103, 387-397.
65. Masuda, Y.; Kikuzaki, H.; Hisamoto, M.; Nakatani, N. Antioxidant properties of gingerol related compounds from ginger. Biofactors 2004, 21, 293-296.
66. Shin, D.; Kinoshita, K.; Koyama, K.; Takahashi, K. Antiemetic principles of Alpinia officinarum. J. Nat. Prod. 2002, 65, 1315-1318.
67. Bode, A. M.; Ma, W. Y.; Surh, Y. J.; Dong, Z. Inhibition of epidermal growth factor-induced cell transformation and activator protein 1 activation by [6]- gingerol. Cancer Res. 2001, 61, 850-853.
68. Nurtjahja-Tjendraputra, E.; Ammit, A. J.; Roufogalis, B. D.; Tran, V. H.; Duke, C. C. Effective anti-platelet and COX-1 enzyme inhibitors from pungent constituents of ginger. Thromb. Res. 2003, 111, 259-265.
69. Kim, S. O.; Kundu, J. K.; Shin, Y. K.; Park, J. H.; Cho, M. H.; Kim, T. Y.; Surh, Y. J. [6]-Gingerol inhibits COX-2 expression by blocking the activation of p38 MAP kinase and NF-kappaB in phorbol ester-stimulated mouse skin. Oncogene 2005, 24, 2558-2567.
70. Young, H. Y.; Luo, Y. L.; Cheng, H. Y.; Hsieh, W. C.; Liao, J. C.; Peng, W. H. Analgesic and anti-inflammatory activities of [6]-gingerol. J. Ethnopharmacol. 2005, 96, 207-210.
71. Park, K. K.; Chun, K. S.; Lee, J. M.; Lee, S. S.; Surh, Y. J. Inhibitory effects of [6]-gingerol, a major pungent principle of ginger, on phorbol ester-induced inflammation, epidermal ornithine decarboxylase activity and skin tumor promotion in ICR mice. Cancer Lett. 1998, 129, 139-144.
72. Dedov, V. N.; Tran, V. H.; Duke, C. C.; Connor, M.; Christie, M. J.; Mandadi, S.; Roufogalis, B. D. Gingerols: a novel class of vanilloid receptor (VR1) agonists. Br. J. Pharmacol. 2002, 137, 793-798.
77
73. Kim, S. O.; Chun, K. S.; Kundu, J. K.; Surh, Y. J. Inhibitory effects of [6]- gingerol on PMA-induced COX-2 expression and activation of NF-kappaB and p38 MAPK in mouse skin. Biofactors 2004, 21, 27-31.
74. Lee, J. L.; Mukhtar, H.; Bickers, D. R.; Kopelovich, L.; Athar, M. Cyclooxygenases in the skin: pharmacological and toxicological implications. Toxicol. Appl. Pharmacol. 2003, 192, 294-306.
75. Joe, B.; Vijaykumar, M.; Lokesh, B. R. Biological properties of curcumin-cellular and molecular mechanisms of action. Crit. Rev. Food Sci. Nutr. 2004, 44, 97-111.
76. Ammon, H. P.; Anazodo, M. I.; Safayhi, H.; Dhawan, B. N.; Srimal, R. C. Curcumin: a potent inhibitor of leukotriene B4 formation in rat peritoneal polymorphonuclear neutrophils (PMNL). Planta Med. 1992, 58, 226.
77. Grant, K. L.; Schneider, C. D. Turmeric. Am. J. Health Syst. Pharm. 2000, 57, 1121-1122.
78. Sasaki, Y.; Fushimi, H.; Cao, H.; Cai, S. Q.; Komatsu, K. Sequence analysis of Chinese and Japanese Curcuma drugs on the 18S rRNA gene and trnK gene and the application of amplification-refractory mutation system analysis for their authentication. Biol. Pharm. Bull. 2002, 25, 1593-1599.
79. Aggarwal, B. B.; Kumar, A.; Bharti, A. C. Anticancer potential of curcumin: preclinical and clinical studies. Anticancer Res. 2003, 23, 363-398.
80. Chainani-Wu, N. Safety and anti-inflammatory activity of curcumin: a component of tumeric (Curcuma longa). J. Altern. Complement Med. 2003, 9, 161-168.
81. Deodhar, S. D.; Sethi, R.; Srimal, R. C. Preliminary study on antirheumatic activity of curcumin (diferuloyl methane). Indian J. Med. Res. 1980, 71, 632-634.
82. Lal, B.; Kapoor, A. K.; Agrawal, P. K.; Asthana, O. P.; Srimal, R. C. Role of curcumin in idiopathic inflammatory orbital pseudotumours. Phytother. Res. 2000, 14, 443-447.
83. Lal, B.; Kapoor, A. K.; Asthana, O. P.; Agrawal, P. K.; Prasad, R.; Kumar, P.; Srimal, R. C. Efficacy of curcumin in the management of chronic anterior uveitis. Phytother. Res. 1999, 13, 318-322.
84. Satoskar, R. R.; Shah, S. J.; Shenoy, S. G. Evaluation of antiinflammatory property of curcumin (diferuloyl methane) in patients with postoperative inflammation. Int. J. Clin. Pharm. Th. 1986, 24, 651-654.
78
85. Cheng, A. L.; Hsu, C. H.; Lin, J. K.; Hsu, M. M.; Ho, Y. F.; Shen, T. S.; Ko, J. Y.; Lin, J. T.; Lin, B. R.; Ming-Shiang, W.; Yu, H. S.; Jee, S. H.; Chen, G. S.; Chen, T. M.; Chen, C. A.; Lai, M. K.; Pu, Y. S.; Pan, M. H.; Wang, Y. J.; Tsai, C. C.; Hsieh, C. Y. Phase I clinical trial of curcumin, a chemopreventive agent, in patients with high-risk or pre-malignant lesions. Anticancer Res. 2001, 21, 2895- 2900.
86. Sharma, R. A.; Euden, S. A.; Platton, S. L.; Cooke, D. N.; Shafayat, A.; Hewitt, H. R.; Marczylo, T. H.; Morgan, B.; Hemingway, D.; Plummer, S. M.; Pirmohamed, M.; Gescher, A. J.; Steward, W. P. Phase I clinical trial of oral curcumin: biomarkers of systemic activity and compliance. Clin. Cancer Res. 2004, 10, 6847-6854.
87. Sharma, R. A.; McLelland, H. R.; Hill, K. A.; Ireson, C. R.; Euden, S. A.; Manson, M. M.; Pirmohamed, M.; Marnett, L. J.; Gescher, A. J.; Steward, W. P. Pharmacodynamic and pharmacokinetic study of oral Curcuma extract in patients with colorectal cancer. Clin. Cancer Res. 2001, 7, 1894-1900.
88. He, X. G.; Lin, L. Z.; Lian, L. Z.; Lindenmaier, M. Liquid chromatography electrospray mass spectrometric analysis of curcuminoids and sesquiterpenoids in turmeric (Curcuma longa). J. Chromatogr. A 1998, 818, 127-132.
89. Hiserodt, R.; Hartman, T. G.; Ho, C. T.; Rosen, R. T. Characterization of powdered turmeric by liquid chromatography mass spectrometry and gas chromatography mass spectrometry. J. Chromatogr. A 1996, 740, 51-63.
90. Park, S. Y.; Kim, D. S. Discovery of natural products from Curcuma longa that protect cells from beta-amyloid insult: a drug discovery effort against Alzheimer's disease. J. Nat. Prod. 2002, 65, 1227-1231.
91. Chattopadhyay, I.; Biswas, K.; Bandyopadhyay, U.; Banerjee, R. K. Turmeric and curcumin: biological actions and medicinal applications. Curr. Sci. India 2004, 87, 44-53.
92. Gupta, A. P.; Gupta, M. M.; Kumar, S. Simultaneous determination of curcuminoids in Curcuma samples using high performance thin layer chromatography. J. Liq Chromatogr. R. T. 1999, 22, 1561-1569.
93. Inoue, K.; Hamasaki, S.; Yoshimura, Y.; Yamada, M.; Nakamura, M.; Ito, Y.; Nakazawa, H. Validation of LC/electrospray-MS for determination of major curcuminoids in foods. J. Liq. Chromatogr. R. T. 2003, 26, 53-62.
79
94. Jayaprakasha, G. K.; Rao, L. J. M.; Sakariah, K. K. Improved HPLC method for the determination of curcumin, demethoxycurcumin, and bisdemethoxycurcumin. J. Agr. Food Chem. 2002, 50, 3668-3672.
95. Lechtenberg, M.; Quandt, B.; Nahrstedt, A. Quantitative determination of curcuminoids in Curcuma rhizomes and rapid differentiation of Curcuma domestica Val. and Curcuma xanthorrhiza Roxb. by capillary electrophoresis. Phytochem. Analysis 2004, 15, 152-158.
96. Sun, X. H.; Gao, C. L.; Cao, W. D.; Yang, X. R.; Wang, E. K. Capillary electrophoresis with amperometric detection of curcumin in Chinese herbal medicine pretreated by solid-phase extraction. J. Chromatogr. A 2002, 962, 117- 125.
97. Richmond, R.; PomboVillar, E. Gas chromatography mass spectrometry coupled with pseudo-sadtler retention indices, for the identification of components in the essential oil of Curcuma longa L. J. Chromatogr. A 1997, 760, 303-308.
98. Bansal, R. P.; Bahl, J. R.; Garg, S. N.; Naqvi, A. A.; Kumar, S. Differential chemical compositions of the essential oils of the shoot organs, rhizomes and rhizoids in the turmeric Curcuma longa grown in Indo-grangetic plains. Pharm. Biol. 2002, 40, 384-389.
99. Raina, V. K.; Srivastava, S. K.; Jain, N.; Ahmad, A.; Syamasundar, K. V.; Aggarwal, K. K. Essential oil composition of Curcuma longa L. cv. Roma from the plains of northern India. Flavour Frag. J. 2002, 17, 99-102.
100. Cao, H.; Sasaki, Y.; Fushimi, H.; Komatsu, K. Molecular analysis of medicinally- used Chinese and Japanese Curcuma based on 18S rRNA gene and trnK gene sequences. Biol. Pharm. Bull. 2001, 24, 1389-1394.
101. Araujo, C. C.; Leon, L. L. Biological activities of Curcuma longa L. Mem. Inst. Oswaldo. Cruz. 2001, 96, 723-728.
102. Miquel, J.; Bernd, A.; Sempere, J. M.; Diaz-Alperi, J.; Ramirez, A. The Curcuma antioxidants: pharmacological effects and prospects for future clinical use. a review. Arch. Gerontol. Geriatr. 2002, 34, 37-46.
103. Miyakoshi, M.; Yamaguchi, Y.; Takagaki, R.; Mizutani, K.; Kambara, T.; Ikeda, T.; Zaman, M. S.; Kakihara, H.; Takenaka, A.; Igarashi, K. Hepatoprotective effect of sesquiterpenes in turmeric. Biofactors 2004, 21, 167-170.
104. Roth, G. N.; Chandra, A.; Nair, M. G. Novel bioactivities of Curcuma longa constituents. J. Nat. Prod. 1998, 61, 542-545.
80
105. Tilak, J. C.; Banerjee, M.; Mohan, H.; Devasagayam, T. P. Antioxidant availability of turmeric in relation to its medicinal and culinary uses. Phytother. Res. 2004, 18, 798-804.
106. Egan, M. E.; Pearson, M.; Weiner, S. A.; Rajendran, V.; Rubin, D.; Glockner- Pagel, J.; Canny, S.; Du, K.; Lukacs, G. L.; Caplan, M. J. Curcumin, a major constituent of turmeric, corrects cystic fibrosis defects. Science 2004, 304, 600- 602.
107. Gescher, A. J.; Sharma, R. A.; Steward, W. P. Cancer chemoprevention by dietary constituents: a tale of failure and promise. Lancet Oncol. 2001, 2, 371-379.
108. Jordan, W. C.; Drew, C. R. Curcumin--a natural herb with Anti-HIV activity. J Natl. Med. Assoc. 1996, 88, 333.
109. Kelloff, G. J.; Crowell, J. A.; Hawk, E. T.; Steele, V. E.; Lubet, R. A.; Boone, C. W.; Covey, J. M.; Doody, L. A.; Omenn, G. S.; Greenwald, P.; Hong, W. K.; Parkinson, D. R.; Bagheri, D.; Baxter, G. T.; Blunden, M.; Doeltz, M. K.; Eisenhauer, K. M.; Johnson, K.; Knapp, G. G.; Longfellow, D. G.; Malone, W. F.; Nayfield, S. G.; Seifried, H. E.; Swall, L. M.; Sigman, C. C. Strategy and planning for chemopreventive drug development: clinical development plans II. J. Cell Biochem. Suppl. 1996, 26, 54-71.
110. Yang, F.; Lim, G. P.; Begum, A. N.; Ubeda, O. J.; Simmons, M. R.; Ambegaokar, S. S.; Chen, P. P.; Kayed, R.; Glabe, C. G.; Frautschy, S. A.; Cole, G. M. Curcumin inhibits formation of amyloid beta oligomers and fibrils, binds plaques, and reduces amyloid in vivo. J. Biol. Chem. 2005, 280, 5892-5901.
111. Ji, M. J.; Choi, J.; Lee, J.; Lee, Y. Induction of apoptosis by Ar-turmerone on various cell lines. Int. J. Mol. Med. 2004, 14, 253-256.
112. Chan, M. M. Inhibition of tumor necrosis factor by curcumin, a phytochemical. Biochem. Pharmacol. 1995, 49, 1551-1556.
113. Brennan, P.; O'Neill, L. A. Inhibition of nuclear factor kappaB by direct modification in whole cells--mechanism of action of nordihydroguaiaritic acid, curcumin and thiol modifiers. Biochem. Pharmacol. 1998, 55, 965-973.
114. Jobin, C.; Bradham, C. A.; Russo, M. P.; Juma, B.; Narula, A. S.; Brenner, D. A.; Sartor, R. B. Curcumin blocks cytokine-mediated NF-kappa B activation and proinflammatory gene expression by inhibiting inhibitory factor I-kappa B kinase activity. J. Immunol. 1999, 163, 3474-3483.
81
115. Plummer, S. M.; Holloway, K. A.; Manson, M. M.; Munks, R. J.; Kaptein, A.; Farrow, S.; Howells, L. Inhibition of cyclo-oxygenase 2 expression in colon cells by the chemopreventive agent curcumin involves inhibition of NF-kappaB activation via the NIK/IKK signalling complex. Oncogene 1999, 18, 6013-6020.
116. Joe, B.; Lokesh, B. R. Effect of curcumin and capsaicin on arachidonic acid metabolism and lysosomal enzyme secretion by rat peritoneal macrophages. Lipids 1997, 32, 1173-1180.
117. Huang, M. T.; Newmark, H. L.; Frenkel, K. Inhibitory effects of curcumin on tumorigenesis in mice. J Cell Biochem. Suppl. 1997, 27, 26-34.
118. Rao, C. V.; Rivenson, A.; Simi, B.; Reddy, B. S. Chemoprevention of colon cancer by dietary curcumin. Ann. N. Y. Acad. Sci. 1995, 768, 201-204.
82
APPENDIX A
Manuscript “Metabolic profiling and phylogenetic analysis of medicinal Zingiber species: tools for authentication of ginger (Zingiber officinale Rosc.)”, in press
Phytochemistry.
83
Metabolic profiling and phylogenetic analysis of medicinal Zingiber species: tools for authentication of ginger (Zingiber officinale Rosc.) Hongliang Jianga,b,c,d, Zhengzhi Xiea,b,c,d, Hyunjo Kooa,b,c, Steven P. McLaughlina,b, Barbara N. Timmermanna,e, and David R. Ganga,b,c*
a Arizona Center for Phytomedicine Research, College of Pharmacy, University of Arizona, Tucson, AZ 85721, USA b Department of Plant Sciences, College of Agriculture and Life Sciences, University of Arizona, Tucson, AZ 85721, USA c Bio5 Institute, University of Arizona, Tucson, AZ 85721, USA d Department of Pharmaceutical Sciences, College of Pharmacy, University of Arizona, Tucson, AZ 85721, USA e Department of Pharmacology and Toxicology, College of Pharmacy, University of Arizona, Tucson, AZ 85721, USA
Molecular and chemical data can be used to distinguish ginger from other medicinal plants in Zingiber. In addition, phylogeny based on the DNA sequences (A) matched closely the phylogeny based on chemical profiles (B) for these Zingiber species.
L1 L1 AB63.7 Z. officinale 50.2 49.0 L15 Z. zerumbet L37 30.8
53.7 L37 Z. montanum L15 55.0
L46 Z. mioga L46
L31 Z. spectabile L31
84
Title: Metabolic profiling and phylogenetic analysis of medicinal Zingiber species: tools
for authentication of ginger (Zingiber officinale Rosc.)
Hongliang Jianga,b,c,d, Zhengzhi Xiea,b,c,d, Hyun Jo Kooa,b,c, Steven P. McLaughlina,b,
Barbara N. Timmermanna,e, and David R. Ganga,b,c*
aArizona Center for Phytomedicine Research, College of Pharmacy, University of
Arizona
bDepartment of Plant Sciences, College of Agriculture and Life Sciences, University of
Arizona
cBIO5 Institute, University of Arizona
dDepartment of Pharmaceutical Sciences, College of Pharmacy, University of Arizona
eDepartment of Pharmacology and Toxicology, College of Pharmacy, University of
Arizona
*Corresponding author:
David R. Gang
Department of Plant Sciences and BIO5 Institute, University of Arizona, Tucson, AZ
85721-0036, USA
Tel: 520-621-7154
Fax: 520-621-7186
email: [email protected]
85
Abstract
Phylogenetic analysis and metabolic profiling were used to investigate the diversity of plant material within the ginger species and between ginger and closely related species in the genus Zingiber (Zingiberaceae). In addition, anti-inflammatory data were obtained for the investigated species. Phylogenetic analysis demonstrated that all Z. officinale samples from different geographical origins were genetically indistinguishable. In contrast, other
Zingiber species were significantly divergent, allowing all species to be clearly distinguished using this analysis. In the metabolic profiling analysis, the Z. officinale samples derived from different origins showed no qualitative differences in major volatile compounds, although they did show some significant quantitative differences in non- volatile composition, particularly regarding the content of [6]-, [8]-, and [10]-gingerols, the most active anti-inflammatory components in this species. The differences in gingerol content were verified by HPLC. The metabolic profiles of other Zingiber species were very different, both qualitatively and quantitatively, when compared to Z. officinale and to each other. Comparative DNA sequence/chemotaxonomic phylogenetic trees showed that the chemical characters of the investigated species were able to generate essentially the same phylogenetic relationships as the DNA sequences. This supports the contention that chemical characters can be used effectively to identify relationships between plant species. Anti-inflammatory in vitro assays to evaluate the ability of all extracts from the
Zingiber species examined to inhibit LPS-induced PGE2 and TNF-α production suggested that bioactivity may not be easily predicted by either phylogenetic analysis or
86
gross metabolic profiling. Therefore, identification and quantification of the actual
bioactive compounds are required to guarantee the bioactivity of a particular Zingiber
sample even after performing authentication by molecular and/or chemical markers.
Key words: Zingiber officinale, Zingiberaceae, ginger, authentication, metabolic profiling GC/MS, HPLC, gingerols, anti-inflammatory, trnL-F, rps16
1. Introduction
Ginger (Zingiber officinale Rosc.), a member of the tropical and sub-tropical
Zingiberaceae, has been cultivated for thousands of years as a spice and for medicinal purposes. It is used extensively in Traditional Chinese Medicine to treat headaches, nausea and colds and in Ayurvedic and Western herbal medicinal practice for the treatment of arthritis, rheumatic disorders and muscular discomfort (Dedov et al., 2002).
This species contains biologically active constituents including the main pungent principles, the gingerols and shogaols. The gingerols, a series of chemical homologs differentiated by the length of their unbranched alkyl chains, were identified as the major active components in the fresh rhizome (Govindarajan, 1982), with [6]-gingerol (5- hydroxy-1-[4´-hydroxy-3´-methoxyphenyl] decan-3-one) being the most abundant. In addition, the shogaols, another homologous series and the dehydrated form of the gingerols, that result from the elimination of the OH group at C-5 and the consequent formation of a double bond between C-4 and C-5, are the predominant pungent
87 constituents in dried ginger (Connell and Sutherland, 1969; Mustafa et al., 1993). [6]-
Gingerol has been found to possess various pharmacological and physiological effects including anti-inflammatory, analgesic, antipyretic, gastroprotective, cardiotonic, and antihepatotoxic activities (Bhattarai et al., 2001; Jolad et al., 2004). Due to these properties, ginger has gained considerable attention as a botanical dietary supplement in the USA and Europe in recent years, and especially for its use in the treatment of chronic inflammatory conditions.
Despite ginger’s widespread medicinal and culinary uses, the authentication of ginger samples remains a difficult problem due to heterogeneity of the plant material, contamination with similar looking plants and by the purposeful adulteration of some commercial samples. Because of these problems, authentication of the raw material is very important to ensure that specific batches of dried, chipped or ground ginger are of the quality desired for the manufacture of reliable botanical dietary supplements.
To help address this problem, we investigated a large diversity of plant material within the ginger species and between ginger and several closely related species in the genus
Zingiber, using molecular and metabolic profiling coupled with anti-inflammatory activity data. The species analyzed were selected because they are have documented medicinal uses, specifically anti-inflammatory, are used as substitutes for ginger as spices/flavoring agents, or because they are sometimes mistaken in the popular market
88 for ginger (Ando, et al., 2005; Nakamura, et al., 2004; Miyoshi, et al., 2003; Murakami, et al., 2004; Yang, et al., 1999.).
2. Results and discussion
The major goals of this project were to investigate the effect of genetic diversity instead of environmental influences on the chemistry and bioactivity (anti-inflammatory activity) of ginger samples and to set a framework for the authentication of these important botanicals. To accomplish these goals, fresh frozen greenhouse-grown Zingiber samples were used for DNA sequence-based phylogenetic analysis, GC/MS-based metabolic profiling, HPLC quantitation of gingerols, and anti-inflammatory assays. As described in the Experimental section, all samples were grown at the same time under identical conditions in the same greenhouse to ensure that environmental effects were eliminated in this study.
2.1. Phylogenetic analysis based on molecular data
One of our objectives was to investigate the genetic variability of ginger obtained from different sources and of the interspecific differences between ginger (Z. officinale) and other medicinal Zingiber species. A phylogeny of 104 species in 41 genera representing all four tribes of the Zingiberaceae was reported by Kress et al. (2002). That study, which was based on DNA sequences of the nuclear internal transcribed spacer (ITS) and
89
plastid matK regions, did not include Z. officinale (ginger), Z. mioga, Z. montanum, Z. spectabile and Z. zerumbet. These two regions, as described in the Experimental section, are unsuitable for the phylogenetic determination of ginger specimens. The reason is that they do not amplify well using standard primers (for the matK-trnK flanking intergenic spacer regions) or because the gene (ITS) is present in more than one copy in the genome of ginger thus, leading to undeterminable sequence data from PCR-product based sequencing runs. We, therefore, used solely the rps16 and trnL-F regions for our analysis of Z. officinale, Z. mioga, Z. montanum, Z. spectabile, Z. zerumbet and Alpinia galanga.
All species were chosen because of their traditional use as medicinal plants and/or because they are used as adulterants of ginger (Langner et al., 1998). A. galanga also served as the outgroup for our phylogenetic analyses.
In the plant specimens that we examined, the intron of rps16 had a total aligned length of
742 bp, and the trnL-F region contained a total aligned length of 891 bp. The combined matrix with the indel characters contained 1633 bp. Phylogenetic analysis (using maximum parsimony, see Experimental) of the joined sequences from the intron of rps16 and the trnL-F region resulted in a single consensus unrooted parsimonious tree (Fig. 1A).
The consensus tree produced when rps16 and trnL-F regions were used independently to produce parsimonious trees did not differ from the tree produced when the datasets were combined (data not shown). All of the ginger (Z. officinale) samples contained the identical sequence over this entire region, resulting in a single, undifferentiated clade for these samples in the phylogenetic analysis, even though many of these sample lines had
90 been obtained from very different geographical origins (see Table 1). However, the sequence of the ginger samples differed from those of Z. mioga, Z. montanum, Z. spectabile, Z. zerumbet, and the outgroup Alpinia galanga for both the rps16 intron and the trnL-F region, resulting in clear delineation of all species in our analysis. In particular, there are many single nucleotide polymorphisms in these sequences, allowing us to distinguish ginger from other species based on the sequence data. Some of these sequence differences are illustrated in Fig. 2.
2.2. GC/MS-based metabolic profiling and fingerprinting of Zingiber species
Because medicinal ginger is often sold as ground rhizome powders or as alcoholic or non-polar solvent extracts from these powders, it may be very difficult to obtain DNA evidence to demonstrate adulteration or authentication of particular samples. Thus, we evaluated the utility of chemical characters derived from metabolic profiling experiments to reconstruct the phylogeny of the medicinal Zingiber species and to distinguish between species. Non-polar compounds were extracted with Methyl t-Butyl Ether (MTBE) and thereafter analyzed by GC/MS. Compounds detected, identified and quantified by
GC/MS are listed in Table 2. Many compounds present in small quantities were not included in this analysis because they could not be readily identified due to insufficient mass spectrum quality or because their relative concentration could not be adequately evaluated.
91
Based on the compounds that were detected and/or identified from the different samples,
we found that all ginger (Z. officinale) samples showed very similar metabolic
fingerprints, i.e., there was no apparent qualitative difference in their GC/MS
chromatograms. These plants were grown at the same time under identical conditions in
the same greenhouse. Thus, when all differential environmental factors were eliminated,
the ginger samples appeared to be chemically very similar, at least at the metabolic
fingerprint level (they produced the same compounds), even though these lines were
originally obtained from very diverse populations around the country. This result
matched the molecular data based on the joined sequences of the rps16 intron and the
trnL-F region. At the metabolic profile level, however, clear differences between lines
could be observed, where many of the compounds were found at different levels in the
different ginger lines. This suggests that genetic factors control not only which specific
compounds are produced, but at what levels. This is not surprising, but does suggest that
ginger obtained from different sources may have significantly different levels of active
compounds. This is addressed further below in section 2.3.
Interestingly, the phylogenetic trees generated using the metabolic profiling and the DNA
data were almost identical in structure (see Fig. 1B). The only observed difference was
that Z. zerumbet was found to be more closely related to Z. officinale based on molecular data whereas Z. montanum was more closely related to Z. officinale based on the chemical data. However, bootstrap support for these differences was not very strong.
Thus, major volatile chemical markers were very effective at distinguishing Zingiber
92
species and at reconstructing essentially the same phylogeny as was obtained using the
DNA sequence data, at least when the chemical characters were considered on a
presence/absence basis. Other phylogenetic studies using both molecular and chemical
data have been performed, such as with the genera Peltigera (Peltigeraceae) and Sticta
(Stictaceae). Studies of these two genera also showed that both types of characters are
useful and complementary (McDonald et al., 2003 and Miadlikowska and Lutzoni, 2000).
Many of the compounds identified in our metabolic profiling analysis could be used as
marker compounds to distinguish between the different Zingiber species. For example,
many of these compounds were only detected in Z. officinale samples and not in the other species that we examined. These included the gingerols and their derivatives, the shogaols and paradols, which have not been reported in any species besides Z. officinale,
and which represent unique marker compounds for this species. In addition, citronellal,
(E)- and (Z)-citral, (+)-cyclosativene, zingiberene, α-cubebene, germacrene D, cedr-8-ene,
and α-farnesene, among others (see Table 2), were also only present in Z. officinale.
These compounds, which were not detected in the other examined Zingiber species, could
also be used as chemical markers to distinguish Z. officinale from other Zingiber species.
Similarly, a number of other compounds were present in extracts from only one of the
other Zingiber species. For example, 3-carene and limonene were detected only in Z.
zerumbet; 1,3-cyclohexadiene, 1-methyl-4-(1-methyl)- and 4-isopropyl-1-methyl-2-
cyclohexen-1-ol were detected only in Z. montanum; and 1,4-Bis(methoxy)-triquinacene,
93
was found only in Z. mioga. These compounds can be used as markers for the
identification of ginger (Z. officinale) samples that have been adulterated by these other
species. Z. spectabile did not offer any known compound that distinguished it from the
other species. We observed, however, the presence of several unknown compounds that
were found only in Z. spectabile (Table 2). These compounds, identified as DRG-GM1-
N1-8.86-136-93-121, DRG-GM1-N1-9.03-134-119-105 and DRG-GM1-N1-9.11-136-
93-68, were named following the nomenclature rules outlined by Bino et al. (2004) for the naming of unknown compounds in metabolic profiling investigations.
We also identified a number of compounds that were not detected in Z. officinale but were detected in more than one of the other Zingiber species. These included 3-thujene;
1-methyl-4-(1-methylethyl)-1,4-cyclohexadiene; and p-menth-1-en-4-ol, among others
(Table 2). These compounds could also be used as markers to identify supposedly pure samples of Z. officinale that have actually been adulterated with Z. spectabile (or the other Zingiber species in which they were found).
2.3. Quantitation of gingerols using HPLC
In order to further investigate potential variation of chemical composition in samples of Z. officinale from different sources, three major bioactive components [6]-, [8]-, and [10]- gingerols were quantitatively determined (metabolite target analysis). Calibration curves were derived from three independent injections of five concentrations of [6], [8]-, and
94
[10]-gingerols versus the peak area. Linearity was found in the concentration range between 25 and 200 µg·ml-1, with high reproducibility and accuracy. Regression analysis
of the experimental data points showed a linear relationship with excellent correlation
coefficients (r2) for [6]-, [8]-, and [10]-gingerols, being 0.999 for each, suggesting high
precision in this analysis. The linear regression equations for the curves of [6]-, [8]-, and
[10]-gingerols concentrations were y = 52.85x – 95.27, y = 46.02x -73.92, and y = 43.57x
-59.58, respectively, where x was the concentration of standard gingerol (µg·ml-1) and y was the total peak area. Although the gingerols are able to undergo dehydration reactions, leading to formation of the shogaols, this only occurs at elevated temperatures or if solutions of the gingerols are dried in air. We saw no evidence for the presence of the shogaols in the standard solutions used for this analysis (although we were able to detect, identify and quantify the shogaols in extracts from ginger rhizomes, see Table 2). Thus, we believe that are results are not only precise, but are accurate representations of the actual content of the gingerols in these samples.
The content of gingerols in Z. officinale samples, as summarized in Table 3, were determined by HPLC for 10 accessions that were obtained from different origins. The total gingerols content varied from 1.931 to 3.577 mg/g, with [6]-, [8]-, and [10]- gingerols ranging from 1.284 to 1.905, 0.220 to 0.595, and 0.310 to 1.128 mg/g, respectively (see Table 3). The observed differences were significant for most pair-wise comparisons, as determined by ANOVA analysis (P<0.05). Thus, although qualitatively not different by GC/MS-based metabolic fingerprinting analysis (no difference in
95
absence/presence of major volatile metabolites), more detailed metabolic profiles for
select compounds in these Z. officinale samples demonstrated that the chemical compositions of these accessions were in fact quantitatively different. Because these plants were grown under identical conditions at the same time in the same greenhouse, variation in metabolite composition could not be attributed to environmental factors.
Instead, it appears that genetic variation, which we were not able to measure by sequencing two conserved genes (trnL and rps16), must be responsible for the differential accumulation of these major pungent principles. Variation in expression or activity of genes/enzymes involved in the production of these metabolites can explain these differences. Furthermore, this result supports the contention that ginger from different sources is in fact likely to possess different properties, both in flavor and pungency, but also in potential bioactivity and health benefits. Identification of genes involved in regulating the production of these compounds is a major goal of future research in this area.
2.4. Anti-inflammatory assays
To address the issue of potential variation in bioactivity within Z. officinale, and to determine if there were significant differences in bioactivity between different species in the genus Zingiber, we measured the ability of MTBE and MeOH extracts of Zingiber samples to inhibit LPS- induced in vitro production of PGE2 and TNF-α in cultured
human promonocytic U937 cells (Ilieva et al., 2004). All extracts from the Zingiber
96
species examined were much more effective at inhibiting PGE2 production than at inhibiting TNF-α, thus only data for PGE2 are shown in Table 4.
No statistically significant difference was found between the ginger (Z. officinale) crude
MeOH and MTBE extracts in their ability to inhibit PGE2 production, as determined by
one-way ANOVA. The IC50 for inhibition of PGE2 production of ginger MeOH extracts
ranged from 0.058 µg·ml-1 to 0.629 µg·ml-1 and of MTBE extracts ranged from 0.062
-1 -1 2 µg·ml to 0.499 µg·ml . No correlation existed (R <0.001) in the PGE2 inhibitory
activity between the MeOH and the MTBE extracts from individual samples. For
example, the MeOH extracts from accessions L1 and L55 were among the most effective,
whereas the MTBE extracts from these accessions were among the least effective (see
Table 4) at inhibiting PGE2 production. In contrast, both the MeOH and MTBE extracts
from accessions L2 and L56 were among the most effective. Nevertheless, the IC50 for
inhibition of PGE2 production of the MeOH and MTBE extracts from most ginger
samples was roughly comparable to the IC50 for indomethacin in our assay system.
Furthermore, studies with purified gingerols, shogaols and paradols found IC50 values
between 0.290 and 7.35 µg·ml-1 for inhibition of inflammation response (Park et al.,
1998; Young et al., 2005). These data support the observation that ginger is an effective anti-inflammatory botanical (Henderson and Panush, 1999; Srivastava and Mustafa,
1992), but raise questions regarding the nature of the compounds responsible for this activity.
97
Previous investigations have concluded that the gingerols are the major anti-
inflammatory compounds in ginger (Park et al., 1998; Young et al., 2005). Indeed, our
results demonstrate that the compounds responsible for the anti-inflammatory activity
(measured in this assay system) were present in both the MTBE and MeOH extracts,
which were obtained in parallel from the same ground rhizome powders. Compounds
such as the gingerols, shogaols, and paradols fit this description, in that they are exactable
by both MTBE and MeOH. However, detailed metabolite target analysis of ginger
accessions obtained from different origins but grown under identical conditions
(described above) demonstrated that gingerol concentrations varied significantly between
ginger samples (see Tables 3 and 4). In fact, no correlation existed between gingerol
concentration (either total gingerol or for individual gingerols) and the level of PGE2 inhibitory activity for these ginger samples (R2 between 0.06 and 0.32 for all
comparisons). Thus, total bioactivity of crude MTBE and MeOH extracts appears be the
result of a combination of compounds, which appear to differ by accession. Some of these contributing compounds may well be the gingerols (or shogaols or paradols, which
are derived from the gingerols), but our results clearly demonstrate that other compounds
must also be involved in producing the observed anti-inflammatory activity. These could
be diarylheptanoids. Curcumin, the major diarylheptanoid of turmeric, has been show to
possess anti-inflammatory activity (Chainani-Wu, 2003). Ginger contains significant
levels of a wide range of diarylheptanoids (Kikuzaki et al., 1991; Ma et al., 2004).
Because these compounds are not amenable to GC/MS analysis, they were not included
in our analysis. In addition, many species of Alpinia have been shown to contain a wide
98
range of diarylheptanoids. These compounds may be responsible for the observed in vitro
anti-inflammatory activity.
One major difference between the ginger MeOH and MTBE extracts was that MeOH
extracts displayed much less cytotoxicity than did the MTBE extracts, as is obvious by
the fact that most of these extracts were not cytotoxic at the highest concentrations
measured. The compounds responsible for this cytotoxicity are likely to be rather
nonpolar and non-extractable with MeOH, ruling out the gingerols, shogaols, and
paradols, and most diarylheptanoids. The cytotoxic components in ginger MTBE extracts
have not been identified.
Z. zerumbet MeOH and MTBE extracts also displayed good PGE2 inhibition activity,
which was comparable to most ginger samples. In contrast, other Zingiber species, including Z. mioga, Z. montanum, and Z. spectabile, showed lower anti-inflammatory activity. Interestingly, Z. zerumbet was the closest species to Z. officinale in the phylogenetic tree based on DNA sequence. In contrast, A. galanga, the most distant species from Z. officinale in the phylogenetic tree, demonstrated the highest PGE2 inhibitory activity. On the other hand, the A. galanga extracts also showed strong cytotoxicity. The gingerols and related molecules are not present in A. galanga or Z. zerumbet. Thus, as discussed above, other compounds in these species must be responsible for the observed anti-inflammatory activity. These compounds have not yet been identified.
99
3. Conclusions
Using DNA sequence data, we were able to produce a phylogeny of the medicinal
Zingiber species. This DNA sequence-based phylogeny matched very closely the
phylogeny produced from chemical markers obtained from metabolic profiling
experiments using extracts from the rhizomes. Thus, chemotaxonomic investigations can
be very suitable for specific groups of plants and can uncover relationships that match
those determined by DNA molecular data. Furthermore, we were able to identify several
compounds that may serve as important markers for specific Zingiber species and which
may be used to identify adulterated ginger samples. However, the anti-inflammatory
activity in the analyzed species did not necessarily follow the phylogeny. Although the
closely related Z. zerumbet had similar anti-inflammatory activity when compared to Z.
officinale, so did A. galanga, which was the most distant from Z. officinale in the
phylogenetic trees. These observations, as well as the results from anti-inflammatory
assays with extracts from different ginger accessions, suggest that more than one set of
compounds that are not common to these plants are responsible for the anti-inflammatory activity. Lack of correlation between gingerol content and observed total anti-
inflammatory activity further confirms that other compounds contribute to this activity.
Some of these compounds may be diarylheptanoids, although other, as yet
uncharacterized compounds may also contribute to this activity. Whether these
compounds in ginger are acting alone or in synergy with the gingerols is still an open
question and can only be investigated after these compounds are separated, identified and
100 assayed. Furthermore, identification and quantitation of the actual bioactive compounds or at least measurement of the bioactivity in some assay system (such as we performed here) must be performed to ensure that a given botanical actually possesses the activity desired (Steinke et al., 1993; Gong et al., 2004), whatever that desired activity may be.
These results also demonstrate the importance of having access to a living collection of closely related medicinal plants to work with when performing a comparative study of this type. The results obtained here could not have been obtained by working with commercial extracts alone.
4. Experimental
4.1. Plant material Plant identification was performed by one of us (Steven P.
McLaughlin), based on comparison to the literature of morphological characters, including rhizome structure and color, leaf size, shape, and coloring, and especially on inflorescence and flower characters. Voucher specimens were deposited in the University of Arizona Herbarium. Individual rhizomes of Z. officinale, Z. mioga, Z. montanum, Z. spectabile and Z. zerumbet were obtained from different sources (see Table 1) and grown side by side in the same greenhouse at the University of Arizona through at least one whole year’s growth seasons, from germination to dormancy. After onset of dormancy, the rhizomes were harvested and replanted to eliminate any carryover affects of prior growth conditions (from original source) on the chemical composition of the rhizomes and to ensure that uniform growth conditions were applied equally to all plant specimens
101
that were evaluated in this investigation. The plants used for this analysis were grown in
5 gallon pots, in Scott’s Metromix soil, and watered by drip irrigation. Fresh young leaves were collected on the same day in the middle of October (one month prior to
dormancy onset) for DNA isolation and fresh rhizomes were collected on the same day in
the middle of May (during the middle of the growing season). The collected plant
material was immediately frozen in liquid nitrogen, and kept at -80 °C until analyzed.
4.2. Chemicals and Reagents Acetonitrile and methanol were from Burdick & Jackson
(Muskegon, MI). Methyl t-Butyl Ether (MTBE, High Purity Solvent) was purchased from
EMD Chemicals Inc (Gibbstown, NJ). Authentic standards of [6]-gingerol, [8]-gingerol, and [10]-gingerol were purchased from ChromaDex, Inc. (Santa Ana, CA).
4.3. Phylogenetic Analysis Genomic DNA was isolated from frozen young leaf tissue using the DNeasy plant mini kit (Qiagen, Valencia, CA) according to the manufacturer's protocol. The isolated DNA was quantified in water using a UV spectrophotometer
(Lambda 35 UV/vis Spectrometer, PerkinElmer Instruments, Shelton, CT), by measuring the absorbance at 260 and 280 nm and calculating the concentration using the method from the DNeasy Plant Mini Kit and DNeasy Plant Maxi Kit Handbook (Qiagen). The quality of the DNA samples were also examined using 1% agarose gel electrophoresis in the presence or absence of endonuclease (EcoR I) digestion. All of the DNA samples were diluted with double distilled water into working stock solutions (5 ng/µl), and stored at -80 °C before use.
102
Four separate DNA sequence regions were originally chosen for use in this study. These
included: 1) the nuclear internal transcribed spacer (ITS) locus, 2) the flanking intergenic
spacer regions of the plastid encoded genes matK-trnK, and two other noncoding
chloroplast regions, 3) the trnL-F region and 4) the intron of rps16. The ITS region of
~800 bp was amplified with primer pair ITS4 and ITS5 (Kress et al, 2002). The matK region of ~2500 bp was amplified with primers trnK1F (Manos and Steele, 1997) and
trnK2R (Steele and Vilgalys, 1994). The trnL-F region consists of the trnL intron and the
trnL-trnF intergenic spacer (Taberlet et al., 1991). The entire trnL-F region of ~900 bp
was amplified with primer pair tRNc and tRNf (Wallander and Albert, 2000). The intron
of rps16 of 800-900 bp was amplified with primer pair rpsF and rpsR2 (Wallander and
Albert, 2000). PCR amplification of the ITS region produced two bands with mixed
sequence in each, suggesting multiple copies of this gene region in ginger and related
species. The matK regions did not amplify consistently for all the plant samples from the
different species in this investigation. This made use of these two gene regions
impractical for our investigation, and they were consequently dropped from further
analysis, although they had been the regions used in the phylogeny reported by Kress et
al. (2002), which had not included any of the species that we examined.
PCR was preformed in a 25 µl reaction volume using the Taq kit from Fisher Scientific
(Fair Lawn, NJ) according to the manufacturer’s directions accept that 3mM MgCl2, 0.75
U of Taq, and 5 pmol of each primer were used. These amplifying reactions were all performed in a PTC-200 DNA Engine DYADTM Peltier Thermal Cycler (MJ Research,
103
Waltham, MA) using the same program (an initial step of 30 s at 94°C, followed by 30
cycles of 5 s at 94°C, 10 s at 56°C , 3.5 min at 72°C , then a final 7 min extension at
72°C ).
PCR products were incubated with 3 units of exonuclease I (USB Corporation, Cleveland,
Ohio) and 0.6 unit of shrimp alkaline phosphatase (USB) at 37 °C for 5 min followed by
72 °C for 15 min in order to remove unincorporated dNTPs and primers. Sequencing reactions were performed using 2ul of enzymatically cleaned PCR product, 3 pmol/µl sequencing primer (one of the primers used in PCR amplification) and BigDye
Terminator (v3.0, Applied Biosystems, Foster City, CA) in a 96 well PTC-200 thermal cycler with 35 cycles of 10 s at 96°C, 5 s at 50°C, and 4 min at 60°C. Unincorporated terminators were removed by ethanol precipitation followed by 70% ethanol wash.
Purified sequencing reactions were resuspended with 10ul of HIDI (Applied Biosystems) and loaded on a ABI3730xl automatic sequencer to separate and collect sequences.
The forward and reverse sequences were checked and edited using ChromasPro MFC
Application version 1.15 (Conor McCarthy, Griffith University, Australia). Based on the forward and reverse sequences, consensus sequences from each of the two chloroplast loci were aligned separately. Alignment of the joined sequences for each sample was performed using ClustalX version 1.83 (Thompson et al., 1997). Maximum parsimony analyses of the trnL-F region and the intron of rps16 sequence data were conducted using
PHYLIP 3.63 (Felsenstein, 2004).
104
The discrete character “0” was used to represent “not detectable” and “1” was used to
represent “detectable” for compounds listed in Table 2. The Seqboot, Pars, and Consense
programs in the PHYLIP 3.63 package (Felsenstein, 2004), were used to produce a
consensus phylogenetic tree from these discrete characters based on parsimony.
Bootstrap support for the resulting tree was also determined using these programs.
4.4. Metabolic profiling, fingerprinting and targeted metabolite analysis Frozen fresh
ginger rhizome was ground into a fine powder in the presence of liquid nitrogen with a
mortar and pestle and divided into 1 g aliquots for extraction with methyl t-butyl ether
(MTBE) or MeOH, transferred to a 4 ml glass vial, covered with 2 ml of solvent (MTBE
or MeOH) and capped with PTFE lined cap. After extraction overnight at room
temperature with shaking at 200 rpm, each sample was centrifuged in the same glass vial
in a SORVALL RC-5 Superspeed Refrigerated Centrifuge, GSA Rotor (Du Pont
instruments, Norwalk, CT, USA) at 1500 rpm for 25 min to pellet the ground plant
material and then filtered through an Acrodisc® CR 13 mm syringe filter with 0.20 µm
PTFE membrane. The MTBE filtrate was used for GC/MS analysis and for anti-
inflammatory activity assays. The MeOH filtrate was used for anti-inflammatory activity
assays. A similar extraction procedure produced MeOH extracts for gingerol quantitation
by HPLC, the only difference being that the samples were extracted by sonication for 30 min instead of overnight with shaking. After that, samples were centrifuged and filtered as described above. Triplicate extracts were used for quantitation analysis.
105
GC/MS data were recorded with a ThermoFinnigan Trace DSQ GC/MS (ThermoElectron,
San Jose, CA). The gas chromatograph was fitted with an Alltech ECONO-CAPTM-
ECTM-5 (30m × 0.25mm ID × 0.25µm) capillary column, with 5 m guard column.
Operating conditions: column oven temperature programmed at 40°C for 2 min, then to
100°C at 8°C/min, then to 280°C at 3°C/min, then to 300°C at 10°C/min and hold for 3.5 min; Injector/transfer line/ion source temperatures 220/250/200°C, respectively; electron voltage, 70 eV. UHP helium was used as the carrier gas at a flow rate of 1.2 ml/min.
Injection volume was 3 µl and split ratio was 10. Eluted compounds were identified using the NIST Mass Spectral library Version 2.0 (NIST/EPA/NIH, USA) and by referring to a publication from Jolad et al. (2004). Compound identifications indicated by the library search program as being >80% probable were viewed as being likely hits. Spectra for each eluting compound were then compared (by hand) to the standard spectrum for the best hit to determine if the molecular ion peaks and the fragmentation patterns did in fact match. If any discrepancies were observed, then the compound was designated as an unknown compound and included in Supplementary Table 1.
An Agilent HPLC system was used for gingerol quantitation. Detector: DAD; Column:
Luna C18 (2), 5 µm, 25cm×4.6 mm (Phenomenex); Guard column: Security Guard AJO-
4287 (Phenomenex); Mobil phase: nanopure water (A) and HPLC grade acetonitrile (B);
The gradient elution had the following profile: 0-8 min, 45-50% B; 8-17 min, 50-65% B;
17-32 min, 65-100% B; 32-38 min, 100% B; flow rate: 1ml/min; temperature 48°C ;
Injection volume: 20 µl; Detection: 210, 230, and 280 nm. Triplicate injections for each
106
of three replicate extracts were performed for each accession to ensure accuracy and
reproducibility in this analysis.
4.5. Anti-inflammatory activity assays. These assays were performed as previously
described (Jolad et al., 2004), with a different cell type. Human promonocytic U937 cells
were cultured in Iscove’s Modified Dulbecco’s Medium with 4 mM L-glutamine, 1.5 g/l
6 -1 sodium bicarbonate, and 20% fetal bovine serum at 37°C, 5% CO2. Cells (1 × 10 ·ml )
growing actively were distributed into 48 well plates (0.5 ml/well) and cultured with
phorbol myristate acetate (PMA, 10 nM) for 24 h at 37°C, 5% CO2 to differentiate the
cells. Cells were washed with culture media. Lipopolysaccharide (LPS, 1µg·ml-1) and
different concentrations of plant extract were then added to duplicate plates. Cells were
cultured for another 24 h. Supernatants were taken and stored at -80 °C before assays for
human TNF-α and PGE2.
Immunoassay kits were ordered from R & D System. ODs at 450 nm for TNF-α and 405
nm for PGE2 (570 nm reference) were measured using a Spectra max Plus plate reader
(Molecular Devices, Ramsey, MT). Data were analyzed using Molecular Devices plate reader software.
For cytotoxicity assays, human promonocytic U937 cells were cultured as described above. Cells (1 × 105 cells·ml-1) were distributed in 96 well plates (0.1 ml/well) and
cultured with PMA (10 nM) for 24 h. Cells were washed with culture media. LPS
107
(1µg·ml-1) and different concentrations of extract were added to duplicate plates. Cells
were cultured for another 24 h. Then, for MTT (C, N-diphenyl-N'-4,5-dimethyl thiazol-2-
yl tetrazolium bromide) assays, 20 µl of MTT (5 mg·ml-1) were added to each well. After the plates were then cultured for another 4 h, supernatants were aspirated and 100 µl of
isopropanol-HCl (0.04% HCl) were added to each well. The plates were protected from
the light at RT overnight. The OD was measured at 570 nm (660 nm reference). For
modified tetrazolium salt XTT (sodium 3'-[1-[(phenylamino)-carbonyl]-3,4-tetrazolium]- bis(4-methoxy-6- nitro)benzene-sulfonic acid hydrate) assay, 25 µl of XTT (1 mg·ml-1 with phenazine methosulfate) were added to each well and the plates were cultured for another 4 h in the dark. O.D. was measured at 450 nm (650 nm as reference wavelength).
Acknowledgements
The authors acknowledge financial assistance from the National Science Foundation
Plant Genome Program, grant DBI-0227618 to D.R.G., and the National Institutes of
Health NCCAM/ODS, grants #5 P50 AT 000474-04 and 3 P50 AT 000474-03 S1 to
B.N.T. We also thank Veronica Rodriguez and Brenda Jackson for assistance with chemical analysis, Dr. Guan Jie Chen for assistance with anti-inflammatory assays, and
Betsy Lewis, Brenda Jackson, and Dr. Maria del Carmen Ramirez for their assistance in
collecting some of the plant samples used in this investigation and help with GC/MS. The
contents of this publication are solely the responsibility of the authors and do not
necessarily represent the official views of NCCAM, ODS, or the National Institutes of
Health.
108
Fig. 1. A: An unrooted phylogenetic tree of samples from genus Zingiber based on trnL and rps16 sequences; B: An unrooted phylogenetic tree of samples from genus Zingiber based on chemical profiles identified using GC/MS. L8 from genus Alpinia is used as an outgroup. Numbers above the lines are bootstrap values in the phylogenetic tree.
109
Fig.2. Some sequence differences of the trnL intron, the trnL-trnF intergenic spacer and the intron of rps16 from Zingiber species and Alpinia galanga. L*: represents L1, L2, L5,
L6, L18, L32, L45, L54, L55, and L56
110
Table 1. The origin of Zingiber and Alpinia samples used in this analysis
Accession Species Original Source Voucher L1 Zingiber officinale Rosc. Alden Botanica, Moreno Valley, CA S. P. McLaughlin 9501
Flowering plant confirmed in L2 Zingiber officinale Rosc. ABCO, Tucson, AZ greenhouse, not vouchered L5 Zingiber officinale Rosc. 17th Street Farmers Market, Tucson, AZ B. Lewis 217
Flowering plant confirmed in L6 Zingiber officinale Rosc. Tucson Cooperative Warehouse, Tucson, AZ greenhouse, not vouchered L18 Zingiber officinale Rosc. Stokes Tropicals, New Iberia, LA Plants died prior to flowering L32 Zingiber officinale Rosc. Pacific Botanicals, Grants Pass, OR B. Lewis 218 L45 Zingiber officinale Rosc. Plantation Gardens, Clermont, FL Plants have not flowered L54 Zingiber officinale Rosc. Super K-Mart, Tucson, AZ Plants have not flowered L55 Zingiber officinale Rosc. Trader Joe's, Tucson, AZ Plants have not flowered L56 Zingiber officinale Rosc. Trader Joe's, Tucson, AZ B. Lewis 216 L15 Zingiber zerumbet Smith Fairchild Tropical Garden, Miami, FL S. P. McLaughlin 9769
Zingiber montanum (Koenig) L37 Theilade Gingerwood Nursery, St. Gabriel, LA B. Lewis 208
Zingiber mioga (Thunberg) L46 Roscoe Plantation Gardens, Clermont, FL S. P. McLaughlin 9977 L31 Zingiber spectabile Griff. Stokes Tropicals, New Iberia, LA Plants have not flowered L8 Alpinia galanga (L.) Sw. Fairchild Tropical Garden, Miami, FL B. Lewis 256
Table 2. Compounds detected and/or identified from Zingiber and Alpinia samples using GC/MS.
RT Compound Name M.W. L1 L2 L5 L6 L18 L32 L45 L54 L55 L56 L15 L37 L46 L31 L8 1,7,7-Trimethyl- 6.95 136 1 1 1 1 1 1 1 1 1 1 1 0 0 0 0 tricyclo[2.2.1.0(2,6)]heptane 7.06 3-Thujene 136 ○ 0 0 0 0 0 0 0 0 0 0 0 2 0 1 0 7.20 1R-α-Pinene 136 2 2 2 2 2 2 2 2 2 2 2 2 3 2 2 7.42 R(-)3,7-Dimethyl-1,6-octadiene 138 ● 1 1 1 1 1 1 1 1 1 1 0 0 0 0 0 7.51 Camphene 136 3 3 3 2 2 2 3 3 3 3 3 1 1 0 0 8.00 4(10)-Thujene 136 1 1 1 1 1 1 1 1 1 1 0 3 2 3 1 8.06 2(10)-Pinene 136 1 1 1 1 1 1 1 1 1 1 1 1 3 3 2 8.35 α-Pinene 136 2 2 2 2 2 2 2 2 2 2 2 2 1 1 0 2-Methyl-5-(1-methylethyl)-1,3- 8.62 136 1 1 1 1 1 2 1 1 1 2 1 1 1 0 0 cyclohexadiene 8.74 3-Carene 136 ○ 0 0 0 0 0 0 0 0 0 0 2 0 0 0 0 DRG-GM1-N1-8.86-136-93- 8.86 136 ○ 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 121 1-Methyl-4-(1-methylethyl)-1,3- 8.88 136 ○ 0 0 0 0 0 0 0 0 0 0 0 2 0 0 0 cyclohexadiene 9.03 Cymene 134 1 1 1 1 1 1 1 1 1 1 1 1 1 0 0 DRG-GM1-N1-9.03-134-119- 9.03 134 ○ 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 105 9.11 Limonene 136 ○ 0 0 0 0 0 0 0 0 0 0 2 0 0 0 0 9.11 DRG-GM1-N1-9.11-136-93-68 136 ○ 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 9.12 m-Mentha-6,8-diene 136 ○ 0 0 0 0 0 0 0 0 0 0 0 2 0 0 0 3-Methylene-6-(1-methylethyl)- 9.14 136 3 3 3 3 3 3 3 3 3 3 0 0 3 0 0 cyclohexene 9.16 Cineole 154 1 1 1 1 1 1 1 2 1 3 2 1 0 0 3 1-Methyl-4-(1-methylethyl)-1,4- 9.75 136 ○ 0 0 0 0 0 0 0 0 0 0 0 2 0 2 0 cyclohexadiene 9.91 Terpineol 154 ○ 0 0 0 0 0 0 0 0 0 0 0 2 0 0 0 10.34 p-Mentha-1,4(8)-diene 136 1 1 1 1 1 1 1 0 1 1 0 2 0 0 0 10.58 α-Linalool 154 1 2 2 1 2 1 1 2 2 1 1 1 0 0 0 trans-4-Isopropyl-1-methyl-2- 11.11 154 ○ 0 0 0 0 0 0 0 0 0 0 0 1 0 0 111 cyclohexen-1-ol 0
cis-4-Isopropyl-1-methyl-2- 11.57 154 ○ 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 cyclohexen-1-ol 11.66 (-)-Alcanfor 152 1 1 1 1 1 1 1 1 1 1 2 0 0 0 0 11.83 Citronellal 154 ● 1 1 1 1 1 1 1 1 1 1 0 0 0 0 0 12.22 (1S,2R,4S)-(-)-Borneol 154 ● 2 2 1 1 1 2 2 1 1 3 0 0 0 0 0 12.62 p-Menth-1-en-4-ol 154 ○ 0 0 0 0 0 0 0 0 0 0 0 3 0 1 1 12.85 p-Menth-1-en-8-ol 154 1 2 1 2 2 1 2 2 2 2 0 1 0 0 0 14.24 (Z)-α-Citral 152 ● 2 2 2 2 2 2 2 2 2 2 0 0 0 0 0 15.16 (E)-α-Citral 152 ● 2 3 2 3 3 3 2 3 3 3 0 0 0 0 0 15.82 2-Undecanone 170 ● 1 1 1 1 1 1 1 1 1 1 0 0 0 0 0 17.53 4-Allylphenyl acetate 176 ○ 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 17.64 p-Menth-1-en-8-ol, acetate 196 ○ 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 18.21 (+)-Cyclosativene 204 ● 1 1 1 1 1 1 1 1 1 1 0 0 0 0 0 18.56 α-Cubebene or Copaene 204 ● 2 2 2 2 2 2 1 2 2 3 0 0 0 0 0 19.56 Zingiberene 204 ● 1 1 1 1 1 1 1 1 1 1 0 0 0 0 0 20.10 Caryophyllene 204 ○ 0 0 0 0 0 0 0 0 0 0 2 0 0 0 1 21.22 α-Caryophyllene 204 ○ 0 0 0 0 0 0 0 0 0 0 3 0 0 0 1
α-Farnesene (1,6,10- 21.34 Dodecatriene, 7,11-dimethyl-3- 204 2 1 1 1 1 2 2 1 1 2 0 0 0 0 2 methylene-)
22.24 Germacrene D 204 ● 2 2 1 2 1 2 1 1 1 2 0 0 0 0 0 22.32 2-Methyl-6-p-tolyl-2-heptene 202 ● 2 3 3 2 3 2 3 3 3 3 0 0 0 0 0 22.65 Eudesma-4(14),11-diene 204 1 2 2 2 2 1 2 2 2 1 0 1 0 0 1 22.80 Cedr-8-ene 204 ● 3 3 3 3 3 3 3 3 3 3 0 0 0 0 0 22.82 Cadinene 204 ● 0 1 1 0 0 0 0 2 2 0 0 0 0 0 0
1-Methyl-4-(5-methyl-1- 23.20 methylene-4-hexenenyl)- 204 ○ 0 0 0 0 0 0 0 0 0 0 0 0 0 0 2 cyclohexene,
α-Farnesene (1,3,6,10- 23.30 Dodecatetraene, 3,7,11- 204 ● 3 3 3 3 3 3 3 3 3 3 0 0 0 0 0 trimethyl-)
(+)-Epi- 23.36 204 ● 1 1 1 1 1 1 1 1 1 1 0 0 0 0 bicyclosesquiphellandrene 0 0 23.53 Panasinsen 204 ● 1 1 1 1 1 1 1 1 1 1 0 0 0 0 112 23.71 Cadina-1(10),4-dien 204 ○ 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0
23.86 α-Sesquiphellandrene 204 3 3 3 3 3 3 3 3 3 3 0 2 0 0 2 25.42 1,4-Bis(methoxy)-triquinacene 204 ○ 0 0 0 0 0 0 0 0 0 0 0 2 0 0 0 25.82 Caryophyllene oxide 220 ○ 0 0 0 0 0 0 0 0 0 0 2 0 0 0 0 28.21 Eudesm-4(14)-en-11-ol 222 ○ 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 2,6,9,9-Tetramethyl-2,6,10- 31.49 220 ○ 0 0 0 0 0 0 0 0 0 0 3 0 0 0 0 cycloundecatrien-1-one 46.69 [6]-Paradol 278 ● 1 1 1 1 1 1 1 1 1 1 0 0 0 0 0 48.37 [6]-Shogaol 276 ● 2 2 2 2 2 2 2 2 2 2 0 0 0 0 0 49.37 [7]-Paradol 292 ● 2 2 2 1 2 2 2 2 2 2 0 0 0 0 0 51.11 [6]-Gingerol 294 ● 3 3 2 2 2 2 2 2 3 3 0 0 0 0 0 52.85 Acetoxy-[6]-gingerol 336 ● 1 1 1 1 1 1 1 1 1 1 0 0 0 0 0 53.95 [8]-Shogaol 304 ● 1 1 1 1 1 1 1 1 1 1 0 0 0 0 0 54.48 Diacetoxy-[6]-gingerdiol 380 ● 1 1 1 1 2 1 1 2 2 1 0 0 0 0 0 54.84 [9]-Paradol 320 ● 1 1 1 1 1 2 1 1 1 1 0 0 0 0 0 56.56 [8]-Gingerol 322 ● 2 2 2 1 1 2 1 1 2 2 0 0 0 0 0 59.08 [10]-Shogaol 332 ● 2 2 2 2 1 2 2 2 2 2 0 0 0 0 0
60.03 [11]-Paradol 348 ● 2 2 2 2 2 2 2 2 2 2 0 0 0 0 0
Note: 0 indicates absence, 1 indicates < 0.5%, 2 indicates 0.5% – 5%, and 3 indicates > 5% of total integrated peak area of TIC of a particular sample; ● indicates compounds detected only in ginger (Z. officinale) samples; ○ indicates compounds not detected in ginger.
113
Table 3. Content (mg/g) of the major gingerols in ginger samples from different lines, as determined by HPLC
Ginger Line Compound L1 L2 L5 L6 L18 L32 L45 L54 L55 L56 [6]-gingerol 1.669±0.021 1.407±0.013 1.457±0.014 1.442±0.013 1.454±0.020 1.905±0.015 1.508±0.021 1.284±0.022 1.388±0.001 1.854±0.015 [8]-gingerol 0.397±0.004 0.220±0.002 0.526±0.004 0.352±0.004 0.410±0.010 0.580±0.004 0.462±0.008 0.370±0.005 0.392±0.001 0.595±0.004 [10]-gingerol 0.588±0.004 0.310±0.003 1.021±0.005 0.653±0.008 0.738±0.018 0.804±0.006 0.805±0.012 0.549±0.005 0.658±0.011 1.128±0.008 Total 2.654±0.029 1.931±0.017 3.004±0.021 2.447±0.024 2.602±0.047 3.289±0.025 2.775±0.041 2.023±0.033 2.438±0.011 3.577±0.025
Note: all values are the average of 3 replicates ± standard error 114
Table 4. Anti-inflammatory activities and cytotoxicity of Zingiber and Alpinia MeOH and MTBE extracts
MeOH extract MTBE extract
IC50(PGE2) Cytotoxic Dose IC50(PGE2) Cytotoxic -1 -1 -1 -1 Accession Species (µg·ml ) (µg·ml ) IC50(PGE2)/Cytotoxicity (µg·ml ) Dose (µg·ml ) IC50(PGE2)/Cytotoxicity L1 Z. officinale 0.058 >50 <0.001 0.335 5 0.067 L2 Z. officinale 0.059 >50 <0.001 0.072 0.1 0.72 L5 Z. officinale 0.167 >50 <0.003 0.31 0.1 3.1 L6 Z. officinale 0.629 50 0.013 0.483 10-50 0.01-0.048 L18 Z. officinale 0.074 >50 <0.002 0.077 5-10 0.008-0.015 L32 Z. officinale 0.146 >50 0.003 0.098 5-10 0.01-0.020 L45 Z. officinale 0.069 50 0.001 0.163 0.1 1.63 L54 Z. officinale 0.073 50 0.002 0.079 1 0.079 L55 Z. officinale 0.058 10 0.006 0.499 1-5 0.1-0.499 L56 Z. officinale 0.065 1-5 0.013-0.065 0.062 1 0.062 L15 Z. zerumbet 0.079 5-10 0.008-0.016 0.096 0.1 0.96 L37 Z. montanum 7.678 >50 0.154 33.822 50 0.676 L46 Z. mioga * >50 * * 10-50 * L31 Z. spectabile 1.171 1 1.171 0.142 1-5 0.028-0.142 L8 A. galanga 0.055 1-5 0.011-0.055 0.053 0.1 0.53 Ref. Compd. Indomethacin 0.055 >5 <0.011 Note: * indicates no inhibitory activity.
. 115
116
References
Ando, S., Matsuda, H., Morikawa, T., Yoshikawa, M., 2005. 1´S-1´-Acetoxychavicol
acetate as a new type inhibitor of interferon-β production in lipopolysaccharide-
activated mouse peritoneal macrophages. Bioorgan. Med. Chem. 13, 3289-3294.
Bhattarai, S., Tran, V.H., Duke, C.C., 2001. The stability of gingerol and shogaol in
aqueous solutions. J. Pharm. Sci. 90 (10), 1658-1664.
Bino, R.J., Hall, R.D., Fiehn, O., Kopka, J., Saito, K., Draper, J., Nikolau, B.J., Mendes,
P., Roessner-Tunali, U., Beale, M.H., Trethewey, R.N., Lange, B.M., Wurtele,
E.S., Sumner, L.W., 2004. Potential of metabolomics as a functional genomics
tool. Trends Plant Sci. 9, 418-425.
Chainani-Wu, N., 2003. Safety and anti-inflammatory activity of curcumin: a component
of turmeric (Curcuma longa). J. Altern. Complem. Med. 9, 161-168.
Connell, D.W., Sutherland, M.D., 1969. A re-examination of gingerol, shogaol, and
zingerone, the pungent principles of ginger (Zingiber officinale Roscoe). Aust. J.
Chem. 22, 1033-1043.
Dedov, V.N., Tran, V.H., Duke, C.C., Connor, M., Christie, M.J., Mandadi, S.,
Roufogalis, B.D., 2002. Gingerols: a novel class of vanilloid receptor (VR1)
agonists. Br. J. Pharmacol. 137 (6), 793-798.
Felsenstein, J., 2004. PHYLIP (Phylogeny Inference Package) version 3.63. Distributed
by the author. Department of Genome Sciences, University of Washington, Seattle,
USA
117
Gong, F., Fung, Y.S., Liang, Y.Z., 2004. Determination of volatile components in ginger
using gas chromatography-mass spectrometry with resolution improved by data
processing techniques. J. Agric. Food Chem. 52, 6378-6383.
Govindarajan, V.S., 1982. Ginger-chemistry, technology and quality evaluation: Part I.
Crit. Rev. Food Sci. Nutr. 17, 1-96.
Ilieva, I., Ohgami, K., Shiratori, K., Koyama, Y., Yoshida, K., Kase, S., Kitamei, H.,
Takemoto, Y., Yazawa, K., Ohno, S., 2004. The effects of Ginkgo biloba extract
on lipolysaccharide-induced inflammation in vitro and in vivo. Exp. Eye Res. 79,
181-187.
Herderson, C.J., Panush, R.S., 1999. Diets, dietary supplements, and nutritional therapies
in rheumatic diseases. Rheum. Dis. Clinic. N. Am. 25, 937-968.
Jolad, S.D., Lantz, R.C., Solyom, A.M., Chen, G.J., Bates, R.B., Timmermann, B.N.,
2004. Fresh organically grown ginger (Zingiber officinale): composition and
effects on LPS-induced PGE2 production. Phytochemistry 65 (13), 1937-1954.
Kikuzaki, H., Kobayashi, M., Nakatani, N., 1991. Diarylheptanoids from rhizomes of
Zingiber officinale. Phytochemistry 30 (11), 3647-3651.
Kress, W.J., Prince, L.M., Williams, K.J., 2002. The Phylogeny and a new classification
of the gingers (Zingiberaceae): evidence from molecular data. Am. J. Bot. 89
(11), 1682-1696.
Langner, E., Greifenberg, S., Gruenwald, J., 1998. Ginger: history and use. Adv. Ther. 15
(1), 25-44.
118
Ma, J., Jin, X., Yang, L., Liu, Z., 2004. Diarylheptanoids from the rhizomes of Zingiber
officinale. Phytochemistry. 65 (8), 1137-1143.
Manos, P.S., Steele, P.S., 1997. Phylogenetic analyses of “higher” Hamamelididae based
on platid sequence data. Am. J. Bot. 84, 1407-1419.
McDonald, T., Miadlikowska, J., Lutzoni, F., 2003. The lichen genus Sticta in the Great
Smoky Mountains: a phylogenetic study of morphological, chemical, and
molecular data. Bryologist 106, 61-79.
Miadlikowska, J., Lutzoni, F., 2000. Phylogenetic Revision of the genus Peltigera
(lichen-forming Ascomycota) based on morphological, chemical, and large
subunit nuclear ribosomal DNA data. Int. J. Plant Sci. 161, 925-958.
Miyoshi, N., Nakamura, Y., Ueda, Y., Abe, M., Ozawa, Y., Uchida, K., Osawa, T., 2003.
Dietary ginger constituents, galanals A and B, are potent apoptosis inducers in
Human T lymphoma Jurkat cells. Cancer Lett. 199, 113-119.
Murakami, A., Miyamoto, M., Ohigashi, H., 2004. Zerumbone, an anti-inflammatory
phytochemical, induces expression of proinflammatory cytokine genes in human
colon adenocarcinoma cell lines. Biofactors. 21, 95-101.
Mustafa, T., Srivastava, K.C., Jensen, K.B., 1993. Drug development report (9):
pharmacology of ginger, Zingiber officinale. J. Drug Dev. 6, 25-39.
Nakamura, Y., Yoshida, C., Murakami, A., Ohigashi, H., Osawa, T., Uchida, K., 2004.
Zerumbone, a tropical ginger sesquiterpenes, activates phase II drug metabolizing
enzymes. FEBS Lett. 572, 245-250.
119
Park, K.K., Chun, K.S., Lee, J.M., Lee, S.S., Surh, Y.J., 1998. Inhibitory effects of [6]-
gingerol, a major pungent principle of ginger, on phorbol ester-induced
inflammation, epidermal ornithine decarboxylase activity and skin tumor
promotion in ICR mice. Cancer Lett. 129, 139-144.
Srivastava, K.C., Mustafa, T., 1992. Ginger (Zingiber officinale) in rheumatism and
musculoskeletal disorders. Medical Hypoth. 39, 342-348.
Steele, K.P., Vilgalys, R., 1994. Phylogenetic analyses of Polemoniaceae using
nucleotide sequences of the plastid gene matK. Syst. Bot. 19, 126-142.
Steinke, B. Muller, B., Wagner, H., 1993. Biological standardization of Ginkgo extracts.
Planta Med. 59, 155-160.
Taberlet, P., Gielly, L., Pantou, G., Bouvet, J., 1991. Universal primers for amplification
of three non-coding regions of chloroplast DNA. Plant Mol. Biol. 17 (5), 1105-
1109.
Thompson, J.D., Gibson, T.J., Plewniak, F., Jeanmougin, F. and Higgins, D.G., 1997.
The ClustalX windows interface: flexible strategies for multiple sequence
alignment aided by quality analysis tools. Nucl. Acids Res. 24, 4876-4882.
Wallander, E., Albert, V.A., 2000. Phylogeny and classification of Oleaceae based on
rps16 and trnL-F sequence data. Am. J. Bot. 87 (12), 1827-41.
Yang, X., Eilerman, R.G., 1999. Pungent principal of Alpinia galanga (L.) Swartz and its
applications. J. Agric. Food Chem. 47, 1657-1662.
120
Young, H.Y., Luo, Y.L., Cheng, H.Y., Hsieh, W.C., Liao, J.C., Peng, W.H., 2005.
Analgesic and anti-inflammatory activities of [6]-gingerol. J Ethnopharmacol. 96,
207-10.
Supplementary Table 1. Unknown compounds detected from Zingiber and Alpinia samples using GC/MS.
RT Compound Name M.W. L1 L2 L5 L6 L18 L32 L45 L54 L55 L56 L15 L37 L46 L31 L8 8.86 DRG-GM1-N1-8.86-136-105-120 136 ○ 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 9.70 DRG-GM1-N1-9.70-154-93-136 154 ○ 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 11.76 DRG-GM1-N1-11.76-152-95-71 152 ○ 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 12.49 DRG-GM1-N1-12.49-154-9571 154 ● 1 1 1 1 1 1 1 1 1 1 0 0 0 0 0 12.85 DRG-GM1-N1-12.85-154-59-79 154 ○ 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 14.20 DRG-GM1-N1-14.20-136-121-93 136 ○ 0 0 0 0 0 0 0 0 0 0 0 2 0 0 0 15.63 DRG-GM1-N1-15.63-154-95-136 154 ○ 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 16.04 DRG-GM1-N1-16.04-150-93-136 150 ○ 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 16.26 DRG-GM1-N1-16.26-150-97-69 150 ○ 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 17.44 DRG-GM1-N1-17.44-212-109-137 212 ○ 0 0 0 0 0 0 0 0 0 0 0 0 0 0 3 17.91 DRG-GM1-N1-17.91-212-109-197 212 ○ 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 18.09 DRG-GM1-N1-18.09-212-109-126 212 ○ 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 19.11 DRG-GM1-N1-19.11-204-93-67 204 1 1 1 1 2 2 2 1 1 2 1 0 2 0 0 19.38 DRG-GM1-N1-19.38-170-139-95 170 ○ 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 19.72 DRG-GM1-N1-19.72-204-105-189 204 ○ 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 20.60 DRG-GM1-N1-20.6-204-119-93 204 ○ 0 0 0 0 0 0 0 0 0 0 0 0 0 0 2 21.49 DRG-GM1-N1-21.49-204-91-105 204 ● 1 1 1 1 1 1 1 1 1 1 0 0 0 0 0 22.00 DRG-GM1-N1-22.00-204-189-133 204 ● 1 1 1 1 1 1 1 1 1 1 0 0 0 0 0 22.20 DRG-GM1-N1-22.20-204-161-91 204 ○ 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 22.75 DRG-GM1-N1-22.75-204-121-93 204 ○ 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 22.95 DRG-GM1-N1-22.95-204-93-121 204 ○ 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 24.00 DRG-GM1-N1-24.00-204-107-93 204 1 1 1 1 1 1 1 1 1 1 0 0 0 0 2 24.60 DRG-GM1-N1-24.60-178-131-103 178 ○ 0 0 0 0 0 0 0 0 0 0 0 2 0 0 0 24.89 DRG-GM1-N1-24.89-204-121-93 204 2 2 2 2 2 2 2 2 2 2 0 1 0 0 0 25.20 DRG-GM1-N1-25.20-204-69-93 204 ● 1 1 1 1 1 1 1 1 1 1 0 0 0 0 0 25.97 DRG-GM1-N1-25.97-176-115-144 176 ○ 0 0 0 0 0 0 0 0 0 0 0 2 0 0 0 26.28 DRG-GM1-N1-26.28-192-177-161 192 ○ 0 0 0 0 0 0 0 0 0 0 0 2 0 0 0 26.42 DRG-GM1-N1-26.42-220-93-121 220 ○ 0 0 0 0 0 0 0 0 0 0 2 0 0 0 0 26.78 DRG-GM1-N1-26.78-220-67-109 220 ○ 0 0 0 0 0 0 0 0 0 0 2 0 0 0 0 0 27.44 DRG-GM1-N1-27.44-244-81-67 244 ○ 0 0 0 0 0 0 0 0 0 0 1 0 0 0 121 28.30 DRG-GM1-N1-28.30-192-132-150 192 ○ 0 0 0 0 0 0 0 0 0 0 0 0 0 0 3
28.63 DRG-GM1-N1-28.63-220-67-81 220 ○ 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 28.74 DRG-GM1-N1-28.74-236-123-165 236 ○ 0 0 0 0 0 0 0 0 0 0 0 0 0 0 2 31.32 DRG-GM1-N1-31.32-220-189-174 220 ○ 0 0 0 0 0 0 0 0 0 0 0 2 0 0 0 32.36 DRG-GM1-N1-32.36-222-207-176 222 ○ 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 33.47 DRG-GM1-N1-33.47-264-180-162 264 ○ 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 34.17 DRG-GM1-N1-34.17-220-189-174 220 ○ 0 0 0 0 0 0 0 0 0 0 0 2 0 0 0 35.00 DRG-GM1-N1-35.00-234-192-149 234 ○ 0 0 0 0 0 0 0 0 0 0 0 0 0 0 2 36.07 DRG-GM1-N1-36.07-208-177-146 208 ○ 0 0 0 0 0 0 0 0 0 0 0 2 0 0 0 39.88 DRG-GM1-N1-39.88-250-159-190 250 ○ 0 0 0 0 0 0 0 0 0 0 0 3 0 0 0 50.02 DRG-GM1-N1-50.02-302-137-95 302 1 1 1 1 1 1 1 1 1 1 1 0 2 3 1 54.15 DRG-GM1-N1-54.15-318-109-232 318 ○ 0 0 0 0 0 0 0 0 0 0 0 0 2 0 0 54.33 DRG-GM1-N1-54.33-300-109-132 300 ○ 0 0 0 0 0 0 0 0 0 0 0 0 2 0 0 54.69 DRG-GM1-N1-54.69-318-95-67 318 ○ 0 0 0 0 0 0 0 0 0 0 0 0 2 0 0 55.17 DRG-GM1-N1-55.2-270-168-137 270 2 2 2 2 2 2 2 2 2 2 1 1 1 1 1 65.84 DRG-GM1-N1-65.84-380-190-159 380 ○ 0 0 0 0 0 0 0 0 0 0 0 2 0 0 0 66.65 DRG-GM1-N1-66.65-380-190-159 380 ○ 0 0 0 0 0 0 0 0 0 0 0 2 0 0 0 67.52 DRG-GM1-N1-67.52-410-220-189 410 ○ 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 68.5 DRG-GM1-N1-68.50-410-220-189 410 ○ 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 68.82 DRG-GM1-N1-68.82-410-220-189 410 ○ 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 69.56 DRG-GM1-N1-69.56-410-220-189 410 ○ 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 70.21 DRG-GM1-N1-70.21-440-220-189 440 ○ 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 70.98 DRG-GM1-N1-70.98-440-220-189 440 ○ 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 71.71 DRG-GM1-N1-71.71-414-207-281 414 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
Note: 0 indicates absence, 1 indicates < 0.5%, 2 indicates 0.5% – 5%, and 3 indicates > 5% of total integrated peak area of TIC of a particular sample; ● indicates compounds detected only in ginger (Z. officinale) samples; ○ indicates compounds not detected in ginger.
122
123
APPENDIX B
Manuscript “Instrument dependence of ESI ionization and MS/MS fragmentation of the
gingerols: A cautionary tale for metabolomics investigations”, submitted to the Journal of
the American Society for Mass Spectrometry.
124
Title: Instrument Dependence of ESI Ionization and MS/MS
Fragmentation of the Gingerols: A Cautionary Tale for Metabolomics
Investigations
Hongliang Jiang1–4, Arpad Somogyi5, Barbara N. Timmermann1,6, and David R. Gang1–3*
1Arizona Center for Phytomedicine Research, 2Department of Plant Sciences, 3Bio5
Institute, 4Department of Pharmaceutical Science, 5Department of Chemistry,
6Department of Pharmacology and Toxicology, and University of Arizona, Tucson,
Arizona, 85721
Running title: Instrument Dependence of ESI Ionization and MS/MS Fragmentation of the Gingerols
Address reprint requests to Dr. David R. Gang
Department of Plant Sciences and Bio5 Institute, University of Arizona, Tucson, AZ
85721-0036, USA
Tel: 520-621-7154
Fax: 520-621-7186 email: [email protected]
125
ABSTRACT
The gingerols including [6]-, [8]-, and [10]-gingerols, a series of chemical homologs
differentiated by the length of their unbranched alkyl chains, have been identified as
major active components in fresh ginger rhizome. The purpose of this study was to
investigate the utility of ion trap LC-MS/MS as an online tool to identify and quantify
these compounds in raw or processed ginger rhizome samples. Negative mode
electrospray ionization was used in MS, MS/MS and MSn experiments in quadrupole ion
trap instruments from two different manufacturers and in high resolution and accurate
mass MS and MS/MS experiments in a Fourier transform ion cyclotron resonance (FT-
ICR) mass spectrometer to elucidate the ionization and fragmentation mechanisms of
these compounds in these instruments. Positive mode electrospray ionization, which
generated many more fragment ions in full scan MS even under gentle ionization
conditions, was also used in LC-MS and MS/MS experiments and in direct infusion MS
and MS/MS experiments. Consistent and predictable ionization and fragmentation
behaviors were observed for all gingerols when analyzed in the same instrument.
Instruments from different manufacturers, however, had different ionization mechanisms and produced very different fragmentation patterns. The major difference between instruments was their ability to form covalent dimer adducts of the gingerols. These results clearly demonstrate that LC-MS instruments produce data that cannot necessarily be replicated in other laboratories, especially if those labs do not have the same
instrument model from the same manufacture. This presents major problems for
metabolite target analysis, metabolic profiling and metabolomics investigations, which
126 would benefit from LC-MS mass spectrum libraries as they do from GC/MS mass spectrum libraries, because such libraries may not be valid across platforms.
Introduction
The rhizomes of ginger (Zingiber officinale, Rosc.) have long been valued as a spice and as an herbal medicine in China, Japan, India, and other countries [1, 2]. Major active components of ginger are the gingerols, including [6]-, [8]-, and [10]-gingerols, a series of chemical homologs differentiated by the length of their unbranched alkyl chains. [6]-
Gingerol (5-hydroxy-1-[4′-hydroxy-3′-methoxyphenyl] decan-3-one) is the most abundant of these in normal ginger samples [3]. [6]-Gingerol has been found to possess various pharmacological and physiological effects including anti-inflammatory, analgesic, antipyretic, gastroprotective, cardiotonic, antihepatotoxic, antifungal, antioxidant, anti- ulcer, anti-tumor promoting, anti-platelet aggregation, molluscicidal, and anti- schistosomal activities [4-13]. Due to these properties of the gingerols, ginger has gained considerable attention as a botanical dietary supplement in the USA and Europe in recent years, and especially for its use in the treatment of chronic inflammatory conditions.
Analysis of the gingerols in ginger samples is thus crucial to determine the quality of the raw material or its processed products. Such analysis is also required for metabolomics- based investigations that seek to elucidate the biochemical pathways involved in gingerol formation.
Although gas-chromatography coupled with mass spectrometry (GC/MS) has been used quite often to analyze ginger samples, it has been mostly used to determine the
127
composition of essential oil [1, 14]. Due to their chemically labile properties, gingerols
can usually only be determined by analysis of their degradation products in GC/MS [15].
Gingerols could also be identified by modification of ginger extracts and/or partially
purified fractions to their trimethylsilyl (TMS) derivatives with a view to improve their volatility, stability, and separation in the GC/MS [16-18]. In addition, [6]-gingerol has been determined in underivatized oleoresin of ginger rhizome using GC/MS [19].
Furthermore, [4]-, [6]-, [7]-, and [8]-gingerols were also determined by GC/MS in underivatized fractions from ginger rhizome crude extract. In these analyses of underivatized ginger rhizome fractions using GC/MS, gingerols with longer side chains, such as [10]- and [12]-gingerol, could not be detected due to their thermal degradation
with longer exposure to high temperature in the GC. In addition, GC/MS has not been applied in the analysis of gingerols from crude ginger rhizome extracts. Other methods to analyze the gingerols include thin-layer chromatography (TLC), high performance liquid chromatography (HPLC), and cyclodextrin-modified micellar electrokinetic chromatography (MEKC) [20-22]. However, without comparison of the retention times with authentic standards, which are hard to come by because most gingerols are not commercially available and are difficult to synthesize or isolate, these methods (TLC,
HPLC, and MEKC) cannot provide unequivocal identification.
Liquid chromatography coupled to other forms of detection, especially LC-MS and LC/NMR, have the potential for online identification of unknown peaks in the analysis of the non-volatile chemicals without the need for either extensive sample purification or synthesis of standards. Because of these capabilities, these techniques
128
have gained in popularity in recent years for use in conjunction with GC/MS-based
analysis in metabolomics and metabolic profiling investigations. On-line and off-line LC-
NMR analysis has been conducted on a methanolic ginger extract, although no gingerols
were identified [23]. LC-MS based analyses of these compounds from crude ginger
rhizome extract have also been conducted [24, 25]. Among these methods, LC-MS
provides fast and accurate on-line analysis of non-volatile components from complex
matrices. However, because only single dimensional MS analysis with only positive
ionization has been reported so far (providing only parent ion molecular weight
information), no detailed structural information of the analyzed compounds could be
obtained. This may lead to ambiguous peak identifications. Furthermore, single-
dimensional LC-MS cannot provide information to potentially identify unknown
compounds, a significant challenge for metabolomics-based investigations.
The purpose of this investigation was to study the fragmentation behavior of the
gingerols, including [6]-, [8]-, [10]-, and [12]-gingerols, in ion trap LC-MS/MS, using instruments from different manufacturers, so that general ionization and fragmentation mechanisms or instrument dependent ionization and fragmentation mechanisms could be characterized for these compounds. Such information is required for accurate metabolite target analysis, metabolic profiling or metabolomics investigations. A similar investigation with the curcuminoids, major active principles in turmeric (Curcuma longa
L., also in the ginger family) that appear to be biosynthetically related to the gingerols, was able to identify clear ionization and fragmentation mechanisms for these compounds in several LC-MS/MS instruments [26]. In the present report, we describe LC-MS/MS
129
analysis of the gingerols and direct infusion ESI-MSn (n=2-7) experiments with the
gingerols that were carried out to interpret and confirm the ionization and fragmentation
behavior observed in the LC-MS/MS runs. High resolution and accurate mass
spectrometry (FT-ICR MS/MS) was also used to confirm the fragmentation behavior of
these compounds. The verified ionization and fragmentation mechanism of the gingerols
in ion trap LC-MS/MS instruments from different manufactures can be used for
unequivocal determination of gingerols with different alkyl chain lengths in raw plant
material or in processed products of ginger.
Experimental
Reagents and Chemicals
HPLC grade acetonitrile and methanol were from Burdick & Jackson (Muskegon, MI,
USA). Formic acid was from J. T. Baker (Mallinkrodt Baker, Inc., Phillipsburg, NJ,
USA). Ammonium formate was from Fisher Scientific (Fair Lawn, NJ, USA). Deionized water was re-distilled. Authentic standards of [6]-, [8]-, and [10]-gingerols were purchased from ChromaDex, Inc. (Santa Ana, CA).
LC-ESI-MS/MS analysis of gingerols
LC-ESI-MS/MS analysis of gingerols was performed on two separate ion trap mass
spectrometers: i) a ThermoElectron (Finnigan) Surveyor MS HPLC coupled to a
ThermoElectron LCQ Advantage ion trap (San Jose, CA, USA), and ii) an Agilent 1100
130
HPLC system coupled to an Agilent LC-MSD-Trap-SL ion trap mass spectrometer (Palo
Alto, CA, USA).
LC separation of gingerols—Pure gingerols were dissolved in MeOH (0.1 mg/ml)
or ginger rhizome samples were extracted with MeOH [27]. For both instruments, the
same column and elution parameters were used to achieve baseline separation of the
eluted gingerols, with almost identical elution volumes. Chromatographic separation was performed on a Discovery® HS C18 column (3 µm, 15 cm × 2.1 mm, Supelco,
Bellefonte, PA, USA) preceded by a Discovery® HS C18 guard column (3 µm, 2 cm ×
2.1 mm). Mobile phase A was water containing 5mM ammonium formate and 0.1%
formic acid buffer. Mobile phase B was Acetonitrile. The column temperature was 40°C.
The HPLC flow rate was 0.25 ml/min. 5 µl sample solution as injected into the HPLC. A
mobile phase gradient was used with the percentage of B in A varying as follows: 0-2
min, 5% B; 2-57 min, 5-100% B; 57-60 min, 100% B; 60-65 min, 100-5% B; 65-75 min,
5% B. For the LCQ Advantage instrument, a 1:10 split after the photo diode array
detector yielded a 25 µl/min flow rate into the ion trap that reduced background solvent
ions and greatly enhanced sensitivity.
MS and MS2 parameters for Thermo Electron (Finnigan) LCQ Advantage—The acquisition parameters for positive and negative mode were: sheath gas flow 29 (positive),
36 (negative) (both in arbitrary units); aux/sweep gas flow 6 (positive), 0 (negative);
source voltage 5 kV (positive), 4.5 kV (negative); source current 80 µA (positive and
negative); capillary voltage 36 V (positive), -32 V (negative); capillary temperature 270
°C; tube lens offset 45 V (positive), -25 V (negative); collision gas pressure ca. 10-5 torr;
131
Q-value 25%; mass range measured: 100-1000 m/z. These were the optimized parameters
for the maximum transmission of the gingerol-derived ions.
MS and MS2 parameters for Agilent LC-MSD-Trap-SL—The acquisition parameters for positive and negative mode were: drying N2 temperature, 350°C, 10 l/min;
nebulizer pressure 60 psi; HV capillary 4500 V; HV end plate offset -500 V; capillary
current 65.9 nA (positive mode), 62.3 (negative mode); end plate current 1482.7 nA
(positive mode), 1378.7 (negative mode); capillary exit RF amplitude 99.3 V (positive
mode), -99.3 V (negative mode); skimmer 40.0 V (positive mode), -40.0 V (negative
mode); mass range measured: 50-900 m/z. These were the optimized parameters for the
maximum transmission of the gingerol-derived ions.
Direct infusion ESI-MSn analysis of gingerols
[6]-, [8]-, and [10]-gingerols were analyzed by direct infusion ESI on three separate ion
trap instruments: a ThermoElectron LCQ Classic ESI-ion trap (San Jose, CA, USA) and
the same ThermoElectron LCQ Advantage ESI-ion trap and Agilent LC-MSD-Trap-SL
ion trap mass spectrometers used in the LC-MS/MS experiments, but with the HPLC disconnected. In order to investigate the formation of the parent ions and the fragmentation behavior of the parent ions of the gingerols in the MSn experiments, the
ionization conditions in the ThermoElectron LCQ Advantage ion trap and Agilent LC-
MSD-Trap-SL ion trap mass spectrometer were similar to those in the LC-MS/MS
experiments but optimized for ion formation and transfer under direct infusion conditions.
Ionization conditions for the Thermoelectron (Finnigan) LCQ classic instrument were as
132
follows: needle voltage: 4.5 kV (positive), -4.0 V (negative), sheath gas flow: 60
(positive), 30 (negative) (both in arbitrary units); other parameters were optimized for the
instrument and did not differ significantly from those described for the LCQ Advantage
as described above. For these analyses, the gingerols were dissolved at several different
concentrations from 5 µM to 2 mM in MeOH: H2O (8:2) containing 0.1% formic acid and 5 mM ammonium formate. The infusion flow rate in all cases was 10 µl/min.
FT-ICR analysis of gingerols
High resolution and accurate mass measurements were carried out on an IonSpec 4.7 T
FT-ICR instrument (Lake Forest, CA). Ions were generated using an Analytica (Branford,
CT) second generation electrospray (ESI) source using diluted (5:1, in MeOH:H2O 5:1)
solutions of those that were analyzed in the ion trap instruments. Direct infusion with a
flow rate of 2 µl/min was applied to generate negatively charged ions by ESI. The exact
masses (elemental composition) of the deprotonated molecules of [6]-, [8]-, and [10]-
gingerols were determined using alkyl sulphonates as internal standards. Tandem MS/MS
fragmentation was achieved using the sustained off-resonance irradiation (SORI)
technique with N2 as collision gas. The SORI excitation time was 1500 ms with a N2 pressure of ca. 5 × 10-6 torr in the ICR cell.
133
RESULTS AND DISCUSSION
LC-(–)ESI-MS and LC-(–)ESI-MS/MS analysis of [6]-, [8]-, and [10]-
gingerols in a ThermoElectron LCQ Advantage ion trap mass spectrometer
When we used our ThermoElectron LCQ Advantage instrument to perform LC-(–)ESI-
MS-based metabolic profiling analysis of extracts from ginger rhizomes, we were very
surprised to find that we could not detect the predicted parent ions [M-H]- at m/z of 293,
321, and 349, respectively, of [6]-, [8]-, and [10]-gingerols, although these are among the
most abundant compounds produced by this plant tissue. UV detection in the diode array
detector, however, clearly showed that these compounds were present and were very
abundant. Attempts to tune the instrument on pure standard compounds also failed to
produce the expected parent ions [M-H]- for these compounds. Instead, we detected ions
with m/z of 585, 641, and 697, respectively, in the attempts to tune the instrument. These
corresponded to ions that appeared to be covalent dimers of the respective gingerols
[(2M-2H)-H]-. From the total ion chromatograms (TICs) from LC-MS analysis of
methanolic extracts of ginger samples or of standard gingerol compounds, we detected
these ions, respectively, in the full MS spectra at elution volumes of 8.075 ml, 9.700 ml,
and 11.150 ml using the HPLC gradient profile described in Experimental section. These
elution volumes were the same as those observed using diode array detection of pure
compounds analyzed individually, supporting the identification of these compounds and confirming that the covalent dimer ions did indeed represent the only ionized form of the gingerols in this instrument. Thus, it appeared that the LCQ Advantage was able to form the covalent dimer and not the monomer ions of the gingerols during the ESI ionization
134
process. In contrast, the biosynthetically related curcuminoids ionized as expected in this
instrument, with the monomer [M-H]- as the major observed ion and no evidence of
covalent dimer formation [26]. We have often seen non-covalent dimer [2M-H]- adducts accompanying the monomer [M-H]- ion in our LC-MS analysis, with the LCQ Advantage and with other instruments. Thus, this unusual ionization did not appear to be the result of
some miscalibration or faulty setup of the instrument. However, formation of the
apparent covalent dimers of the gingerols was very intriguing and led to further
examination of the ionization and fragmentation of the gingerols that is described in this manuscript.
Further analysis of the gingerols in the LCQ Advantage using negative mode LC-
ESI-MS/MS suggested that the parent ions observed (m/z of 585, 641, and 697) were indeed covalent dimers. These ions did not break apart to form the monomer parent ion
[M-H]- when altered source conditions (in source fragmentation) or different
fragmentation energies in the trap were applied. Instead, the observed ions at m/z of 585,
641, and 697 all behaved as covalent entities in MS/MS experiments. The fragmentation of these ions occurs through just a few fragmentation channels producing simple MS/MS spectra (Fig. 1) and demonstrating the same profiles of fragmentation for all gingerols
examined. As shown in Fig. 1, two daughter ions B and D are present as the major
fragments. These appeared to represent loss of water and loss of a larger functional group,
which was later determined to be the alkyl side chain of the gingerols (see below). Three
additional diagnostic daughter ions C, E, and F were also observed in the MS/MS spectra
of the [(2M-2H)-H]- ion (covalent dimer) of each gingerol.
135
Direct infusion (–)ESI-MSn analysis of [6]-, [8]-, [10]- and [12]-gingerols
in a ThermoElectron LCQ Advantage ion trap mass spectrometer
Because we were not able to detect the monomer [M-H]- or the non-covalent dimer [2M-
H]- ions of the gingerols in negative mode LC-ESI-MS or LC-ESI-MS/MS in the LCQ
Advantage but instead only detected what appeared to be covalent dimers [(2M-2H)-H]-, we proceeded to characterize the ionization and fragmentation of these compounds in this instrument using direct infusion ESI ionization. Evaluation of standard gingerols at several different dilutions, from very concentrated to barely detectable, gave the same ionization results: in full scan MS we were only able to detect the formation of ions that appeared to be the covalent dimer ions [(2M-2H)-H]- of the gingerols. We were never
able to observe the monomer [M-H]-or non-covalent dimer [2M-H]- ions. We then
proceeded to use higher order (–)ESI-MSn experiments to characterize the structure of the
putative covalent dimers of [6]-, [8]-, and [10]-gingerols and their corresponding
daughter ions. As shown in Table 2, MSn experiments up to level MS7 were performed to
characterize the structure of these ions. Based on the fragmentation behavior of the parent
ions in these experiments, we proposed and verified the structure of ion A [(2M-2H)-H]-
(covalent dimers) and its corresponding daughter ions (Scheme 1).
In (–)ESI-MS2 experiments with ion A [(2M-2H)-H]- (covalent dimer), major daughter ions B and D were observed, just as we had seen in LC-(–)ESI-MS/MS analyses.
The formation of daughter ion B can proceed by loss of one H2O (Scheme 1) from A. The
other daughter ion D can be formed via McLafferty rearrangement plus loss of a neutral moiety (Scheme 1). This corresponds to the alkyl side chain of the respective gingerol
136
with C-C bond cleavage between carbons 4 and 5, which is β to the ketone functional
group. In (–)ESI-MS3 experiments, daughter ion B from the (–)ESI-MS2 experiments was
further fragmented, leading to the formation of daughter ions C (loss one H2O) and E
(McLafferty rearrangement followed by the loss of the same neutral moiety as in
formation of D from A, Scheme 1). The other daughter ion D from the (–)ESI-MS2
experiments was also subjected to further fragmentation in the MS3 experiments and two
daughter ions E and F were thus formed by loss of H2O and McLafferty rearrangement
plus loss of a neutral moiety (Scheme 1), respectively. As indicated in Scheme 1,
daughter ion E can be obtained from both daughter ions B and D. All daughter ions,
including B, C, D, E, and F, formed in the (–)ESI-MS2 or MS3 experiments, were also
observed in the LC-(–)ESI-MS/MS analyses and were produced by loss of one or more
H2O and/or via McLafferty rearrangement plus loss one or more neutral moieties
(Scheme 1).
To further confirm the proposed structures of ions A, B, C, D, E, and F, we carried out (–)ESI-MSn experiments up to (–)ESI-MS7. In the (–)ESI-MS4 experiments, daughter ion C from the (–)ESI-MS3 experiments was subjected to further fragmentation, leading to the formation of daughter ion G by rearrangement and loss of two radical moieties (Scheme 1). Similarly, daughter ion I was formed in the (–)ESI-MS4
experiments by fragmentation of daughter ion E. Daughter ion H was formed by loss of a
methyl radical from ion F (Scheme 1). In (–)ESI-MS5 to MS7 experiments, some
daughter ions were produced via multiple routes as indicated in Scheme 1 and Table 2.
Daughter ion I, obtained in the (–)ESI-MS4 experiments from daughter ion E, could also
137
be achieved from daughter ion H by loss of an acetyl group in the (–)ESI-MS5
experiments (Scheme 1). Daughter ion J, however, was only produced from daughter ion
I by loss of methyl radical in the (–)ESI-MS5 experiments (Scheme 1). Daughter ion K
could be produced from daughter ion G by loss of methyl and acetyl groups in the (–
)ESI-MS5 experiments or from daughter ion J by the loss of acetyl radical in the MS6 experiments (Scheme 1 and Table 2). Furthermore, daughter ions L, M, and N were observed in the (–)ESI-MS7 experiment of daughter ion K by loss of one hydroxy group
or loss of one or two CO molecules, respectively (Scheme 1 and Table 2).
As proposed in Scheme 1, ion A [(2M-2H)-H]- (covalent dimer) of the gingerols
is believed to be formed from two monomers via C-C (biphenyl) instead of C-O-C
(phenylether) coupling. The reason for this proposal is that daughter ions, including H
and J, were formed via cleavage of C-O-C bonds in the (–)ESI-MSn experiments, and it
would be expected that cleavage of phenylether coupled dimers would also occur under
these conditions. No such cleavage was observed. Instead, daughter ions L, M, and N in the (–)ESI-MS7 experiments were observed, which still contained the moiety of the
covalent dimer, suggesting that this covalent bond is quite stable. In addition, the C-C
coupling is likely to occur at position C-5′ of the phenyl ring in the gingerols because of
weaker steric hindrance from the side chain and methoxy group (Scheme 1). Finally,
based on the respective fragmentation patterns of the gingerols described above, it was
not possible for this covalent bond to exist as an ether involving the side chain Os.
138
LC-(–)ESI-MS/MS analysis of [6]-, [8]-, [10]-, and [12]-gingerols in an
Agilent LC-MSD-Trap-SL ion trap mass spectrometer
In order to determine whether the formation of parent ions A [(2M-2H)-H]- (covalent
dimers) and their fragmentation behaviors depend solely on the nature of the compounds
themselves or also on the instrument used for analysis, we used the same exact
methanolic extracts of ginger samples and standard gingerols dissolved in methanol for
LC-(–)ESI-MS analysis on an Agilent LC-MSD-Trap-SL ion trap mass spectrometer.
From the TICs of these samples, we were not able to detect the covalent dimer parent ion
A (m/z of 585, 641, and 697) for any of the gingerols. Instead, we observed the predicted monomer parent ion O [M-H]- at m/z of 293.5, 321.7, and 349.7, respectively, for [6]-,
[8]-, and [10]-gingerol (see Table 3). Ions O were observed at delayed elution volumes
(~0.625 ml delay) compared to the ThermoElectron LCQ Advantage, due to differences
in the HPLC pump designs for these two instruments (the ThermoElectron Surveyor MS
Pump has a significantly smaller gradient mixing cell than the Agilent 1100 series HPLC
pump). In addition, using the Agilent LC-MSD-Trap-SL ion trap we were able to detect
ion O [M-H]- at m/z of 377.5 of [12]-gingerol (Table 3) at elution volume 12.575 ml in
the full MS TICs of methanolic extracts of ginger. In LC-(–)ESI-MS/MS analysis, the gingerols demonstrated identical fragmentation profiles. As shown in Fig. 2, ion Q (m/z of 193.2) is the base peak in the MS/MS spectra of all deprotonated [M-H]- gingerol
monomers. Three additional characteristic ions, P, R, and S, are present as minor peaks
in the MS/MS spectra of ion O [M-H]-.
139
Direct infusion (–)ESI-MSn analysis of [6]-, [8]-, and [10]-gingerols in an
Agilent LC-MSD-Trap-SL ion trap mass spectrometer
Extensive efforts to tune the Agilent LC-MSD-Trap-SL in an attempt to form ion A
[(2M-2H)-H]- (covalent dimer) of the gingerols failed to produce this ion. Instead, only
the deprotonated monomer ion O [M-H]- and small amounts of the deprotonated non-
covalent dimer adduct [2M-H]- were formed in this instrument for either diluted or
concentrated standard gingerol solutions. (–)ESI-MSn experiments (Table 3) via direct
infusion of [6]-, [8]-, and [10]-gingerols were then conducted to elucidate the structures
of the corresponding daughter ions observed in the LC-(–)ESI-MS/MS analyses. Based
on the fragmentation behavior of parent ion O of the gingerols in these experiments, we
have proposed the structures and fragmentation patterns of the corresponding daughter
ions as outlined in Scheme 2. In the (–)ESI-MS2 experiments with ion O [M-H]-,
daughter ions P, Q, and S were observed. The formation of daughter ion P can proceed
2 by loss of one H2O. Ion Q, the base peak in the (–)ESI-MS spectra of ion O, can be
formed via McLafferty rearrangement plus loss of neutral moiety, as was observed in the
ThermoElectron LCQ Advantage. Ion S is a minor peak in the (–)ESI-MS2 spectra of ion
O, produced by ionization of the lost side chain of ion O.
In order to further confirm the structure of the major daughter ion Q, (–)ESI-MS3
experiments were performed to fragment this ion, leading to the formation of ion R by
loss of a radical moiety (Scheme 2). Ion R was also observed in the LC-(–)ESI-MS/MS
spectra of ion O (Fig. 2). Further fragmentation of ion R was carried out in (–)ESI-MS4 experiments, leading to the formation of ion T by loss of an acetyl group (Scheme 2).
140
When Schemes 1 and 2 are compared, it becomes quite clear that ions A [(2M-2H)-H]-
- and O [M-H] fragment similarly by loss of H2O and/or radical moiety, and/or
McLafferty rearrangement. Therefore, the formation of parent ions A and O in the
negative mode ESI-MS appear to be instrument dependent, while the subsequent
fragmentation of these parent ions appears to be instrument independent and relies,
instead, on the properties of the constituent moieties of the gingerols.
Direct infusion (–)ESI-MSn analysis of [6]-, [8]-, and [10]-gingerols in a
ThermoElectron LCQ Classic ion trap mass spectrometer
To further confirm our observations about the formation of parent ions and corresponding
fragmentation behavior of the gingerols, we performed direct infusion (–)ESI-MSn experiments in a ThermoElectron LCQ Classic ion trap mass spectrometer. However, the deprotonated parent ions of the monomer were more abundant in this instrument in the (–
)ESI-MS/MS analysis under certain buffer conditions and concentrations of the gingerols, where the sample was diluted three fold in methanol over standard sample concentration conditions and a very small amount (a very few microliters) of NH4OH was added.
Interestingly, we were able to observe both the A [(2M-2H)-H]- and O [M-H]- parent ions
in (–)ESI full MS spectra by optimizing the instrument set up and sample concentration
(data not shown). The same fragmentation profiles observed in the ThermoElectron LCQ
Advantage and the Agilent LC-MSD-Trap-SL and shown in Fig. 1 and Fig. 2,
respectively, were observed in (–)ESI-MS/MS experiments for ions A and O formed in the LCQ Classic. These results further verified our observation that the formation of
141
parent ions (A or O) of the gingerols in (–)ESI-MS is instrument dependent, while subsequent fragmentation in MSn experiments is not. This also suggests that formation of
ion A occurs in the source or heated capillary and not in the trap, especially because
formation of A over O in the LCQ Classic depended on gingerol concentration, where
only covalent dimer (A) was found at higher concentrations and monomer (O) could be
observed at lower concentrations.
LC-(+)ESI-MS and LC-(+)ESI-MS/MS analysis of [6]-, [8]-, and [10]-
gingerols in a ThermoElectron LCQ Advantage ion trap mass spectrometer
Compared to the negative (–)ESI spectra, the positive (+)ESI full MS spectra of the
gingerols are more complicated. From the TICs of methanolic extracts of ginger samples or of standard gingerols, we detected several different positively charged ions for each
gingerol in our HPLC buffer system. These ions included the protonated parent ions of
the covalent dimer [(2M-2H)+H]+ and monomer [M+H]+. Additional adducts ions, such
+ + as ammonium adducts of the monomer [M+NH4] and covalent dimer [(2M-2H)+NH4] ions, and sodium adduct of the covalent dimer [(2M-2H)+Na]+ ion, as well as an ion
+ corresponding to the loss of water from the protonated monomer ion, [M+H-H2O] , (see
Table 1) in the full MS spectra at the same elution volumes as observed in LC-(–)ESI-MS
for the corresponding gingerols. Direct infusion (+)ESI-MS analysis of these gingerols
was also performed in this instrument, yielding the same results as obtained for the LC-
based analysis. In both analyses, the profiles of protonated ions and other adduct ions
depended on the composition of the buffer used to dissolve or elute the sample. When,
142
for example, the buffer contained formic acid and ammonium formate, at whatever
concentration evaluated, the protonated ions ([(2M-2H)+H]+ and [M+H]+) were present
as only minor peaks in (+)ESI full MS spectra. Interestingly, we also observed the
covalent dimer [(2M-2H)+H]+ ion, even though this ion was present as a minor peak in
the (+)ESI full MS spectra.
LC-(+)ESI-MS/MS experiments were performed to further characterize the
+ + ammonium adducts of the monomer [M+NH4] and covalent dimer [(2M-2H)+NH4]
+ ions and loss of water fragment of the monomer ion [M+H-H2O] that were observed in
full scan LC-(+)ESI-MS. In these experiments, the monomer ammonium adduct
+ [M+NH4] ion lost NH3 and H2O. The covalent dimer ammonium adduct [(2M-
2H)+NH4]+ ion, on the other hand, lost species with 100, 128 or 156 daltons for [6]-, [8]-
or [10]-gingerol, respectively, in LC-(+)ESI-MS/MS. These processes are triggered by
the McLafferty rearrangement followed by the loss of a neutral moiety, as observed in (–
)ESI-MS/MS and shown in Scheme 1. The loss of water fragment of the monomer
+ [M+H-H2O] ion also lost species with 100, 128 or 156 daltons, respectively, resulting in
formation of an ion at m/z 177 for all three gingerols. However, because of the prior loss
of water, eliminating the hydroxyl group on the alkyl side chain, a McLafferty rearrangement was not possible. Instead, loss of these fragments in (+)ESI-MS/MS presumably occurs by a different rearrangement mechanism, as shown in Scheme 3.
Direct infusion (+)ESI-MS/MS of the gingerols in this instrument also supported these observations.
143
LC-(+)ESI-MS and LC-(+)ESI-MS/MS analysis of [6]-, [8]-, and [10]-
gingerols in an Agilent LC-MSD-Trap-SL ion trap mass spectrometer
As for the ThermoElectron LCQ Advantage, more complicated spectra were obtained in
(+)ESI full MS analysis of the gingerols in the Agilent LC-MSD-Trap-SL than were
produced in (–)ESI full MS analysis. The protonated monomer [M+H]+ parent ion, and
+ + other adducts ions including ammonium adduct [M+NH4] and sodium adduct [M+Na] of the monomer ion, sodium adduct of the non-covalent dimer [2M+Na]+ ion, and loss of
+ water fragment of the monomer [M+H-H2O] ion (see Table 1) were detected in the
(+)ESI full MS spectra at the same elution volumes as observed for each gingerol in LC-
(–)ESI-MS/MS. Direct infusion (+)ESI-MS analysis of these gingerols was also
conducted. As for the LCQ Advantage, changes in composition of the buffer used to dissolve or elute sample affected the profiles of the protonated and other adduct ions.
Unlike the LCQ Advantage, however, the Agilent LC-MSD-Trap-SL did not produce
ions or adduct ions that were derived from covalent dimers of the gingerols, just as was
seen in negative ionization analysis.
As was done with the ThermoElectron LCQ Advantage, LC-(+)ESI-MS/MS
experiments were performed in the Agilent LC-MSD-Trap-SL to further characterize the
ions and adducts formed in LC-(+)ESI-MS analysis of the gingerols. The sodium adduct
of the non-covalent dimer [2M+Na]+ ion was targeted for fragmentation and yielded loss
of one non-covalently interacting monomer and retention of the sodium adduct. The
+ + ammonium adduct [M+NH4] and loss of water fragment of the monomer [M+H-H2O] ions showed the same fragmentation pattern as was described above for (+)ESI-MS/MS
144
in the ThermoElectron LCQ Advantage. As described above, the sodium adduct of the
monomer [M+Na]+ ion lost species with 100, 128 or 156 daltons in a McLafferty
rearrangement resulting in a sodiated daughter ion at m/z 217 for all three gingerols.
Direct infusion (+)ESI-MS/MS of the gingerols in this instrument also supported these
observations.
Direct infusion (+)ESI-MS and (+)ESI-MS/MS analysis of [6]-, [8]-, and
[10]-gingerols in a ThermoElectron LCQ Classic ion trap mass
spectrometer
As was observed in the other two instruments, more complicated spectra were obtained in
(+)ESI full MS analysis of the gingerols in the ThermoElectron LCQ Classic than were
produced in (–)ESI full MS analysis. This instrument produced the protonated parent ions
of the covalent dimer [(2M-2H)+H]+ and monomer [M+H]+, and additional adducts ions
+ including the ammonium adduct of the monomer [M+NH4] , and sodium adducts of the
monomer [M+Na]+ and the noncovalent dimer [2M+Na]+, and the loss of water fragment
+ of the monomer [M+H-H2O] ions direct infusion (+)ESI-MS analysis of the gingerols.
As for the other two instruments, changes in composition of the buffer used to dissolve
the sample affected the profiles of the protonated and other adduct ions. Under all buffer conditions tested that contained formic acid and ammonium formate, the protonated ions
([(2M-2H)+H]+ and [M+H]+) were present only as minor peaks in the (+)ESI full MS
spectra. Interestingly, both the covalent dimer [(2M-2H)+H]+ and non-covalent dimer
145
[2M+Na]+ ions of the gingerols were observed in this instrument, even though the
covalent dimer [(2M-2H)+H]+ ion was a very minor peak in the (+)ESI full MS spectra.
In (+)ESI-MS/MS analysis in the ThermoElectron LCQ Classic instrument, the
ammonium adducts, sodium adducts, and loss of water fragment of the monomer ions
showed the same fragmentation behavior as was observed from in the ThermoElectron
LCQ Advantage and Agilent LC-MSD-Trap-SL instruments. Therefore, (+)ESI-MS/MS experiments could be performed on these ions. Interestingly, a daughter ion at m/z 195 was observed in the (+)ESI-MS/MS of all three gingerols, due to McLafferty rearrangement and loss of species with 100, 128 or 156 daltons, respectively. This observation suggested that protonation of the gingerols can occur on different oxygen atoms of the molecule. If the protonation occurs on the hydroxy group of the alkyl side chain, subsequent loss of water occurs preventing McLafferty rearrangement, as discussed above. If on the other hand, protonation of the phenolic or methoxyl oxygens
(attached to the aromatic ring) occurs, McLafferty rearrangement occurs.
FT-ICR analysis of [6]-, [8]-, and [10]-gingerols
To confirm the identity of the parent and daughter ions produced by (–)ESI in the Agilent
and/or ThermoElectron ion trap instruments, we performed high resolution and accurate mass FT-ICR MS measurements of negatively charged ions of the gingerols. The resulting ions, nominal masses, accurate masses, and corresponding molecular formulae
and predicted masses are summarized in Table 4. Ions O, Q, and S were observed from
146
the FT-ICR SORI MS/MS spectra in the analysis of gingerols. Among these ions, ion Q
is present as the base peak. The molecular formulae of these daughter ions support our
proposed structures (shown in Scheme 2). The covalent dimer [(2M-2H)-H]- ion, formed in the ThermoElectron but not in the Agilent ion trap mass spectrometers, was not produced in negative mode ESI by the FT-ICR instrument. The formation and fragmentation of the parent ions of the gingerols in (–)ESI in the FT-ICR MS/MS analysis demonstrated the same behavior as that observed in the (–)ESI-MS/MS analysis in the Agilent LC-MSD-Trap-SL, confirming the results obtained from this instrument.
This also supports the conclusion that formation of parent ions A and O and the corresponding fragmentation behaviors of these parent ions in negative mode ESI-MSn is instrument dependent.
CONCLUSIONS
Ion trap-based ESI-MS/MS is a very sensitive and precise method for identifying the
gingerols. Consistent and predictable ionization and fragmentation behaviors were
observed for all gingerols when analyzed in the same instrument. In negative mode ESI-
MS, ion trap instruments from different manufacturers produced different parent ions
from the gingerols, indicating that different ion formation mechanisms occur during ESI
in these instruments. On the other hand, the very similar fragmentation patterns of the
negatively charged gingerol ions in MS/MS or MSn experiments in these instruments
show that fragmentation mechanisms appear to be instrument independent. This was
supported by similar experiments with the same compounds using a high resolution and
147 accurate mass ESI-FT-ICR instrument. A similar situation was observed for positive mode ESI-MS and ESI-MS/MS in these instruments. These results clearly demonstrate that LC-MS instruments produce data that cannot necessarily be replicated in other laboratories, especially if those laboratories do not have the same instrument model from the same manufacturer. This presents potential problems for metabolic profiling and metabolomics investigations. Such investigations rely on rapid and high throughput analysis of complex samples, which often are left in rather crude form, for compound identification and analysis. In such investigations, GC/MS analysis has been commonly used because standard libraries can be utilized across platforms due to the relative uniformity of ionization, fragmentation, and detection in EI-quadrupole mass spectrometers. However, GC/MS is not able to identify, detect and certainly not quantitate compounds of all classes. For this reason, LC-MS based approaches have gained in popularity in recent years [28-30]. However, our results clearly show that great caution must be taken in such LC-MS and LC-MS/MS based analyses, where great differences in ionization can lead to very different types of ions being formed from the same molecule in similar instruments made by different manufacturers (e.g., the non- covalent vs. covalent dimers of the gingerols observed in this investigation). This makes production of standard mass spectrum libraries a very difficult task, because such libraries are likely to not be completely valid across platforms, even though a standard framework for metabolomics and metabolic profiling experiments has begun to be established [31]. In addition, analyses performed on the same samples by different laboratories may yield very different results for certain compounds, if these laboratories
148 do not use identical instruments. This may especially be the case for metabolite target analysis investigations using selective ion recording/monitoring experiments, where the predicted ion may not actually be formed, depending on the instrument. In such cases, care must be taken to ensure that the predicted analyte can be observed in the instrument chosen, whether it exists as the predicted ion or not.
ACKNOWLEDGEMENTS
The authors acknowledge financial assistance from the National Science Foundation
Plant Genome Program, grant DBI-0227618 to D.R.G., and the National Institutes of
Health NCCAM/ODS, grants #5 P50 AT 000474-04 and 3 P50 AT 000474-03 S1 to
B.N.T. We wish to thank Dr. Aniko Solyom for access to the Agilent LC-MSD-Trap-SL.
The contents of this publication are solely the responsibility of the authors and do not necessarily represent the official views of NCCAM, ODS, or the National Institutes of
Health.
149
References
1. Gong, F.; Fung, Y. S.; Liang, Y. Z. Determination of Volatile Components in
Ginger Using Gas Chromatography-Mass Spectrometry with Resolution
Improved by Data Processing Techniques. J. Agr. Food Chem. 2004, 52, 6378-
6383.
2. Mustafa, T.; Srivastava, K. C.; Jensen, K. B. Drug Development Report.9.
Pharmacology of Ginger, Zingiber-officinale. J. Drug Dev. 1993, 6, 25-39.
3. Govindarajan, V. S. Ginger--Chemistry, Technology, and Quality Evaluation:
Part 1. Crit. Rev. Food. Sci. Nutr. 1982, 17, 1-96.
4. Adewunmi, C. O.; Oguntimein, B. O.; Furu, P. Molluscicidal and
Antischistosomal Activities of Zingiber officinale. Planta Med. 1990, 56, 374-376.
5. Bhattarai, S.; Tran, V. H.; Duke, C. C. The Stability of Gingerol and Shogaol in
Aqueous Solutions. J. Pharm. Sci. 2001, 90, 1658-1664.
6. Ficker, C.; Smith, M. L.; Akpagana, K.; Gbeassor, M.; Zhang, J.; Durst, T.;
Assabgui, R.; Arnason, J. T. Bioassay-Guided Isolation and Identification of
Antifungal Compounds from Ginger. Phytother. Res. 2003, 17, 897-902.
7. Hikino, H.; Kiso, Y.; Kato, N.; Hamada, Y.; Shioiri, T.; Aiyama, R.; Itokawa, H.;
Kiuchi, F.; Sankawa, U. Antihepatotoxic Actions of Gingerols and
Diarylheptanoids. J. Ethnopharmacol. 1985, 14, 31-39.
8. Koo, K. L.; Ammit, A. J.; Tran, V. H.; Duke, C. C.; Roufogalis, B. D. Gingerols
and Related Analogues Inhibit Arachidonic Acid-Induced Human Platelet
Serotonin Release and Aggregation. Thromb. Res. 2001, 103, 387-397.
150
9. Masuda, Y.; Kikuzaki, H.; Hisamoto, M.; Nakatani, N. Antioxidant Properties of
Gingerol Related Compounds from Ginger. Biofactors 2004, 21, 293-296.
10. Ruedi, P.; Juch, M. Chemistry and Biological Activities of Long-Chain Alkyloxy-
Catechols of the [n]-Gingerol-Type. Curr. Org. Chem. 1999, 3, 623-646.
11. Surh, Y. J. Anti-tumor Promoting Potential of Selected Spice Ingredients with
Antioxidative and Anti-Inflammatory Activities: a Short Review. Food Chem.
Toxicol. 2002, 40, 1091-1097.
12. Yamahara, J.; Mochizuki, M.; Rong, H. Q.; Matsuda, H.; Fujimura, H. The Anti-
ulcer Effect in Rats of Ginger Constituents. J. Ethnopharmacol. 1988, 23, 299-
304.
13. Young, H. Y.; Luo, Y. L.; Cheng, H. Y.; Hsieh, W. C.; Liao, J. C.; Peng, W. H.
Analgesic and Anti-inflammatory Activities of [6]-Gingerol. J. Ethnopharmacol.
2005, 96, 207-210.
14. Chen, C. C.; Ho, C. T. Gas-Chromatographic Analysis of Volatile Components of
Ginger Oil (Zingiber-officinale Roscoe) Extracted with Liquid Carbon-Dioxide. J.
Agr. Food Chem. 1988, 36, 322-328.
15. Chen, C. C.; Rosen, R. T.; Ho, C. T. Chromatographic Analysis of Gingerol
Compounds in Ginger (Zingiber-officinale-Roscoe) Extracted by Liquid Carbon-
Dioxide. J. Chromatogr. 1986, 360, 163-173.
16. Harvey, D. J. Gas-Chromatographic and Mass-Spectrometric Studies of Ginger
Constituents-Identification of Gingerdiones and New Hexahydrocurcumin
Analogs. J. Chromatogr. 1981, 212, 75-84.
151
17. Masada, Y.; Inoue, T.; Hashimot, K.; Fujioka, M.; Shiraki, K. Studies on Pungent
Principles of Ginger (Zingiber officinale Rosc.) by GC/MS. Yakugaku Zasshi
1973, 93, 318-321.
18. Masada, Y.; Inoue, T.; Hashimot, K.; Fujioka, M.; Uchino, C. Studies on
Constituents of Ginger (Zingiber-officinale-Roscoe) by GC-MS. Yakugaku Zasshi
1974, 94, 735-738.
19. Singh, G.; Maurya, S.; Catalan, C.; de Lampasona, M. P. Studies on Essential Oils,
Part 42: Chemical, Antifungal, Antioxidant and Sprout Suppressant Studies on
Ginger Essential Oil and its Oleoresin. Flavour Frag. J. 2005, 20, 1-6.
20. Balladin, D. A.; Headley, O.; Chang-Yen, I.; McGaw, D. R. High Pressure Liquid
Chromatographic Analysis of the Main Pungent Principles of Solar Dried West
Indian Ginger (Zingiber officinale Roscoe). Renew. Energ. 1998, 13, 531-536.
21. Connell, D. W.; McLachlan, R. Natural Pungent Compounds.4. Examination of
Gingerols, Shogaols, Paradols and Related Compounds by Thin-Layer and Gas-
Chromatography. J. Chromatogr. 1972, 67, 29-35.
22. Huang, H. Y.; Kuo, K. L.; Hsieh, Y. Z. Determination of Cinnamaldehyde,
Cinnamic Acid, Paeoniflorin, Glycyrrhizin and [6]-Gingerol in the Traditional
Chinese Medicinal Preparation Kuei-chih-tang by Cyclodextrin-Modified
Micellar Electrokinetic Chromatography. J. Chromatogr. A 1997, 771, 267-274.
23. Saha, S.; Smith, R. M.; Lenz, E.; Wilson, I. D. Analysis of a Ginger Extract by
High-Performance Liquid Chromatography Coupled to Nuclear Magnetic
152
Resonance Spectroscopy Using Superheated Deuterium Oxide as the Mobile
Phase. J. Chromatogr. A 2003, 991, 143-150.
24. He, X. G.; Bernart, M. W.; Lian, L. Z.; Lin, L. Z. High-Performance Liquid
Chromatography Electrospray Mass Spectrometric Analysis of Pungent
Constituents of Ginger. J. Chromatogr. A 1998, 796, 327-334.
25. Hiserodt, R. D.; Franzblau, S. G.; Rosen, R. T. Isolation of 6-, 8-, and 10-
Gingerol from Ginger Rhizome by HPLC and Preliminary Evaluation of
Inhibition of Mycobacterium Avium and Mycobacterium Tuberculosis. J. Agr.
Food Chem. 1998, 46, 2504-2508.
26. Jiang, H.; Somogyi, A.; Timmermann, B. N.; Gang, D. R. ESI-MS/MS Analysis
of Curcuminoids. Submitted to Rapid Commun. Mass Spectrom. 2005.
27. Jiang, H.; Xie, Z.; Koo, H.; McLaughlin, S. P.; Timmermann, B. N.; Gang, D. R.
Metabolic Profiling and Phylogenetic Analysis of Medicinal Zingiber species:
Tools for Authentication of Ginger (Zingiber officinale Rosc.). Submitted to
Phytochemistry 2005.
28. Bino, R. J.; Hall, R. D.; Fiehn, O.; Kopka, J.; Saito, K.; Draper, J.; Nikolau, B. J.;
Mendes, P.; Roessner-Tunali, U.; Beale, M. H.; Trethewey, R. N.; Lange, B. M.;
Wurtele, E. S.; Sumner, L. W. Potential of Metabolomics as a Functional
Genomics tool. Trends Plant Sci. 2004, 9, 418-425.
29. Halket, J. M.; Waterman, D.; Przyborowska, A. M.; Patel, R. K.; Fraser, P. D.;
Bramley, P. M. Chemical Derivatization and Mass Spectral Libraries in Metabolic
Profiling by GC/MS and LC/MS/MS. J. Exp. Bot. 2005, 56, 219-243.
153
30. Sumner, L. W.; Mendes, P.; Dixon, R. A. Plant Metabolomics: Large-Scale
Phytochemistry in the Functional Genomics Era. Phytochemistry 2003, 62, 817-
836.
31. Jenkins, H.; Hardy, N.; Beckmann, M.; Draper, J.; Smith, A. R.; Taylor, J.; Fiehn,
O.; Goodacre, R.; Bino, R. J.; Hall, R.; Kopka, J.; Lane, G. A.; Lange, B. M.; Liu,
J. R.; Mendes, P.; Nikolau, B. J.; Oliver, S. G.; Paton, N. W.; Rhee, S.; Roessner-
Tunali, U.; Saito, K.; Smedsgaard, J.; Sumner, L. W.; Wang, T.; Walsh, S.;
Wurtele, E. S.; Kell, D. B. A Proposed Framework for the Description of Plant
Metabolomics Experiments and Their Results. Nat. Biotechnol. 2004, 22, 1601-
1606.
154
Table 1. Parent and adduct ions of [6]-, [8]-, [10]-, and [12]-gingerols in the (–)ESI and
(+)ESI full MS spectra from LC-ESI-MS analysis in a ThermoElectron LCQ Advantage
ion trap mass spectrometer and a Agilent LC-MSD-Trap-SL ion trap mass spectrometer
coupled to HPLC.
Compounds Mr Ionsa Ionsb Data from ThermoFinnigan LCQ Advantage ion trap mass spectrometer - + + [6]-gingerol 294 585.3 [(2M-2H)-H] 312.1 [M+NH4] ; 277.1 [M+H-H2O] ; + + 604.0 [(2M-2)+NH4] ; 609.3 [(2M-2)+Na] ; 587.0 [(2M-2)+H]+ ; 294.9 [M+H]+ - + + [8]-gingerol 322 641.2 [(2M-2H)-H] 340.1 [M+NH4] ; 305.1 [M+H-H2O] ; + + 660.0 [(2M-2)+NH4] ; 665.3 [(2M-2)+Na] ; 643.2 [(2M-2)+H]+ ; 323.0 [M+H]+ - + + [10]-gingerol 350 697.5 [(2M-2H)-H] 368.1 [M+NH4] ; 333.0 [M+H-H2O] ; + + 716.2 [(2M-2)+NH4] ; 721.5 [(2M-2)+Na] ; 699.0 [(2M-2)+H]+ ; 323.0 [M+H]+ Data from Agilent LC-MSD-Trap-SL ion trap mass spectrometer - + + [6]-gingerol 294 293.5 [M-H] 277.4 [M+H-H2O] ; 317.3 [M+Na] ; + + + 312.4 [M+NH4] ; 611.0 [2M+Na] ; 295.3 [M+H] - + + [8]-gingerol 322 321.7 [M-H] 305.6 [M+H-H2O] ; 345.4 [M+Na] ; + + + 340.5 [M+NH4] ; 667.3 [2M+Na] ; 323.5 [M+H] - + + [10]-gingerol 350 349.7 [M-H] 333.5 [M+H-H2O] ; 373.4 [M+Na] ; + + + 368.5 [M+NH4] ; 723.5 [2M+Na] ; 351.5 [M+H] - + + [12]-gingerol 378 377.5 [M-H] 361.6 [M+H-H2O] ; 401.5 [M+Na] ; + + + 396.6 [M+NH4] ; 779.5 [2M+Na] ; 379.6 [M+H]
Ionsa means ions observed from (–)ESI full MS spectra; Ionsb means ions observed from
(+)ESI full MS spectra.
Table 2. Ions formed in direct infusion (–)ESI-MSn experiments with [6]-gingerol, [8]-gingerol, and [10]-gingerol in
ThermoElectron LCQ Advantage ion trap mass spectrometers.
MS MS2 MS3 MS4 MS5 MS6 MS7 567(B) 549(C) 409 (G) 269 (K) * [6]-gingerol 585(A) 467 (E) 327 (I) 312 (J) 269 (K) * 485(D) 467 (E) 327 (I) 312 (J) 269 (K) * 385 (F) 370 (H) 327 (I) 312 (J) 269 (K) 623(B) 605 (C) 437 (G) 269 (K) * [8]-gingerol 641(A) 495 (E) 327 (I) 312 (J) 269 (K) * 513(D) 495 (E) 327 (I) 312 (J) 269 (K) * 385 (F) 370 (H) 327 (I) 312 (J) 269 (K) 679(B) 661 (C) 465 (G) 269 (K) * [10]-gingerol 697(A) 523 (E) 327 (I) 312 (J) 269 (K) * 541(D) 523 (E) 327 (I) 312 (J) 269 (K) * 385 (F) 370 (H) 327 (I) 312 (J) 269 (K) Note: * includes ions 252 (L), 241 (M), and 213 (N) 155
156
Table 3. Ions formed in direct infusion (–)ESI-MSn experiments with [6]-gingerol, [8]- gingerol, and [10]-gingerol in Agilent LC-MSD-Trap-SL ion trap mass spectrometers.
MS MS2 MS3 MS4 275(P) [6]-gingerol 293(O) 193(Q) 178 (R) 135 (T) 99(S) 303(P) [8]-gingerol 321(O) 193(Q) 178 (R) 135 (T) 127(S) 331(P) [10]-gingerol 349(O) 193(Q) 178 (R) 135 (T) 155(S)
157
Table 4. Parent and daughter ion masses and their elemental compositions determined by high resolution and accurate mass FT-ICR measurements for negatively charged precursor and fragment ions of [6]-gingerol, [8]-gingerol and [10]-gingerol. Ions are drawn in scheme 2.
Ion Nominal Mass Measured Mass Calculated Mass Molecular Formula [6]-gingerol
O 293 293.1754 293.1753 C17H25O4
Q 193 193.0866 193.0865 C11H13O3
S 99 99.0815 99.0810 C6H11O [8]-gingerol
O 321 321.2065 321.2066 C19H29O4
Q 193 193.0865 193.0865 C11H13O3
S 127 127.1128 127.1123 C6H11O [10]-gingerol
O 349 349.2378 349.2379 C21H33O4
Q 193 193.0870 193.0865 C11H13O3
158
Figure Captions
Fig. 1. Daughter ion spectra of [6]-gingerol (a), [8]-gingerol (b) and [10]-gingerol (c),
from negative ion (–)ESI-MS/MS measurements from ThermoElectron LCQ Advantage
ion trap mass spectrometer coupled to HPLC. Daughter ion labels correspond to fragments depicted in Scheme 1.
Fig. 2. Daughter ion spectra of [6]-gingerol (a), [8]-gingerol (b), [10]-gingerol (c), and
[12]-gingerol (d) from negative ion (–)ESI-MS/MS measurements from Agilent LC-
MSD-Trap-SL ion trap mass spectrometer coupled to HPLC. Daughter ion labels
correspond to fragments depicted in Scheme 2.
Scheme 1. (–)ESI fragmentation of [6]-gingerol (1, n=4), [8]-gingerol (2, n=6), and [10]-
gingerol (3, n=8) obtained from a ThermoElectron LCQ Advantage ion trap mass
spectrometer.
Scheme 2. (–)ESI fragmentation of [6]-gingerol (1, n=4), [8]-gingerol (2, n=6), [10]-
gingerol (3, n=8), and [12]-gingerol (4, n=10) obtained from an Agilent LC-MSD-Trap-
SL ion trap mass spectrometer.
Scheme 3. (+)ESI fragmentation of [6]-gingerol (1, n=4), [8]-gingerol (2, n=6), and [10]- gingerol (3, n=8) obtained from ThermoElectron LCQ Advantage, ThermoElectron LCQ
Classic, and Agilent LC-MSD-Trap-SL ion trap mass spectrometers.
159
Fig. 1. Daughter ion spectra of [6]-gingerol (a), [8]-gingerol (b) and [10]-gingerol (c),
from negative ion (–)ESI-MS/MS measurements from ThermoElectron LCQ Advantage
ion trap mass spectrometer coupled to HPLC. Daughter ion labels correspond to fragments depicted in Scheme 1.
160
Fig. 2. Daughter ion spectra of [6]-gingerol (a), [8]-gingerol (b), [10]-gingerol (c), and
[12]-gingerol (d) from negative ion (–)ESI-MS/MS measurements from Agilent LC-
MSD-Trap-SL ion trap mass spectrometer coupled to HPLC. Daughter ion labels correspond to fragments depicted in Scheme 2.
161
Scheme 1. (–)ESI fragmentation of [6]-gingerol (1, n=4), [8]-gingerol (2, n=6), and [10]- gingerol (3, n=8) obtained from a ThermoElectron LCQ Advantage ion trap mass spectrometer.
162
Scheme 2. (–)ESI fragmentation of [6]-gingerol (1, n=4), [8]-gingerol (2, n=6), [10]- gingerol (3, n=8), and [12]-gingerol (4, n=10) obtained from an Agilent LC-MSD-Trap-
SL ion trap mass spectrometer.
163
Scheme 3. (+)ESI fragmentation of [6]-gingerol (1, n=4), [8]-gingerol (2, n=6), and [10]- gingerol (3, n=8) obtained from ThermoElectron LCQ Advantage, ThermoElectron LCQ
Classic, and Agilent LC-MSD-Trap-SL ion trap mass spectrometers.
164
APPENDIX C
Manuscript “Analysis of curcuminoids by positive and negative electrospray ionization
and tandem mass spectrometry”, submitted to Rapid Communication in Mass
Spectrometry.
165
TITLE: Analysis of curcuminoids by positive and negative electrospray ionization and
tandem mass spectrometry
Hongliang Jiang1–4, Arpad Somogyi5, Barbara N. Timmermann1,6, and David R. Gang1–3*
1Arizona Center for Phytomedicine Research, 2Department of Plant Sciences, 3Bio5
Institute, 4Department of Pharmaceutical Science, 5Department of Chemistry,
6Department of Pharmacology and Toxicology, and University of Arizona, Tucson,
Arizona, 85721, USA
*Corresponding author:
David R. Gang
Department of Plant Sciences and Bio5 Institute, University of Arizona, Tucson, AZ
85721-0036, USA
Tel: 520-621-7154
Fax: 520-621-7186
email: [email protected]
RUNNING TITLE: ESI-MS/MS Analysis of Curcuminoids
166
ABSTRACT
The curcuminoids are a group of diarylheptanoid molecules that possess important
pharmacological activities, particularly acting as anti-inflammatory agents. The purpose
of this study was to study the fragmentation behavior of the three major curcuminoids in
ion trap LC/MS/MS. Both positive and negative mode electrospray ionization in tandem
and multidimensional MSn experiments in quadrupole ion trap instruments and high
resolution and accurate mass MS and MS/MS experiments in a Fourier transform ion cyclotron resonance mass spectrometer were used to elucidate the fragmentation behavior of these compounds. These experiments, performed with four separate instruments from three manufacturers, yielded essentially the same fragmentation results in all cases for all three curcuminoids. Major and diagnostic fragment ions were identified and their origins proposed. In addition, tautomerism between enol and keto isomers of the curcuminoids, leading to deuteron exchange, was observed in the MS experiments and confirmed by 1H-
NMR.
KEY WORDS: curcumin, turmeric, electrospray tandem mass spectrometry, Fourier transform ion cyclotron resonance mass spectrometry
167
INTRODUCTION
The curcuminoids, bisdemethoxycurcumin (1), demethoxycurcumin (2), and curcumin
(3) (Fig. 1), belong to the group of compounds called the diarylheptanoids 1 and include a
β-diketone moiety. The curcuminoids are the major compounds responsible for the
yellow color of Curcuma longa (turmeric) rhizomes. 2 In addition to their role the
colorants of curries, these compounds have been reported to have anti-inflammatory, anti-arthritic, anti-oxidant, anti-allergic, anti-bacterial, anti-tumor, anti-coagulant, anti-
spasmodic, anti-parasitic, and anti-mutagenic properties. 3-7 For these reasons, turmeric
has been widely used as a food additive, condiment, and health food.
Analysis of the curcuminoids in turmeric samples is crucial to determine the
quality of the raw material or its processed products. Although gas-chromatography
coupled with mass spectrometry (GC/MS) has been used quite often to analyze turmeric samples, it can not be used to determine curcuminoids due to their low volatility and thermally labile properties. 8, 9 A variety of methods including high performance liquid
chromatography (HPLC), and its coupling with mass spectrometry (LC/MS), and
capillary electrophoresis (CE) have been used for the characterization of these
compounds in turmeric. 9-12 Among these methods, LC/MS can be used to determine
even trace amounts curcuminoids in foods or in other complex matrices and provide fast
and accurate analysis as an on-line technique.
LC/MS is rarely applied for full structure characterization except to provide the
molecular mass of the different constituents. Further characterization using LC/MS/MS,
168 where fragmentation patterns can provide additional structural information, can be used to produce firm evidence of identity compounds of interest. In addition, LC/MS/MS can be used for very selective quantitation of compounds of interest when selected fragment ions are monitored in the so-called selective ion monitoring (SIM) mode. This technique not only significantly increases the reliability of quantitative analysis but also improves the sensitivity (detection limit) considerably. Moreover, if curcuminoids, as a group, have the same fragmentation behavior in LC/MS/MS, then this analytical method could be used not only to identify and quantify known curcuminoids but also to identify unknown curcuminoids in extracts from turmeric or related plant material. However, the fragmentation behavior, i.e., the main fragmentation processes of the protonated/deprotonated molecules of the curcuminoids using LC/MS/MS has not been reported in detail.
The purpose of this study was to study the fragmentation behavior of the three major curcuminoids by ion trap LC/MS/MS, using instruments from different manufacturers so that general fragmentation principles, and not machine-specific idiosyncrasies, could be identified. High resolution and accurate mass spectrometry (FT-
ICR/MS/MS) was also used to confirm the fragmentation behavior of these compounds.
169
EXPERIMENTAL
Chemicals and reagents
Authentic standards of curcumin, demethoxycurcumin, and bisdemethoxycurcumin were
obtained from Dr. S. Jolad, Arizona Center for Phytomedicine Research, University of
Arizona, Tucson, AZ. Acetonitrile and methanol were from Burdick & Jackson
(Muskegon, MI, USA). Formic acid was from J. T. Baker (Mallinkrodt Baker, Inc.,
Phillipsburg, NJ, USA). Ammonium formate was from Fisher Scientific (Fair Lawn, NJ,
USA). Deionized water was re-distilled.
LC/ESI-MS/MS analysis of curcuminoids
LC/ESI-MS/MS analysis of curcuminoids was performed on two separate ion trap mass
spectrometers: i) an Agilent 1100 HPLC system coupled to an Agilent LC-MSD-Trap-SL ion trap mass spectrometer (Palo Alto, CA, USA), and ii) a ThermoFinnigan Surveyor
MS HPLC coupled to a ThermoFinnigan LCQ Advantage ion trap (San Jose, CA, USA).
LC separation of curcuminoids—Pure curcuminoids were dissolved in MeOH or turmeric samples were extracted with MeOH. For both instruments, the same column and elution parameters were used to achieve baseline separation of the eluted curcuminoids, with
almost identical elution volumes. For the LCQ Advantage instrument, a 1:10 split after
the photo diode array detector yielded a 25 µl/min flow rate into the ion trap that reduced
background solvent ions and greatly enhanced sensitivity. Column: Discovery® HS C18,
3 µm, 15 cm × 2.1 mm (Supelco, Bellefonte, PA, USA); Guard column: Discovery® HS
170
C18, 3 µm, 2 cm × 2.1 mm (Supelco); Mobile phase: (A) buffer (5mM ammonium
formate, 0.1% formic acid, in ddH2O) and (B) acetonitrile; Gradient (in buffer A): 0-2
min, 5% B; 2-57 min, 5-100% B; 57-60 min, 100% B; 60-65 min, 100-5% B; 65-75 min,
5% B. Flow rate: 0.25 ml/min; temperature, 40°C; Injection volume 5 µl.
MS and MSn parameters for Agilent LC-MSD-Trap-SL—The acquisition parameters for
positive and negative mode were: drying N2 temperature, 350°C, 10 l/min; nebulizer
pressure 60 psi; HV capillary 4500 V; HV end plate offset -500 V; capillary current 65.9
nA (positive mode), 62.3 (negative mode); end plate current 1482.7 nA (positive mode),
1378.7 (negative mode); capillary exit RF amplitude 99.3 V (positive mode), -99.3 V
(negative mode); skimmer 40.0 V (positive mode), -40.0 V (negative mode); mass range
measured: 50-900 m/z. These were the optimized parameters for the maximum
transmission of the curcuminoid-derived ions.
MS and MSn parameters for ThermoFinnigan LCQ Advantage—The acquisition
parameters for positive and negative mode were: sheath gas flow 29 (positive), 36
(negative) (in arbitrary units); aux/sweep gas flow 6 (positive), 0 (negative); source
voltage 5 kV (positive), 4.5 kV (negative); source current 80 µA (positive and negative); capillary voltage 36 V (positive), -32 V (negative); capillary temperature 270 °C; tube lens offset 45 V (positive), -25 V (negative); collision gas pressure ca. 10-5 torr; Q-value
25; mass range measured: 100-1000 m/z. These were the optimized parameters for the
maximum transmission of the curcuminoid-derived ions.
171
Direct infusion ESI-MS/MS analysis of unlabeled and deuterated curcuminoids
Curcuminoids 1, 2, and 3 were analyzed by direct infusion ESI on two separate ion trap
instruments: a Finnigan LCQ Classic ESI-ion trap (San Jose, CA, USA) and the same
ThermoFinnigan LCQ Advantage ESI-ion trap used in the LC/MS/MS experiments, but
with the HPLC disconnected. Both instruments gave essentially the same results for both
negatively and positively charged ions. For this analysis, the curcuminoids were
dissolved in MeOH: H2O (1:1) at a concentration range of 50-80 µM. For solution phase
H/D exchange experiments, the compounds were dissolved at the same concentration range in CD3OD: D2O (1:1) for positive mode. A few drops of NH4OH were added (per
ml of solution) to the same solvent system to improve the ionization in negative mode for
ESI. The infusion flow rate in all cases was 10 µl/min. Instrument set up for the
Thermoelectron (Finnigan) LCQ classic instrument was as follows: needle voltage: 4.5
kV (positive), -4.0 V (negative), sheath gas flow: 60 (positive), 30 (negative) (in arbitrary
units); other parameters were optimized for the instrument and did not differ significantly
from those described for the LCQ Advantage as described above.
FT-ICR analysis of curcuminoids
High resolution and accurate mass measurements were carried out on an IonSpec 4.7 T
FT-ICR instrument (Lake Forest, CA). Ions were generated using an Analytica (Branford,
CT) second generation electrospray (ESI) source by using diluted solutions of those that
were used in the Thermoelectron (Finnigan) LCQ Classic studies. Direct infusion of
diluted (10-20 µM in MeOH:H2O 1:1) individual curcuminoids with a flow rate of 2
172
µl/min was applied to generate positively charged ions by ESI For internal calibration,
first, a mixture of G2R and G3R peptides was used. With reference to the protonated
molecules of the internal standards (289.1624 and 346.1839, respectively), the elemental
composition of the protonated molecules of compounds 1 – 3 was determined. Tandem
MS/MS fragmentation was achieved by using the sustained off-resonance irradiation
(SORI) technique with N2, as collision gas. The SORI excitation time was 500 ms with a
-8 N2 pressure of ca. 5 × 10 torr. First, the selected precursor ion was used to calibrate the
spectra. The accuracy of mass measurements was improved by using a well-defined
fragment in the lower mass region as indicated in Table 1.
1H-NMR analysis of unlabeled curcuminoids
1 Curcuminoids 1, 2, and 3 were dissolved in CD3OD for H-NMR (500 MHz) experiments
performed at room temperature in a Bruker AVANCE DRX500 spectrometer.
RESULTS AND DISCUSSION
The curcuminoids can exist in solution as the keto or the enol forms, depending on the
conditions (temperature, pH, ions present, etc.). Consideration of the tautomerism of
curcuminoids between the keto and enol forms is necessary to understand the ionization
and fragmentation mechanisms of these compounds, even though our purposes were not
targeted at investigating this phenomenon. It is well known that most β-dicarbonyls are in
173
equilibrium with significant amounts of the corresponding enol forms, regardless of
whether they are present as solids, as liquids, in solution or in the gas phase. The kinetics
and mechanisms of enolization and ketonization of many such compounds have been
studied in detail for some time,13 and the purpose of this investigation was not to rehash
this extensive body of work. Instead, we found that an understanding of the phenomenon
of keto-enol tautomerism was required to understand the process of ionization of the curcuminoids in LC-MS experiments. Indeed, the temperature and surface catalytic
activity of ion sources have been reported to be major factors in determining the extent of
keto-enol isomerization in the production of enol or keto ions in such investigations.14
LC/(–)ESI-MS/MS analysis of curcuminoids 1, 2, and 3
The fragmentation of curcuminoids 1 – 3 in negative mode LC/ESI-MS/MS occurs through just a few fragmentation channels producing simple MS/MS spectra (Figure 2), regardless of instrument used; we obtained essentially the same results from the Agilent and the ThermoFinnigan instruments. From the total ion chromatograms (TICs) of methanolic extracts of turmeric samples or of standard compounds dissolved in methanol, we detected the [M-H]- ions (m/z of 307.1, 337.3, and 367.1, respectively) of curcuminoids 1, 2, and 3 at elution volumes of 8.275 ml, 8.475 ml, and 8.625 ml using
the HPLC gradient profile described in Materials and Methods. As shown in Fig. 2, ion B
(m/z 187.1 or 217.1) is the base peak in the MS/MS spectra of deprotonated [M-H]- curcuminoids. The formation of product ion B can proceed via a β-hydrogen shift (shown in Scheme 1) to the double bond, leading to the loss of a neutral moiety. This reaction has
174
also been demonstrated in EI-MS of curcumin.15 The presence of the peak at m/z 217.1 in
the MS/MS spectra of 2 and 3 is due to the presence of the methoxy group in one or two aromatic rings (i.e., the mass shifts by 30 daltons with reference to m/z 187.1). Another characteristic product ion D (m/z 143 and or m/z 173) can be formed from product ion B by rearrangement and loss of one CO2 (Scheme 1). We also observed that ion D is the
major fragment ion in the MS3 spectrum of ion B by direct infusion negative mode ESI-
MSn (spectrum not shown but see discussion below). Product ions C (m/z 145 and/or m/z
175), although very minor in the negative mode LC/MS/MS analysis, were also observed and their relative intensity varied slightly from instrument to instrument. These ions can be formed by a ring closure reaction from the [M-H]- ion (Scheme 1).
The deprotonation of the curcuminoids could theoretically occur at several
positions on the keto and enol forms of the molecules, as shown in Fig. 1. However, the
diketone form, with the negative charge residing on the oxygen atom on one of the
aromatic rings, was the predominant precursor for negative ionization of the
curcuminoids, as determined by the fragmentation pattern of curcuminoids in (−)ESI-
MS/MS analysis, where base peak B was observed, resulting from a β-hydrogen shift. In
contrast, a McLafferty Rearrangement would require the exchange of protons between 3-
O and 5-O and cleavage between carbons 3 and 4 or carbons 4 and 5. This cannot occur
with the curcuminoids, however, because of the presence of a double bond at these
positions in the corresponding enols. And, we did not observe any ions that would be
attributable to such a rearrangement. We have, however, observed McLafferty
rearrangements with related diarylheptanoids and with the gingerols, which do not have
175
the corresponding diketone moiety, but instead possess a β-keto hydroxy moiety, which does not have a double bond between either carbons 3 and 4 or 4 and 5.
Direct infusion (–)ESI-MS/MS analysis of unlabeled and deuterated curcuminoids 1,
2, and 3
We also studied the fragmentation behavior of the curcuminoids in the negative mode (–
)ESI/MS/MS by direct infusion. The MS/MS spectra obtained by fragmentation of the
curcuminoids in the (–)ESI-MS/MS experiments (data not shown) were exactly the same
as those from LC-(–)ESI-MS/MS. These experiments were performed in both the Agilent
and the ThermoFinnigan instruments, yielding essentially the same results.
In addition, the fragmentation behavior of deuterated curcuminoids was investigated using (–)ESI-MS/MS and (–)ESI-MSn experiments to confirm the
fragmentation behavior of unlabeled curcuminoids. Because we observed no instrument
dependent differences in the experiments described above, we only performed these
solution phase H/D exchange experiments in the Finnigan LCQ Classic instrument.
The keto-enol tautomerism of the β-ketide of the curcuminoids in deuterated
solvent would be expected to lead to H→D exchange of two protons at C-4. Another two
protons on the aromatic rings’ p-hydroxyl groups could also be easily exchanged with
deuteriums. Exchange with the aromatic protons of the rings, although possible, is not
very favorable, and would not be expected to be easily observed. As shown below, this
was the case. Thus, there are four protons in the curcuminoids that could readily be
exchanged with deuteriums. For (–)ESI mass spectra, however, we are only able to
176
observe the exchange of three protons with deuteriums, due to the formation of
deprotonated [M-H]- ions by the loss of a proton/deuterium from one of the two aromatic
(phenolic) hydroxyl groups during ionization in (–)ESI mode. This was shown to be the
case by (–)ESI-MS/MS experiments where we dissolved the curcuminoids in CD3OD:
D2O (1:1, with a small amount of NH4OH) and then analyzed them quickly in the LCQ
instrument. The samples were only allowed to sit for a few minutes in the deuterated
solvent prior to (–)ESI-MS/MS analysis. In these experiments, the [M-H(D)]- (parent)
ions were analyzed for curcuminoids 1, 2, and 3. For curcuminoid 1, isotopic forms at
m/z 308 (one proton-deuteron exchanged), 309 (two proton-deuteron exchanged), 310
(three proton-deuteron exchanged) were observed from full mass spectra. For curcuminoids 2 and 3, isotopic forms at m/z 338, 339, and 340 and at m/z 368, 369, and
370, respectively, were observed instead. As shown above, negative ionization of the
curcuminoids in these instruments occurs by abstraction of a proton from an aromatic ring hydroxyl group. Thus, in addition to the exchange of the protons on the p-hydroxyl groups of the aromatic rings, at least two other hydrogen atoms were readily exchanged by deuteriums. This suggested that keto-enol tautomerism of the β-ketide of the curcuminoids is an active process in this solvent system at room temperature. Other experiments were carried out to further confirm these observations.
In the first of these, (–)ESI-MSn experiments in deuterated solvents demonstrated that the fragmentation behavior of the deuterated curcuminoids is same as that of unlabeled curcuminoids. Furthermore, the m/z of product ions of the deuterated forms supports not only the proposed product ions of the unlabeled compounds illustrated in
177
Scheme 1, but also the presence of active keto-enol tautomerism. For example, product
ion B (m/z 187/217) contains only one proton at C-4, and no proton on the p-hydroxyl
group (which possesses the negative charge). The isotopic product ion B was detected at
m/z 188 for deuterated curcuminoid 1, at m/z 188 and 218 for deuterated curcuminoid 2, and at m/z 218 for deuterated curcuminoid 3, independent of the selection of one, two or tree H/D exchanged [M-H]- precursor ions (see Scheme 2). This suggests that the exchangeable proton at C-4 of the non-fragmented curcuminoids is relatively stable after negative ionization in the ESI source and fragmentation in the ion trap. The only other deuterium that could be on this fragment would be on the aromatic hydroxyl group, and it was lost during ionization. In another (–)ESI-MSn experiment, deuterated product ions D,
derived from deuterated product ions B (Scheme 1), were observed. These ions also have
one deuterium, (m/z 144 for deuterated curcuminoid 1, m/z 144 and 174 for deuterated
curcuminoid 2, and m/z 174 for deuterated curcuminoid 3). Again, the difference of 30
Daltons is attributed to the methoxy functional group on one or both aromatic rings. The
decarboxylation required to form D ions would not be expected to lead to the loss of the
deuterium on C-4. In a final MS/MS experiment, product ions C obtained from
deuterated [M-H]- precursor ions A did not contain any of the exchanged deuteriums.
This demonstrates that the deuteriums were indeed present at C-4 in A or on the hydroxyl group of the aromatic ring, and not at any other position in the molecule.
The enol-dione equilibrium has been reported as the general feature for neutral β- diketones, even in the gas phase.15 Although a considerable fraction of the molecules may
have one of the keto groups in the tautomeric enol form at the time of ionization, it is
178
more rational for us to explain the fragmentation behavior of curcuminoids in (–)ESI
mode by invoking the keto form. In fact, several of the ions, especially the base peak at
m/z 187/217, can only be formed from the diketone form of the curcuminoids. Thus, it
appears that the diketone form of the curcuminoids is the dominant form that is ionized in
a manner that allows it to be observed in (–)ESI ion trap mass spectrometry.
LC/(+)ESI-MS/MS and direct infusion (+)ESI-MSn analysis of curcuminoids 1, 2,
and 3
The fragmentation patterns of curcuminoids 1, 2, and 3 in (+)ESI mode are somewhat
similar to each other, but dramatically different from the patterns observed for (–)ESI
mode. In particular, the p-hydroxyl groups and the aromatic rings do not play a direct role
in either ionization or in ion stabilization in (+) mode. Instead, ionization, rearrangements
and new bond formations occur in the heptanoid portion of the molecules. The probable
positions for protonation are indicated in Fig. 1. However, only the precursor ion with the
diketone form can rationalize the fragmentation behavior in (+)ESI-MS/MS.
From the total ion chromatograms (TICs) of methanolic extracts of turmeric samples or of standard compounds dissolved in methanol, we detected the [M+H]+ ions
at m/z of 309, 339, and 369 of curcuminoids 1, 2, and 3, respectively, at the same elution
volumes as was observed in negative mode LC-ESI-MS/MS. Representative (+)ESI
spectra for curcuminoids 1, 2, and 3 in these runs are shown in Fig. 3, and an outline of
the corresponding fragmentation of the protonated molecules is shown in Scheme 3.
179
Eight distinguishing product ions (F, G, H, I, J, K, L, and M) were detected for
all three curcuminoids in LC-(+)ESI-MS/MS. These ions all originate from parent ion E
[M+1], which resulted from protonation of one of the keto groups on the heptanoid chain
(reaction 1, Scheme 3). Low abundance product ions, F, (m/z 291, 321, and 351), were
formed by loss of a H2O molecule from parent ions E (reaction 2, Scheme 3). Product
ions G (m/z 239, 269, and 299) were formed as shown in reaction 3 (Scheme 3) by rearrangement of E and loss of a diketocyclopropane (or equivalent) moiety. Product ions
H (m/z 225, 255, and 285) were formed by rearrangement of E and loss of a 1- hydroxy,3-ketocyclobutene moiety (reaction 4, Scheme 3). Product ions I (m/z 215,
215/245, and 245) were formed by rearrangement of E and neutral loss of one aryl group and a hydrogen atom (reaction 5, Scheme 3). Product ions J (m/z 199, 229, and 259) were formed by rearrangement of E and neutral loss of a 1-hydroxy,5-ketocyclo-1,3-hexadiene moiety (reaction 6, Scheme 3). Product ions K (m/z 189, 189/219, and 219) were formed by H migration on E and neutral loss of an arylethene moiety (reaction 7, Scheme 3).
Product ions L (m/z 147, 147/177, and 177) were formed by 3,4 bond cleavage in rearranged E and neutral loss of a 1-aryl,3-hydroxy-1,3-butadiene moiety (reaction 8,
Scheme 3). Product ions M (m/z 145, 145/175, and 175) were formed by an oxo- migration in rearranged E and neutral loss of a 3-arylcarboxyprop-2-ene moiety (reaction
9, Scheme 3).
Direct infusion (+)ESI-MS/MS and (+)ESI-MS3 analysis of several of these major
ions also confirmed the predicted structures of the product ions. For example, product
ions L lost CO (28 Daltons), supporting the proposed structure of product ions L. In
180
addition, the structures of product ions H, I, and M, were supported in the same manner.
As discussed below (section on FT-ICR MS), accurate mass measurement of these
product ions also confirmed their molecular formulae.
Direct infusion (+)ESI-MS/MS analysis of deuterated curcuminoids 1, 2, and 3
The fragmentation behavior of deuterated curcuminoids in direct infusion (+)ESI-MS/MS
(data not shown) supported that of the unlabeled curcuminoids. In addition, we used this
approach to further investigate the keto-enol tautomerism of the curcuminoids. In this case, positively charged parent ions of deuterated curcuminoids would not be expected to lose one of the deuteriums as a result of ionization. Instead, the neutral molecules would gain only a single proton in non-deuterated solvent, assuming no additional proton exchange with the solvent. Thus, we dissolved deuterated curcumin 3 that resulted from the solution phase H/D exchange experiments described above in unlabeled H2O:MeOH
1:1 (mass of 368 Daltons), and lyophilized to remove the deuterated solvent. We then
recorded the parent ion isotope pattern via direct infusion (+)ESI-MS over time. As
shown in Fig. 4, we observed that m/z 372 is the major form during the first minute of
D/H exchange, although 373 is also very abundant. These represent three and four
deuteriums, respectively. Thus, we see that all four potentially labile protons on curcumin
3 were indeed exchanged by deuteriums while this compound was dissolved in deuterated
solvent. Once transferred to unlabeled solvent, however, these deuteriums rapidly
exchanged with the solvent (see Fig. 4). As mentioned above, after being in unlabeled
solvent for between 30 seconds and one minute, the major forms of deuterated curcumin
181
3 were triply and quadruply labeled (Figure 4A). By 2 to 2.5 minutes, however, m/z 372
(3 deuteriums) was no longer the major form, and m/z 373 (4 deuteriums) was greatly reduced (Figure 4B). By 7.5 minutes, the fraction that was only singly labeled (m/z 370) had increased dramatically (Figure 4C). And by 15 minutes a significant fraction has reverted back to the unlabeled form (m/z 369, Figure 4D). This confirms the results described above for the unlabeled to labeled conversion observed in (–)ESI, where four hydrogens (including the ionizing proton) were predicted to be especially labile.
FT-ICR analysis of curcuminoids 1, 2, and 3
To confirm the identity of the parent and product ions produced by (+)ESI in the Agilent and ThermoFinnigan ion trap instruments, we performed high resolution and accurate mass FT-ICR MS measurements of positively charged ions produced by curcuminoids 1,
2, and 3. The resulting ions, nominal masses, accurate masses, and corresponding molecular formulae and predicted masses are summarized in Table 1. The molecular formulae of product ions support our proposed structures of product ions. All but ions J and K were observed for all curcuminoids in the FT-ICR, and J and K were observed for at least one of the compounds. These results confirm the fragmentation data and fragment ion assignments deduced from direct infusion/ESI-MS/MS and LC/ESI-MS/MS for the curcuminoids. They also support the use of these MS and MS/MS techniques as highly sensitive and specific methods to identify curcuminoids in extracts from plant samples. In addition, the molecular weights and fragmentation pattern obtained by MS and MS/MS
182
measurements, respectively, can be used to characterize previously unknown
curcuminoids or similar compounds.
1H-NMR investigation of keto-enol tautomerism of curcuminoids 1, 2, and 3
To further confirm the MS-based conclusions regarding the keto-enol tautomerism of the
curcuminoids in solution, we also performed 1H-NMR experiments. We dissolved the
unlabeled curcuminoids 1, 2, and 3 in CD3OD just before beginning these experiments,
and we were able to initially observe the proton signal at C-4 of the enol form. The data
from the 1H-NMR spectra are summarized in Table 2. Interestingly, the proton signal at
C-4 of the enol form decreased noticeably within 2.5 hr and completely disappeared within 40 hr. This exchange was much slower than that observed in the ESI-MS/MS experiments. However, this would be expected considering the nature of the solvents
1 used for these two sets of experiments: 100% CD3OD for H-NMR and CD3OD:D2O
(1:1) for ESI-MS/MS. Thus, the H→D exchange phenomenon observed in the 1H-NMR
experiments supports the keto-enol tautomerism of curcuminoids in solution.
CONCLUSIONS
Ion trap-based ESI-MS/MS is a very sensitive and precise method for identifying curcuminoid compounds. Ion trap instruments from different manufacturers produced essentially the same fragmentation patterns, supporting similar ionization and
183
fragmentation mechanisms occurring in these instruments. High resolution and accurate
mass MS and MS/MS measurements in an ESI-FT-ICR instrument confirmed the
fragmentation assignments based on the fragment ions detected in the ion trap
instruments. In addition, solution phase H/D exchange with the perdeuterated solvents,
demonstrated that the curcuminoids undergo keto-enol tautomerism in polar solvents as was proven by both MS and 1H-NMR experiments. However, the fragmentation patterns
exhibited in ESI-MS indicate that the keto form is either more prominent in the gas phase
than the enol form, or it is more readily ionized.
ACKNOWLEDGMENTS
The authors acknowledge financial assistance from the National Science Foundation
Plant Genome Program, grant DBI-0227618 to D.R.G., and the National Institutes of
Health NCCAM/ODS, grants #5 P50 AT 000474-04 and 3 P50 AT 000474-03 S1 to
B.N.T. We also thank Brenda Jackson and Veronica Rodriguez for assistance with
LC/MS and Dr. Shivanand Jolad for authentic curcuminoid standards.
184
REFERENCES
1. Schröder J. Trends Plant Sci. 1997; 2: 373.
2. He X-G, Lin L-Z, Lian L-Z, Lindenmaier M. J. Chromatogr. A. 1998; 818: 127.
3. Sun XH, Gao CL, Cao WD, Yang XR, Wang EK. J. Chromatogr. A. 2002; 962:
117.
4. Shukla Y, Arora A, Taneja P. Mutat. Res-gen. Tox. En. 2002; 515: 197.
5. Sreejayan, Rao MNA. J. Pharm. Pharmacol. 1997; 49: 105.
6. Reddy ACP, Lokesh BR. Mol. Cell Biochem. 1992; 111: 117.
7. Ammon HPT, Wahl MA. Planta Med. 1991; 57: 1.
8. Richmond R, PomboVillar E. J. Chromatogr. A. 1997; 760: 303.
9. Hiserodt R, Hartman TG, Ho CT, Rosen RT. J. Chromatogr. A. 1996; 740: 51.
10. Lechtenberg M, Quandt B, Nahrstedt A. Phytochem. Analysis. 2004; 15: 152.
11. Inoue K, Hamasaki S, Yoshimura Y, Yamada M, Nakamura M, Ito Y, Nakazawa
H. J. Liq. Chromatogr. R. T. 2003; 26: 53.
12. Jayaprakasha GK, Rao LJM, Sakariah KK. J. Agr. Food Chem. 2002; 50: 3668.
13. Rappoport Z. The Chemistry of Enols. Wiley-Interscience: Chichester, 1990; 223-
278.
14. Turecek F, Havlas Z, Maquin F, Gaumann T. Helv. Chim. Acta 1986; 69: 683.
15. van Baar BLM, Rozendal J, van der Goot H. J. Mass Spectrom. 1998; 33: 319.
185
Table 1. Parent and product ion masses and their elemental compositions determined by high resolution and accurate mass FT-ICR measurements for positively charged precursor and fragment ions of curcuminoids 1, 2, and 3.
Nominal Mass Ion Measurement Mass Calculated Mass Elemental Composition Curcuminoid 1 E 309 309.1127* 309.1127* C19H17O4 F 291 291.1004 291.1021 C19H15O4 G 239 239.106 239.1072 C16H15O2 H 225 225.0909 225.0915 C15H13O2 I 215 215.0707 215.0708 C13H11O3 K 189 189.0548 189.0552 C11H9O3 L 147 147.0446* 147.0446* C9H7O2 M 145 145.0653 145.0653 C10H9O Curcuminoid 2 E 339 339.1233* 339.1233* C20H19O5 F 321 321.1135 321.1127 C20H17O4 G 269 269.1182 269.1178 C17H17O3 H 255 255.1023 255.1021 C16H15O3 I 245 245.0817 245.0814 C14H13O4 J 229 229.0868 229.0865 C14H13O3 L 177 177.0553* 177.0552* C10H9O3 L 147 147.0446 147.0446 C9H7O2 M 175 175.0761 175.0759 C11H11O2 Curcuminoid 3 E 369 369.1338* 369.1338* C21H21O6 F 351 351.1241 351.1232 C21H19O5 G 299 299.1291 299.1283 C18H19O4 H 285 285.1129 285.1127 C17H17O4 I 245 245.0812 245.0814 C14H13O4 J 259 259.0977 259.097 C15H15O4 L 177 177.0552* 177.0552* C10H9O3 M 175 175.076 175.0759 C11H11O2
* Ions used for internal calibration (see text for details)
186
Table 2. 1H NMR spectral data of curcuminoids 1, 2, and 3.
H 1 2 3 1 7.59 d (15.5) 7.60 d (15.5)a 7.59 d (15.8) 2 6.61 d (15.5) 6.65 d (15.5)b 6.65 d (15.8) 4 5.96 s* 5.98 s* 5.98 s* 6 6.61 d (15.5) 6.62 d (15.5)b 6.65 d (15.8) 7 7.59 d (15.5) 7.59 d (15.5)a 7.59 d (15.8) 2' 7.51 dd (7.8, 2.0) 7.24 d (2.0) 7.23 d (1.8) 3' 6.83 dd (7.8, 2.0) - - 5' 6.83 dd (7.8, 2.0) 6.83 d (7.8) 6.84 d (8.3) 6' 7.51 dd (7.8, 2.0) 7.13 dd (7.8, 2.0) 7.12 dd (8.3, 1.8) 2'' 7.51 dd (7.8, 2.0) 7.51 dd (7.0, 2.0) 7.23 d (1.8) 3'' 6.83 dd (7.8, 2.0) 6.84 dd (7.0, 2.0) - 5'' 6.83 dd (7.8, 2.0) 6.84 dd (7.0, 2.0) 6.84 d (8.3) 6'' 7.51 dd (7.8, 2.0) 7.51 dd (7.0, 2.0) 7.12 dd (8.3, 1.8) OMe-3' - 3.93 s 3.93 s OMe-3'' - - 3.93 s
J values in parentheses expressed in Hz.
* Proton signal assignable to an enol form. a, b Chemical shifts with the same letter may be interchanged
187
Figure Legends
Figure 1. Keto-enol tautomerism of curcuminoids.
Figure 2. Product ion spectra of curcuminoids 1, 2, and 3 from negative ion (–)ESI-
LC/MS/MS measurements. Product ion labels correspond to fragments depicted in
Scheme 1.
Figure 3. Product ion spectra of curcuminoids 1, 2, and 3 from positive ion (+)ESI-
LC/MS/MS measurements. Product ion labels correspond to fragments depicted in
Scheme 3.
Figure 4. Deuterated curcuminoids undergo rapid D→H exchange with unlabeled solvent
(MeOH:H2O), resulting in generation of less and less labeled curcuminoids as time goes
on. Here, deuterated curcumin 3 was dissolved with MeOH:H2O (1: 1) at room
temperature and monitored continuously by direct infusion (+) ESI in the Finnigan LCQ
Classic ion trap mass spectrometer (a)-(d). Spectrum of unlabeled curcumin 3 is shown in
(e).
Scheme 1. (–)ESI fragmentation of curcuminoids 1, 2, and 3.
188
Scheme 2. Fragmentation graphs obtained from (–)ESI MS/MS measurements for
curcuminoids 1, 2, and 3. (a) non-deuterated curcuminoids; (b) singly deuterated
curcuminoids; (c) doubly deuterated curcuminoids; (d) triply deuterated curcuminoids.
Note: * means ion is absent.
Scheme 3. (+)ESI fragmentation of curcuminoids 1, 2, and 3.
189
H H # # * O O O# O* O # O 1 1 1 R1 3' 7 3'' R2 R1 3' 7 3''R2 R1 3'7 3'' R2 3 * 5 3 5 3 5 HO 4' 4'' OH HO 4' 4'' OH HO 4' 4''OH * * * * * *
keto form enol form
(1): R1=H, R2=H (2): R1=OCH3, R2=H (3): R1=OCH3, R2=OCH3
Figure 1. Keto-enol tautomerism of curcuminoids. Note: * indicates the deprotonation position; # indicates protonation position.
190
Figure 2. Product ion spectra of curcuminoids 1, 2, and 3 from negative ion (–)ESI-
LC/MS/MS measurements. Product ion labels correspond to fragments depicted in
Scheme 1.
191
Figure 3. Product ion spectra of curcuminoids 1, 2, and 3 from positive ion (+)ESI-
LC/MS/MS measurements. Product ion labels correspond to fragments depicted in
Scheme 3.
192
Figure 4 Deuterated curcuminoids undergo rapid D→H exchange with unlabeled solvent (MeOH:H2O), resulting in generation of less and less labeled curcuminoids as time goes on. Here, deuterated curcumin 3 was dissolved with MeOH:H2O (1: 1) at room temperature and monitored continuously by direct infusion (+) ESI in the Finnigan LCQ Classic ion trap mass spectrometer (a)-(d). Spectrum of unlabeled curcumin 3 is shown in (e).
193
OH
O O R1 O C R1 R2 -H O O H C R2 1 H 2 HO OH O 308 (1) O R2 O C 338 (2) D A R2 145 (1) 143 (1) 368 (3) 143, 173 (2) 307 (1) 145, 175 (2) 173 (3) 337 (2) 175 (3) 367 (3) 4
O HO O O O O O H H H C R2 C R2 H R1 R2 3 C R2 H O O O O O O 307 (1) B 187 (1) 337 (2) 187, 217 (2) 367 (3) 217 (3)
Scheme 1. (–)ESI fragmentation of curcuminoids 1, 2, and 3.
194
307 1 (337) 2 (367) 3
145 143 * 187 * * (143) (217) (187) (175) (173) (145) (*) (*) (217) (*) (175) (173) (a)
308 1 (338) 2 3 (368)
* 188 * * 145 144 (144) (218) (188) (175) (174) (145) (*) (*) (218) (*) (175) (174) (b)
309 1 (339) 2 (367) 3
* 188 * * 145 144 (144) (218) (188) (175) (174) (145) (*) (*) (218) (*) (175) (174) (c)
310 1 (340) 2 (370) 3
145 144 * 188 * * (144) (218) (188) (175) (174) (145) (*) (*) (218) (*) (175) (174) (d)
Scheme 2. Fragmentation graphs obtained from (–)ESI MS/MS measurements for curcuminoids 1, 2, and 3. (a) non-deuterated curcuminoids; (b) singly deuterated curcuminoids; (c) doubly deuterated curcuminoids; (d) triply deuterated curcuminoids. Note: * means ion is absent.
O O
Ar + Ar O OH HO Ar 1 2 O OH 1 + F 291 ( ) Ar 1 Ar + Ar Ar + Ar2 2 1 2 1 9 321 (2) M 145 (1) 351 (3) 145, 175 (2 -H2O 175 (3)
+ 2 O O Ar O O O OH O OH Ar 1 + H+ 1 Ar H O 1 Ar Ar + OH Ar + Ar1 Ar2 1 Ar1 2 Ar1 + Ar2 + 2 2 Ar 308 (1) E 309 (1) 2 OH O 338 (2) 339 (2) O 368 (3) 4 369 (3) Ar1H 5 6 OH + O O OH OH O OH O OH + O Ar + Ar Ar + Ar Ar + Ar 1 2 2 2 Ar1 Ar2 2 1 + H Ar2 H H 225 (1) J 199 (1) Ar OH I 215 (1) 255 (2) 1 Ar2 229 (2) 8 215, 245 (2) 285 (3) Ar 259 (3) 7 1 245 (3) Ar2 C=O O 3 O OH CH2=C=O Ar + 2 HO HO + L 147 (1) Ar1 + 147, 177 (2) MeO H Ar K 177 (3) 1 Ar2 R1 R2 G 189 (1) 239 (1) 189, 219 (2) 269 (2) Bisdemethoxycurcumin (1) Ar1 = Ar2 = R1 219 (3) 299 (3) Demethoxycurcumin (2) Ar1=R1, Ar2=R2 ; Ar1=R2,
Curcumin (3) Ar1 = Ar2 = R2
Scheme 3. (+)ESI fragmentation of curcuminoids 1, 2, and 3. 195
196
APPENDIX D
Entitled “Characterization of gingerol-related compounds in ginger rhizome (Zingiber
officinale Rosc.) by high-performance liquid chromatography/electrospray ionization
mass spectrometry”, this manuscript has been submitted to Rapid Communication in
Mass Spectrometry.
197
TITLE: Characterization of gingerol-related compounds in ginger rhizome (Zingiber officinale Rosc.) by high-performance liquid chromatography/electrospray ionization mass spectrometry
Hongliang Jiang1-4, Anikó M. Sólyom1,5, Barbara N. Timmermann1,5, and David R.
Gang1-3*
1Arizona Center for Phytomedicine Research, 2Department of Plant Sciences, 3Bio5
Institute, 4Department of Pharmaceutical Science, 5Department of Pharmacology and
Toxicology, and University of Arizona, Tucson, Arizona, 85721, USA
*Corresponding author:
David R. Gang
Department of Plant Sciences and Bio5 Institute, University of Arizona, Tucson, AZ
85721-0036, USA
Tel: 520-621-7154
Fax: 520-621-7186
email: [email protected]
RUNNING TITLE: LC-ESI-MS/MS Analysis of gingerol-related compounds in ginger
rhizome
198
ABSTRACT
This study sought to determine the utility of LC-ESI-MS/MS coupled with diode array detection in identifying gingerol-related compounds from crude extracts of ginger rhizome. The fragmentation behaviors of compounds in both (−)- and (+)ESI-MS/MS
were used to infer and confirm the chemical structures of several groups of compounds,
including the gingerols, methylgingerols, gingerol acetates, shogaols, paradols,
gingerdiols, mono- and diacetyl gingerdiols, and dehydrogingerdiones. Diode array detection at different wavelengths was used to confirm MS/MS based identification. In total, 31 gingerol-related compounds were identified from the methanolic crude extracts of fresh ginger rhizome in this study. Three of these compounds were found to be new compounds. This study demonstrated that LC-ESI-MS/MS is a powerful on-line tool for identification of gingerol-related compounds, especially for thermally labile compounds that cannot be readily detected by GC/MS analysis.
KEY WORDS: Zingiber officinale; ginger; gingerol; LC-ESI-MS/MS
199
INTRODUCTION
The rhizome of ginger (Zingiber officinale, Rosc.) Zingiberaceae has long served
culinary and medicinal uses [1]. Two major groups of compounds including gingerol-
related compounds and diarylheptanoids have been reported as bioactive components from this plant [2,3]. Gingerol-related compounds, comprising distinct groups
(homologous series) that are differentiated by the length of their unbranched alkyl chains,
have recently gained attention in a variety of biological activity studies [4-7]. Analytical
tools are therefore needed to characterize this group of compounds from diverse sources
including plant material or processed products.
Many analytical methods including gas-chromatography coupled with mass
spectrometry (GC/MS), high performance liquid chromatography (HPLC), and its
coupling to mass spectrometry (LC-MS), thin layer chromatography (TLC) and capillary
electrophoresis (CE) have been used to characterize gingerol-related compounds in
ginger [8-11]. Among these methods, GC/MS has been used quite often to analyze ginger
samples. Nevertheless, gingerol-related compounds with relatively long side chains are
not easily detected by this method, due to their low volatility and thermal lability. LC-MS
has been shown to be an effective method for on-line analysis of this type of compounds
[9]. Single dimension MS analysis, however, cannot provide sufficient information to
accurately identify all known, let alone unknown, compounds. In contrast, our previous
investigation using LC-ESI-MS/MS to characterize and analyze three authentic gingerol
standards [12] suggested that this technique could be employed successfully as a
powerful and specific tool for on-line analysis of gingerol-related compounds.
200
This study sought to use LC-ESI-MS/MS to identify known and unknown
gingerol-related compounds in ginger rhizome. Both negative and positive ionization
ESI-MS/MS were used to obtain fragmentation data, which was used to characterize the
structures of this group of compounds. UV spectra were also obtained by an inline diode
array detector. By comparison of their mass spectra, UV spectra, and chromatographic
characteristics to those of authentic standard compounds and/or against each other, 31
gingerol-related compounds were identified. Three of these were determined as new
compounds.
EXPERIMENTAL
Chemicals and reagents
HPLC grade acetonitrile and methanol were from Burdick & Jackson (Muskegon, MI,
USA). Formic acid was from J. T. Baker (Mallinkrodt Baker, Inc., Phillipsburg, NJ,
USA). Ammonium formate was from Fisher Scientific (Fair Lawn, NJ, USA). Deionized water was re-distilled. Authentic standards of [6]-shogaol and [6]-, [8]-, and [10]- gingerols were purchased from ChromaDex, Inc. (Santa Ana, CA, USA).
Plant material and sample preparation
Fresh ginger rhizomes were collected from plants grown in a greenhouse at the
University of Arizona, frozen in liquid nitrogen, and kept at -80 °C until analyzed, as described previously [13]. Methanolic extracts for LC-ESI-MS/MS analyses were produced from fresh frozen ginger rhizome samples as previously described for [13],
201
with overnight extraction at room temperature and shaking at 200 rpm.
LC-ESI-MS/MS Analysis of ginger extracts
LC-ESI-MS/MS analyses of ginger extracts were performed on an Agilent 1100 HPLC
system coupled to an in-line diode array detector and an Agilent LC-MSD-Trap-SL ion
trap mass spectrometer (Palo Alto, CA, USA).
LC separation of gingerol-related compounds—Column: Discovery® HS C18, 3 µm, 15 cm × 2.1 mm (Supelco, Bellefonte, PA, USA); Guard column: Discovery® HS C18, 3
µm, 2 cm × 2.1 mm (Supelco); Mobile phase: (A) buffer (5mM ammonium formate,
0.1% formic acid, in ddH2O) and (B) acetonitrile; Gradient (in buffer A): 0-2 min, 5% B;
2-57 min, 5-100% B; 57-60 min, 100% B; 60-65 min, 100-5% B; 65-75 min, 5% B. Flow
rate: 0.25 ml/min; temperature, 40°C; Injection volume, 5 µl.
Diode Array Detection—The DAD was set at 425 nm (for signal A), 280 nm (for signal
B), and 230 nm (for signal C), at 4 nm bandwidth individually, with 550 nm reference
wavelength, at 50 nm bandwidth. Full spectral scanning was also performed from 200 to
600 nm, with range step of 2 nm.
MS and MS2 parameters for Agilent LC-MSD-Trap-SL—The acquisition parameters for
positive and negative modes were: drying N2 temperature, 350°C, flow rate 10 l/min;
nebulizer pressure 60 psi; HV capillary 4500 V; HV end plate offset -500 V; capillary
current 65.9 nA (positive mode), 62.3 (negative mode); end plate current 1482.7 nA
(positive mode), 1378.7 (negative mode); capillary exit RF amplitude 99.3 V (positive
mode), -99.3 V (negative mode); skimmer 40.0 V (positive mode), -40.0 V (negative
202 mode); mass range measured: 50-900 m/z. These were the optimized parameters for the maximum transmission of the gingerol-derived ions.
RESULTS AND DISCUSSION
General LC-MS/MS approach
Crude methanolic extracts, produced from fresh-frozen ginger rhizomes, were used directly for LC-ESI-MS/MS analyses, without further sample cleanup. This analysis also included in-line diode array detection. As shown in Fig. 2, both negative and positive ionization ESI-MS were used to detect gingerol related compounds. Obviously, (+)ESI-
MS is more sensitive than (−)ESI-MS in detecting this group of compounds under the condition used for our analysis (Fig. 2 and Table 1). Among the 31 identified compounds,
14 were detected after deprotonation in (−)ESI and only 8 of these provided (−)ESI-
MS/MS spectra (summarized in Table 1). In contrast, all 31 compounds were detected as protonated molecular ions, and/or ammonium/sodium adduct ions in (+)ESI-MS. In addition, (+)ESI-MS/MS spectra were obtained for all 31 compounds (summarized in
Table 1). For the 14 compounds detected in both (−)- and (+)ESI-MS analyses, their molecular weights were confirmed by deprotonated and protonated molecular ions/adduct ions, respectively. For those compounds only detected in (+)ESI, their molecular weights were confirmed by protonated molecular ions and/or their corresponding ammonium/sodium adduct ions. Their fragmentation behaviors in (−)- and/or (+)ESI-
MS/MS were used to confirm their molecular structures. In-line diode array detection set
203
at 425 nm, 280 nm, and 230 nm were also helpful in providing structural confirmation.
The gingerols show a characteristic UV absorption maximum at 280 nm and a shoulder at
230 nm [9]. In addition to these wavelengths, compounds were observed with absorption maxima at ~425 nm, suggesting the presence of an extended conjugation system. In addition to UV absorption, retention times (Rt) in RP-HPLC were also found to be useful
in the structure confirmation and especially for compounds belonging to homologous
series.
Characterization of compounds 1, 2, 7, 17, 24, 3, 11, and 20 (gingerols and
methylgingerols)
Compounds 1, 2, 7, 17, and 24 were observed in both negative and positive mode LC-
ESI-MS analyses. Based on deprotonated molecular ions in (−)ESI-MS and their
corresponding adduct ions in (+)ESI-MS (Table 1), the molecular weights of these
compounds, differentiated by units of 28 Da (-C2H4-), were confirmed. The molecular
weights of compounds 1, 2, 7, 17, and 24 matched those of [4]-, [6]-, [8]-, [10]-, and
[12]-gingerols, respectively. The retention times for all five compounds in RP-HPLC
increased by about 6 min for each additional -C2H2- unit (see Table 1). This observation
further supported that they belonged to a homologous series of increasing hydrophobicity
due to increased alkyl chain length. Compounds 1, 2, 7, and 17 were observed in the
HPLC-DAD chromatograms at 280 nm and 230 nm, characteristic absorption for the
gingerols, confirming our tentative identification (Fig. 2). Compound 24 was not
observed in the HPLC-DAD chromatograms due to its very low abundance in the crude
204
extract.
In addition, all five compounds demonstrated consistent fragmentation behaviors
in both (−)-and (+)ESI-MS/MS analyses, which further confirmed their identities.
Compounds 2, 7, and 17 were confirmed as [6]-, [8]-, and [10]-gingerols by comparison
of their fragmentation behaviors and retention times in RP-HPLC with those of authentic
standards, which have been previously investigated in detail [14]. Compound 24 was
identified as [12]-gingerol because its molecular weight and fragmentation behaviors in
both (−)-and (+)ESI-MS/MS was consistent with the other gingerols. Moreover, the
fragmentation behavior of 24 in (−)ESI-MS/MS has also been previously discussed [14].
Although compound 1 was detected in both (−)- and (+)ESI-MS analyses, its (−)ESI-
MS/MS spectrum was not available due to its low abundance in crude ginger rhizome
extracts. Based on its molecular weight and fragmentation behavior in (+)ESI-MS/MS
(Table 1 and Scheme 1), compound 1 was identified as [4]-gingerol.
In (+)ESI-MS, the protonated molecular ions of the gingerols were not observable.
+ Instead, ion A [M+H-H2O] was detected as a highly abundant peak in (+)ESI-MS due to
the keto (site of protonation) and hydroxy (loss of water) groups on the alkyl chain. In contrast, the ammonium and sodium adduct ions of the gingerols were detectable (Table
1). In addition, the fragmentation behaviors in (+)ESI-MS/MS of the ammonium and sodium adduct ions of the gingerols were different (Scheme 1). For ammonium adduct
+ ions, ion A [M+H-H2O] was formed as the major product ion by the loss of one H2O and one NH3. Ion A was further fragmented by the loss of a neutral alkyl moiety and a
rearrangement (Scheme 1), leading to the formation of a predominant ion B at m/z 177. In
205
contrast, ion C at m/z 217 was produced as the base peak in (+)ESI-MS/MS analysis of sodium adduct ions, which could only be rationalized by a McLafferty rearrangement and
the loss of a neutral alkyl moiety (Scheme 1). No ions formed by loss of H2O were
detected from the (+)ESI-MS/MS of the sodium adduct ions. This observation suggested
+ + that Na and NH4 associate with different functional groups of the gingerol molecules
during the ESI ionization process, with Na+ likely associating with the phenolic hydroxyl
+ group and NH4 associating with the alkyl hydroxyl group.
+ + Compared to compounds 2, 7, and 17, the NH4 and Na adducts of compounds 3,
11, and 20 showed an increase of 14 Da (Table 1), respectively, suggesting an additional
CH3 (Me) group instead of a H atom. Thus, these three compounds could be methyl [6]-
gingerol, methyl [8]-gingerol, and methyl [10]-gingerol or [7]-gingerol, [9]-gingerol, and
[11]-gingerol. Their characteristic product ions B at m/z 191 and C at m/z 231 (Table 1
and Scheme 1), which also demonstrated an increase of 14 Da, suggested that the
additional CH3 group was attached to the phenolic hydroxyl group. Therefore, compounds 3, 11, and 20 were identified as methyl [6]-gingerol, methyl [8]-gingerol, and
methyl [10]-gingerol. Their retention times, differentiated by about 6 min in sequence,
further supported that they were a series of homologs (Table 1). However, support from
HPLC-DAD analysis was only available for compound 3 due to the low abundance of
compounds 11 and 20 in the crude extract (Fig. 2). These three compounds are not likely
to be artifacts of the extraction procedure, which did employ methanol (extractions
performed with shaking at room temperature), because methanolic or aqueous methanolic
solutions of the pure gingerols, whether relatively diluted or concentrated, never show
206 contamination by these compounds, even if stored for prolonged periods of time (months)
(data not shown).
Characterization of compounds 6, 16, 23, and 12 (gingerol acetates)
The ammonium and sodium adduct ions of compounds 6, 16, and 23 observed in (+)ESI-
MS showed an increase of 42 Da when compared to compounds 2, 7, and 17 (Table 1). In addition, the base peak in (+)ESI-MS/MS spectra of these compounds was produced by the loss of 60 Da (AcOH). This information suggested that an acetoxy group instead of an hydroxy group was present on the aliphatic side chain of compounds 6, 16, and 23.
Therefore, the structures of acetoxy-[6]-gingerol, acetoxy-[8]-gingerol, and acetoxy-[10]- gingerol were suggested for compounds 6, 16, and 23, respectively. These proposed structures of compounds 6, 16, and 23 were further confirmed by the formation of product ion D at m/z 137, revealing the presence of the aromatic moiety of these molecules (Scheme 2).
Compared to compound 6, the ammonium adduct ion and all of the corresponding product ions of compound 12 showed an increase of 14 Da (Table 1), suggesting that these two compounds are homologs differentiated only by one CH2 group (CH3 instead of
H). Because of its product ion D (m/z 151 instead of 137), we concluded that compound
12 was methyl acetoxy-[6]-gingerol (Scheme 2).
Characterization of compounds 4 and 8 (monoacetyl gingerdiols)
Compared to compound 6, compound 4 showed an increase of 2 Da for its corresponding
207
+ ammonium and sodium adduct ions in (+)ESI-MS (Table 1). The ion [M+H-H2O] at m/z
321, resulting from the loss of one H2O and NH3 from the molecule, was the base peak in
+ the (+)ESI-MS/MS of its ammonium adduct ion [M+NH4] at m/z 356 (Table 1). This
observation suggested that compound 4 could be 3- or 5-acetoxy-[6]-gingerdiol with a
second aliphatic hydroxy group instead of a keto group on the alkyl chain. The relatively
lower retention time in RP-HPLC for this compound when compared to compound 6 also
supported the presence of a hydroxy group instead of a keto group. The formation of
product ions E at m/z 261 and F at m/z 163 confirmed our tentative identification of
compound 4 as 3- or 5-acetoxy-[6]-gingerdiol (Scheme 3). However, further evidence to
determine the position of the acetyl group on either the 3- or 5- hydroxy was not available.
In the exact same manner, compound 8 was identified as methyl 3- or 5-acetoxy-[6]-
gingerdiol when compared to compound 12.
Characterization of compounds 5, 13, 21, 27, 10, 18, 25, and 29 (diacetyl gingerdiols)
Compounds 5, 13, 21, and 27, respectively, differentiated by 28 Da (-C2H4-) for their
corresponding ammonium and sodium adduct ions in (+)ESI-MS, showed retention time increases of ~5.2 min in the RP-HPLC chromatograms, suggesting that these four compounds were homologs [Table 1]. The formation of product ions [M+H-AcOH]+ and
[M+H-2AcOH]+ in the (+)ESI-MS/MS analysis of ammonium adduct ions from these
compounds suggested two acetoxy groups on the alkyl chain. In addition, compared to
compound 4 (section 3.4), the ammonium and sodium adduct ions of compound 13
showed an increase of 42 Da, revealing that the hydroxy group on the alkyl chain of
208
compound 4 was substituted by one acetoxy group in compound 13. Therefore,
compound 13 was tentatively identified as diacetoxy-[6]-gingerdiol. The formation of
product ions E at m/z 261 and F at m/z 163 in (+)ESI-MS/MS for compounds 4 and 13
allowed us to confirm the structure for compound 13 (Scheme 3). By comparison to
compound 13, the other three homologs, compounds 5, 21, and 27, were identified as
diacetoxy-[4]-gingerdiol, diacetoxy-[8]-gingerdiol, and diacetoxy-[10]-gingerdiol,
respectively (Table 1 and Scheme 3). Among these, diacetoxy-[8]-gingerdiol and
diacetoxy-[10]-gingerdiol were identified as new compounds.
Compared to compounds 5, 13, 21, and 27, the ammonium and sodium adduct
ions in (+)ESI-MS and all their corresponding product ions in (+)ESI-MS/MS of
compounds 10, 18, 25, and 29 demonstrated an increase of 14 Da (-CH2), respectively
(Table 1 and Scheme 3). This observation suggested that compounds 10, 18, 25, and 29
were methyl diacetoxy-[4]-gingerdiol, methyl diacetoxy-[6]-gingerdiol, methyl
diacetoxy-[8]-gingerdiol, and methyl diacetoxy-[10]-gingerdiol, respectively. Among
these, methyl diacetoxy-[8]-gingerdiol was identified as a new compound.
Characterization of compounds 9, 19, 26, 30, and 14 (shogaols and paradols)
Deprotonated molecular ions in (−)ESI-MS and protonated molecular ions in (+)ESI-MS,
differentiated by 28 Da (-C2H4-), were detected for compounds 9, 19, 26, and 30 (Table
1), confirming the molecular weights of these compounds. [6]-, [8]-, [10], and [12]-
shogaol, a series of homologs which have been previously isolated from ginger, matched the molecular weights of compounds 9, 19, 26, and 30, respectively. In addition, these
209
compounds showed retention time increases of ~5.6 min in RP-HPLC in analysis (Table
1), supporting the indication that they were homologs. In the (+)ESI-MS/MS spectra of
these compounds, product ion D at m/z 137 was the only major peak observed (Scheme
4). This observation is reasonable because the keto group on the alkyl chain is the only
group that causes fragmentation in (+)ESI-MS/MS of these protonated molecules. By
comparing the chromatographic (retention time) and spectral data (ESI-MS/MS) with an
authentic standard compound, compound 9 was confirmed to be [6]-shogaol. Compounds
19, 26, and 30 were confirmed as [8]-shogaol, [10]-shogaol, and [12]-shogaol by
comparing their molecular weights, fragmentation behaviors in (+)ESI-MS/MS, and
retention times in RP-HPLC to those of compound 9. Further support from (−)ESI-
MS/MS was not available for these compounds due to their low abundance in the crude
extract of fresh ginger rhizomes. Compared to compound 9, compound 14 showed an
increase of 2 Da of its corresponding protonated ions in (+)ESI-MS (Table 1). This
suggested that compound 14 might be [6]-paradol, with the lack of the double bond
between carbons 4 and 5. Its relative later retention time when compared to [6]-shogaol
in RP-HPLC also supported this hypothesis. In addition, the formation of only major
product ion D at m/z 137 in (+)ESI-MS/MS confirmed compound 14 to be [6]-paradol
(Scheme 3).
Characterization of compounds 15, 22, 28, and 31 (dehydrogingerdiones)
Compounds 15, 22, 28, and 31 were detected in both (−)- and (+)ESI-MS (Table 1). Their
corresponding molecular weights, differentiated by 28 Da (-C2H4-), were thereby
210 obtained. A series of homologs, 1-dehydro-[6]-gingerdione, 1-dehydro-[8]-gingerdione,
1-dehydro-[10]-gingerdione, and 1-dehydro-[12]-gingerdione (Fig. 1), respectively, which have been previously reported from ginger, were suggested for compounds 15, 22,
28, and 31 because of matching molecular weights. In addition, all four compounds were observed in HPLC-DAD at 425 nm (Fig. 2), suggesting the presence of an extended conjugation system. This observation supported the presence of a double bond between carbons 1 and 2 of these homologs. Compared to the corresponding gingerols with alkyl chains of the same length, the dehydrogingerdiones showed relatively large retention times in RP-HPLC, due to the lack of an aliphatic hydroxy group (Table 1). Moreover, the base peak (ion G) at m/z 149 was produced by a β-H shift to the double bond in
(−)ESI-MS/MS, leading to the loss of a neutral moiety (Scheme 5). This rearrangement reaction was also observed for the curcuminoids in (−)ESI-MS/MS [12], closely related compounds which also possess a β-diketone group. Furthermore, the formation of the major product ion H at m/z 177 in (+)ESI-MS/MS also confirmed our tentative identification of these compounds as dehydrogingerdiones (Scheme 5).
CONCLUSIONS
A total of 31 gingerol-related compounds, belonging to different homologous series and differentiated by structural differences on the alkyl chain and the aromatic ring, were identified in methanolic crude extracts from fresh frozen ginger rhizome by LC-ESI-
MS/MS coupled to diode array detection. Interestingly, many of the identified
211
compounds were only detected by the MS detector, therefore, suggesting that the LC-MS
analysis is not only more specific but also more sensitive than diode array analysis for
this group of compounds. Another advantage offered by MS detection is that compounds with very close retention times in RP-HPLC can be distinguished and identified by selective (extracted) ion chromatograms from full MS and MS/MS analysis. This technique is important for the analysis of crude extracts of ginger rhizomes by HPLC without prior fractionation. Nevertheless, many gingerol-related compounds, especially those that lacked available phenolic hydroxy groups for deprotonation, were not
detectable by (−)ESI-MS analysis. Diode array detection was, however, very helpful for
structural distinction and confirmation and especially for compounds with UV absorption at specific wavelengths. For example, the 1-dehydrogingerdiones were easily
distinguished from other gingerol-related compounds in the HPLC-DAD chromatogram
at 425 nm, due to their extended conjugation systems (Fig. 2). Therefore, negative and
positive mode ESI-HPLC-MS/MS analysis coupled to diode array detection was found to
be a powerful and fast on-line tool for the identification of this group of compounds with
relatively complete coverage.
ACKNOWLEDGEMENTS
The authors acknowledge financial assistance from the National Science Foundation
Plant Genome Program, grant DBI-0227618 to D.R.G., and the National Institutes of
Health NCCAM/ODS, grants #5 P50 AT 000474-05 and 3 P50 AT 000474-03 S1 to
212
B.N.T. We also thank Veronica Rodriguez, Hyun Jo Koo, and Brenda Jackson for assistance with chemical analysis. The contents of this publication are solely the responsibility of the authors and do not necessarily represent the official views of
NCCAM, ODS, the NIH, the National Science Foundation.
213
REFERENCES
[1] M. Afzal, D. Al-Hadidi, M. Menon, J. Pesek, M.S. Dhami, Drug Metabol Drug
Interact 18 (2001) 159.
[2] K. Koo, A. Ammit, V. Tran, C. Duke, B. Roufogalis, Thromb. Res. 103 (2001)
387.
[3] Y. Masuda, H. Kikuzaki, M. Hisamoto, N. Nakatani, Biofactors 21 (2004) 293.
[4] V.N. Dedov, V.H. Tran, C.C. Duke, M. Connor, M.J. Christie, S. Mandadi, B.D.
Roufogalis, Br J Pharmacol 137 (2002) 793.
[5] S.O. Kim, J.K. Kundu, Y.K. Shin, J.H. Park, M.H. Cho, T.Y. Kim, Y.J. Surh,
Oncogene 24 (2005) 2558.
[6] P. Ruedi, M. Juch, Curr. Org. Chem. 3 (1999) 623.
[7] H.Y. Young, Y.L. Luo, H.Y. Cheng, W.C. Hsieh, J.C. Liao, W.H. Peng, J
Ethnopharmacol 96 (2005) 207.
[8] D.W. Connell, R. McLachlan, J. Chromatogr. 67 (1972) 29.
[9] X. He, M. Bernart, L. Lian, L. Lin, J. Chromatogr. A 796 (1998) 327.
[10] H. Huang, K. Kuo, Y. Hsieh, J. Chromatogr. A 771 (1997) 267.
[11] S. Jolad, R. Lantz, A. Solyom, G. Chen, R. Bates, B. Timmermann,
Phytochemistry 65 (2004) 1937.
[12] H. Jiang, A. Somogyi, B.N. Timmermann, D.R. Gang, Submitted to Rapid
Commun. Mass Spectrom. (2005).
[13] H. Jiang, Z. Xie, H. Koo, S.P. McLaughlin, B.N. Timmermann, D.R. Gang, In
Press for Phytochemistry (2005).
214
[14] H. Jiang, A. Somogyi, B.N. Timmermann, D.R. Gang, Submitted to J. Am. Soc.
Mass Spectr. (2005).
Table 1. Chromatographic and mass spectral characteristics of gingerol-related compounds detected by LC-ESI-MS in extracts from ginger rhizome
Negative ESI Positive ESI a a tR (min) (-) ESI-MS (m/z) (-) ESI-MS/MS (m/z) (+) ESI-MS (m/z) (+) ESI-MS/MS (m/z) Compound Compound name - c + 25.6 265 [M-H] N/D 249 [M+H-H2O] 177 1 [4]-Gingerol + 284 [M+NH4] 249, 267 289 [M+Na]+ N/D - + 32.5 293 [M-H] 193, 99, 275, 178 277 [M+H-H2O] 177 2 [6]-Gingerol + 312 [M+NH4] 277, 295,. 177 317 [M+Na]+ 217 611 [2M+Na]+ 317 + 35.7 N/D N/D 291 [M+H-H2O] 191 3 Methyl [6]-gingerol + 326 [M+NH4] 291, 309 331 [M+Na]+ 231 + 36.1 N/D 321 [M+H-H2O] 261, 163, 137 4 3- or 5-Acetoxy-[6]-gingerdiol + 356 [M+NH4] 321, 339 361 [M+Na]+ 301, 203 + 36.7 N/D 370 [M+NH4] 293, 233, 163, 353 5 Diacetoxy-[4]-gingerdiol 375 [M+Na]+ N/D 293 [M+H-AcOH]+ 233, 163, 137, 275 + 38.6 N/D 354 [M+NH4] 277, 337, 259, 137 6 Acetoxy-[6]-gingerol 359 [M+Na]+ N/D - + 38.7 321 [M-H] 193, 127, 303, 178 305 [M+H-H2O] 177 7 [8]-Gingerol + 340 [M+NH4] 305, 323 345 [M+Na]+ 217 + 39.1 N/D 370 [M+NH4] 275, 335, 353, 177 8 Methyl 3- or 5-acetoxy-[6]-gingerdiol + 335 [M+H-H2O] 275, 177, 151 39.9 275 [M-H]- N/D 277 [M+H]+ 137 9 [6]-Shogaol + 40.0 N/D 384 [M+NH4] 307, 247, 177, 367 10 Methyl diacetoxy-[4]-gingerdiol + 41.8 N/D 319 [M+H-H2O] 191 11 Methyl [8]-gingerol + 354 [M+NH4] 319, 337 359 [M+Na]+ N/D + 41.9 N/D 368 [M+NH4] 291, 351, 273, 151 12 Methyl acetoxy-[6]-gingerol 373 [M+Na]+ N/D
+ 215 42.2 N/D 398 [M+NH4] 321, 261, 163, 381 13 Diacetoxy-[6]-gingerdiol 403 [M+Na]+ 343, 163, 261, 137, 321, 283
321 [M+H-AcOH]+ 261, 163, 137, 303 42.8 N/D 279 137 14 [6]-Paradol 43.3 289 [M-H]- 149, 134, 139 291 [M+H]+ 177 15 1-Dehydro-[6]-gingerdione 313 [M+Na]+ N/D + 44.4 N/D 382 [M+NH4] 305, 365, 287, 137 16 Acetoxy-[8]-gingerol 387 [M+Na]+ N/D - + 44.6 349 [M-H] 193, 155, 178, 331 333 [M+H-H2O] 177 17 [10]-Gingerol + 368 [M+NH4] 333, 351 373 [M+Na]+ 217 + 45.2 N/D 412 [M+NH4] 335, 275, 177, 395 18 Methyl diacetoxy-[6]-gingerdiol 417 [M+Na]+ 357, 177, 275, 151, 297, 335 335 [M+H-AcOH]+ 275, 177, 151, 317 45.9 303 [M-H]- N/D 305 [M+H]+ 137 19 [8]-Shogaol + 47.4 N/D N/D 347 [M+H-H2O] 191 20 Methyl [10]-gingerol + 382 [M+NH4] 347, 365 387 [M+Na]+ 231 + b 47.6 N/D 426 [M+NH4] 349, 289, 163, 409 21 Diacetoxy-[8]-gingerdiol 431 [M+Na]+ N/D 349 [M+H-AcOH]+ 331, 289, 163, 187 48.8 317 [M-H]- 149, 134, 167 319 [M+H]+ 177 22 1-Dehydro-[8]-gingerdione - + 49.8 391 [M-H] N/D 410 [M+NH4] 333, 393, 137, 315 23 Acetoxy-[10]-gingerol 415 [M+Na]+ 355 - + 50.2 377 [M-H] 193, 183, 178, 361 361 [M+H-H2O] 177 24 [12]-Gingerol 401 [M+Na]+ 217 + b 50.5 N/D 440 [M+NH4] 363, 303, 177, 423 25 Methyl diacetoxyl-[8]-gingerdiol 51.4 331 [M-H]- N/D 333 [M+H]+ 137 26 [10]-Shogaol + b 52.8 N/D 454 [M+NH4] 377, 317, 163, 437 27 Diacetoxy-[10]-gingerdiol 459 [M+Na]+ N/D 53.9 345 [M-H]- 149, 134, 195 347 [M+H]+ 177 28 1-Dehydro-[10]-gingerdione + 55.2 N/D 468 [M+NH4] 391, 331, 177, 451 29 Methyl diacetoxyl-[10]-gingerdiol 56.3 359 [M-H]- N/D 361 [M+H]+ 137 30 [12]-Shogaol 58.0 373 [M-H]- 149, 223, 134 375 [M+H]+ 177 31 1-Dehydro-[12]-gingerdione a product ions shown in each row are given in the order of their relative abundance: the first ion, in each case, is the most abundant b
compound was tentatively identified as a new compound. 216 c N/D indicate that parent ion in ESI-MS and/or product ions in ESI-MS/MS was/were not detectable.
217
Figure Legends
Figure 1. The chemical structures and molecular weights of gingerol-related compounds identified in extracts from ginger rhizome. Note: * indicates new compound.
Figure 2. LC-MS analysis of ginger rhizome extracts. (a) and (b) are total ion current
(TIC) chromatograms from negative ion (–)ESI-HPLC-MS and positive ion (+)ESI-
HPLC-MS, respectively; (c), (d), and (e) are HPLC-DAD chromatograms set at 425 nm,
280 nm, and 230 nm, respectively, of the crude ginger rhizome extract.
Scheme 1. (a): (+)ESI-MS/MS fragmentation of gingerols and methylgingerols 1, 2, 7, 17,
24, 3, 11, and 20 (protonated ions and ammonium adducts).
(b): (+)ESI-MS/MS fragmentation of gingerols and methylgingerols 1, 2, 7,
17, 24, 3, 11, and 20 (sodium adducts).
Scheme 2. (+)ESI-MS/MS fragmentation of gingerol acetates 6, 16, 23, and 12.
Scheme 3. (+)ESI-MS/MS fragmentation of mono- and diacetyl gingerdiols 4, 5, 8, 13,
21, 27, 10, 18, 25, and 29.
Scheme 4. (+)ESI-MS/MS fragmentation of shogaols and paradols 9, 19, 26, 30, and 14.
Scheme 5. a): (–)ESI-MS/MS fragmentation of dehydrogingerdiones 15, 22, 28, and 31.
218 b): (+)ESI-MS/MS fragmentation of dehydrogingerdiones 15, 22, 28, and 31.
219
O OH O OH O MeO MeO MeO (CH2)nCH3 (CH2)nCH3 (CH2)nCH3 HO MeO HO (1) [4]-Gingerol, n=2, MW=266 (3) Methyl [6]-gingerol, n=4, MW=308 (9) [6]-Shogaol, n=4, MW=276 (2) [6]-Gingerol, n=4, MW=294 (11) Methyl [8]-gingerol, n=6, MW=336 (19) [8]-Shogaol, n=6, MW=304 (7) [8]-Gingerol, n=6, MW=322 (20) Methyl [10]-gingerol, n=8, MW=364 (26) [10]-Shogaol, n=8, MW=332 (17) [10]-Gingerol, n=8, MW=350 (30) [12]-Shogaol, n=10, MW=360 (24) [12]-Gingerol, n=10, MW=378
O O O OAc O OAc MeO MeO MeO (CH2)nCH3 (CH2)nCH3 (CH2)nCH3 HO HO MeO (15) 1-Dehydro-[6]-gingerdione, n=4, MW=290 (6) Acetoxy-[6]-gingerol, n=4, MW=336 (12) Methyl acetoxy-[6]-gingerol, n=4, MW=350 (22) 1-Dehydro-[8]-gingerdione, n=6, MW=318 (16) Acetoxy-[8]-gingerol, n=6, MW=364 (28) 1-Dehydro-[10]-gingerdione, n=8, MW=346 (23) Acetoxy-[10]-gingerol, n=8, MW=392 (31) 1-Dehydro-[12]-gingerdione, n=10, MW=374
OAc OH OH OAc OAc OH OH OAc MeO MeO MeO MeO (CH2)nCH3 (CH2)nCH3 (CH2)nCH3 (CH2)nCH3 HO or HO MeO or MeO
(4) 3- or 5-Acetoxy-[6]-gingerdiol, n=4, MW=338 (8) Methyl 3- or 5-acetoxy-[6]-gingerdiol, n=4, MW=352
O OAc OAc OAc OAc MeO MeO MeO (CH2)nCH3 (CH2)nCH3 (CH2)nCH3 HO HO MeO (14) [6]-Paradol, n=4, MW=278 (5) Diacetoxy-[4]-gingerdiol, n=2, MW=352 (10) Methyl diacetoxy-[4]-gingerdiol, n=2, MW=366 (13) Diacetoxy-[6]-gingerdiol, n=4, MW=380 (18) Methyl diacetoxy-[6]-gingerdiol, n=4, MW=394 (21) Diacetoxy-[8]-gingerdiol, n=6, MW=408 * (25) Methyl diacetoxy-[8]-gingerdiol, n=6, MW=422 * (27) Diacetoxy-[10]-gingerdiol, n=8, MW=436 * (29) Methyl diacetoxy-[10]-gingerdiol, n=8, MW=450
Figure 1. The chemical structures and molecular weights of gingerol-related compounds identified in extracts from ginger rhizome. Note: * indicates new compound.
220
6 (a) Total Ion Current (TIC) chromatogram from negative ion (-)ESI-HpLC-MS x10 2 24 1.00 15 17 1 0.75 9 23 26 31 7 19 28 0.50 30 0.25 22 0.00 20.0 30.0 40.0 50.0 60.0 8 x10 (b) Total Ion Current (TIC) chromatogram from positi ve i on (+)ESI-HpLC-MS 11 ,12 16,17 23 0.8 4 29 9,10 14 20,21 24 25 31 1 26 28 0.6 2 6,7,8 13 18 19 22 27 30 0.4 3 15 0.2 5 0.0 20.0 30.0 40.0 50.0 60.0 mAU (c) HPLC-DAD chromatogram (425 nm) 60 15
40 28
Intensity 20 22 31 0 20.0 30.0 40.0 50.0 60.0
mAU (d) HPLC-DAD chromatogram (280 nm) 2 1500 19 9,10 1000 1 3 26 31 6,7,8 27 16,17 28 500 15 25 0 20.0 30.0 40.0 50.0 60.0 (e) HPLC-DAD chromatogram (230 nm) mAU 2
2000 1 28 6,7,8 26 9,10 16,17 27 1000 3 15 19 25 0
20.0 30.0 40.0 50.0 60.0 Retention time (min)
Figure 2. LC-MS analysis of ginger rhizome extracts. (a) and (b) are total ion current
(TIC) chromatograms from negative ion (–)ESI-HPLC-MS and positive ion (+)ESI-
HPLC-MS, respectively; (c), (d), and (e) are HPLC-DAD chromatograms set at 425 nm,
280 nm, and 230 nm, respectively, of the crude ginger rhizome extract.
221
(a) O [M+NH ]+ + O + 4 MeO CH3(CH2)nCHO MeO MeO (CH )nCH + (CH2)nCH3 + 2 3 [M+H-H2O] RO RO RO [M+H]+ A 249 (1; n=2, R=H) 277 (2; n=4, R=H) 177 (1,2,7,17,24; R=H) B 305 (7; n=6, R=H) 191 (3,11,20; R=Me) A 333 (17; n=8, R=H) 361 (24; n=10, R=H) 291 (3; n=4, R=Me) 319 (11; n=6, R=Me) 347 (20; n=8, R=Me)
+ OR OR (b) O OH + Na + CH3(CH2)nCHO + 3' Na Na MeO CH (CH )n (CH )nCH 3 2 OMe OMe 1 3 5 2 3 OH RO O O 4' 266 (1; n=2, R=H) H 289 (1; n=2, R=H) 217 (2,7,17,24; R=H) C 294 (2; n=4, R=H) 317 (2; n=4, R=H) 231 (3,20; R=Me) 322 (7; n=6, R=H) 345 (7; n=6, R=H) 350 (17; n=8, R=H) 373 (17; n=8, R=H) 378 (24; n=10, R=H) 401 (24; n=10, R=H) 308 (3; n=4, R=Me) 331 (3; n=4, R=Me) 336 (11; n=6, R=Me) 359 (11; n=6, R=Me) 364 (20; n=8, R=Me) 387 (20; n=8, R=Me)
Scheme 1. (a): (+)ESI-MS/MS fragmentation of gingerols and methylgingerols 1, 2, 7, 17,
24, 3, 11, and 20 (protonated ions and ammonium adducts).
(b): (+)ESI-MS/MS fragmentation of gingerols and methylgingerols 1, 2, 7,
17, 24, 3, 11, and 20 (sodium adducts).
222
O OAc MeO 3' (CH )nCH [M+NH ]+ [M+H]+ [M+H-AcOH]+ [M+H-AcOH-H2O]+ 1 3 5 2 3 4 RO 4' 336 (6; n=4, R=H) 354 (6) 337 (6) 277 (6) 259 (6) 364 (16; n=6, R=H) Mr 382 (16) 365 (16) 305 (16) 287 (16) 392 (23; n=8, R=H) 410 (23) 393 (23) 333 (23) 315 (23) 350 (12; n=4, R=Me) 368 (12) 351 (12) 291 (12) 273 (12)
H + + + OH OAc O OAc MeO CH 2 MeO MeO (CH2)nCH3 (CH2)nCH3 RO RO RO 137 (6,16,23; R=H) D 151 (12; R=Me)
Scheme 2. (+)ESI-MS/MS fragmentation of gingerol acetates 6, 16, 23, and 12.
223
OAc OAc MeO 3' (CH )nCH [M+NH ]+ [M+H]+ [M+H-AcOH]+ [M+H-2AcOH]+ 1 3 5 2 3 4 RO 4' 352 (5; n=2, R=H) 370 (5) 353 (5) 293 (5) 233 (5) Mr 380 (13; n=4, R=H) 398 (13) 381 (13) 321 (13) 261 (13) 408 (21; n=6, R=H) 426 (21) 409 (21) 349 (21) 289 (21) 436 (27; n=8, R=H) 454 (27) 437 (27) 377 (27) E 317 (27) 366 (10; n=2, R=Me) 384 (10) 367 (10) 307 (10) 247 (10) 394 (18; n=4, R=Me) 412 (18) 395 (18) 335 (18) 275 (18) 422 (25; n=6, R=Me) 440 (25) 423 (25) 363 (25) 303 (25) 450 (29; n=8, R=Me) 468 (29) 451 (29) 391 (29) 331 (29)
OAc OH MeO 3' (CH )nCH 1 3 5 2 3 RO 4' Mr 338 (4; n=4, R=H) 352 (8; n=4, R=Me) [M+NH ]+ [M+H]+ [M+H-H O]+ [M+H-H O-AcOH]+ or 4 2 2 OH OAc 356 (4) 339 (4) 321 (4) 261 (4) MeO 3' E (CH )nCH 370 (8) 353 (8) 335 (8) 275 (8) 1 3 5 2 3 RO 4'
OR OR OR OMe OMe OMe + + MeO MeO (CH2)nCH3 (CH2)nCH3 RO RO + E E H + (CH )nCH + 2 3 163 (5,13,21,27,4; R=H) F 177 (24,3,11,20,8; R=Me)
Scheme 3. (+)ESI-MS/MS fragmentation of mono- and diacetyl gingerdiols 4, 5, 8, 13,
21, 27, 10, 18, 25, and 29.
224
+ OH OH + 3' 1 5 (CH )nCH MeO 2 3 MeO CH2 (CH2)nCH3 HO HO 4' 277 (9; n=4) Mr 305 (19; n=6) 333 (26; n=8) D 137 (9,19,26,30) 361 (30; n=10)
+ OH OH + 3' 1 5 (CH )nCH MeO 2 3 MeO CH2 (CH2)nCH3 HO HO 4' Mr 279 (14; n=4) D 137 (14)
Scheme 4. (+)ESI-MS/MS fragmentation of shogaols and paradols 9, 19, 26, 30, and 14.
225
O H O (a) (CH )nCH 2 3 OMe OMe O O G 149 (15,22,28,31) O O MeO 3' (CH )nCH 1 3 5 2 3 O HO + 4' 290 (15; n=4) O OH MeO + (b) MeO 318 (22; n=6) (CH )nCH 346 (28; n=8) 2 3 HO 374 (31; n=10) HO H 177 (15,22,28,31)
Scheme 5. a): (–)ESI-MS/MS fragmentation of dehydrogingerdiones 15, 22, 28, and 31.
b): (+)ESI-MS/MS fragmentation of dehydrogingerdiones 15, 22, 28, and 31.
226
APPENDIX E
Manuscript “Use of LC-ESI-MS/MS to identify diarylheptanoids in turmeric (Curcuma
longa L.) rhizome”, submitted to the Journal of Chromatography A.
227
Title: Use of LC-ESI-MS/MS to identify diarylheptanoids in turmeric (Curcuma longa L.) rhizome
Hongliang Jianga,b,c,d, Barbara N. Timmermanna,e, and David R. Ganga,b,c*
aArizona Center for Phytomedicine Research, College of Pharmacy, University of
Arizona, Tucson, AZ 85721, USA
bDepartment of Plant Sciences, College of Agriculture and Life Sciences, University of
Arizona, Tucson, AZ 85721, USA
cBIO5 Institute, University of Arizona, Tucson, AZ 85721, USA
dDepartment of Pharmaceutical Sciences, College of Pharmacy, University of Arizona,
Tucson, AZ 85721, USA
eDepartment of Pharmacology and Toxicology, College of Pharmacy, University of
Arizona, Tucson, AZ 85721, USA
*Corresponding author:
David R. Gang
Department of Plant Sciences and BIO5 Institute, University of Arizona, Tucson, AZ
85721-0036, USA
Tel: 520-621-7154
Fax: 520-621-7186
email: [email protected]
228
Abstract
LC-ESI-MS/MS coupled to DAD analysis was used as an on-line tool for identification of diarylheptanoids in fresh turmeric rhizome extracts. Based on their mass spectra, from both negative and positive mode LC-ESI-MS/MS analysis, and supported by their DAD spectra, 19 diarylheptanoids were identified. Among these 19 compounds, curcumin, demethoxycurcumin, and bisdemethoxycurcumin were identified by comparing their chromatographic and spectral data with those of authentic standard compounds. The other diarylheptanoid compounds were identified based on comparison to the three curcuminoids and each other. Twelve of the identified diarylheptanoids have not been previously reported from turmeric and six of these are new compounds.
Keywords: Diarylheptanoids; curcuminoids; curcumin; turmeric; Curcuma longa; rhizome; electrospray ionization; tandem mass spectrometry
229
1. Introduction
Turmeric (Curcuma longa L.), a member of the Zingiberaceae and a well known spice, has long been used in cosmetics, as a coloring agent, and in medicinal preparations in Eastern civilization [1]. Compounds belonging to two important groups of natural products, the diarylheptanoids and sesquiterpenoids, are believed to be responsible for producing many of the important biological and medicinal activities of turmeric [2]. The curcuminoids (see Fig. 1), including curcumin, demethoxycurcumin and bisdemethoxycurcumin, are the major diarylheptanoids and have been shown to contribute to many of the medicinal properties of this species, such as anti-inflammatory, cancer prevention, and anti-HIV activities [1,3-17]. However, it is not clear that these three compounds are the only diarylheptanoids responsible for activity in the Curcuma species.
A variety of methods including gas-chromatography/mass spectrometry (GC/MS), high performance liquid chromatography (HPLC), and its coupling to mass spectrometry
(LC-MS), thin layer chromatography (TLC) and capillary electrophoresis (CE) have been used to analyze the chemical content of various turmeric samples [18-28]. Almost all of these previous reports have focused on the separation and quantitation of the three major curcuminoids from dried turmeric powder [21-24,28], or of sesquiterpenoids found in the essential oil from the rhizome [18,26,27]. Only one report has been published [20] that used LC-MS to identify the three major curcuminoids (plus dihydrocurcumin) and some sesquiterpenoids in turmeric. However, because only single dimensional MS analysis
230
with only positive ionization was reported (yielding only parent ion molecular weight
information), no detailed structural information of the analyzed compounds could be
obtained. As stated in that report, this technique led to ambiguous peak identifications, where at least four different sesquiterpenoids known to be present in turmeric could be assigned to the actual peak [20]. Furthermore, single-dimensional LC-MS cannot provide sufficient information to potentially identify unknown compounds in the rhizome.
In our initial LC-MS-based metabolic profiling investigations with turmeric rhizome extracts, we noticed that, in addition to the three major curcuminoids, a number of other compounds were present that appeared to be diarylheptanoids. Furthermore, because curcumin, demethoxycurcumin, and bisdemethoxycurcumin have been shown to exhibit common fragmentation patterns in ESI-MS/MS [29], we hypothesized that LC-
ESI-MS/MS analysis could provide support for the identification of other known or of unknown diarylheptanoids in turmeric. The purpose of this study was to evaluate the use of both negative and positive ionization in LC-ESI-MS/MS analysis to identify
diarylheptanoids in fresh turmeric samples. Obtaining data from the two modes would provide further support for the assignment of the proposed structures. In addition, we
hypothesized that in-line diode array detection might provide additional evidence to
confirm the proposed structures of certain compounds found using LC-ESI-MS/MS. In
total, 19 diarylheptanoids were identified in our study, six of which were identified as
new compounds and six of the known compounds had not been previously reported from
turmeric.
231
2. Experimental
2.1. Chemicals and reagents
Authentic standards of curcumin, demethoxycurcumin, and
bisdemethoxycurcumin were obtained from Dr. S.D. Jolad, Arizona Center for
Phytomedicine Research, University of Arizona, Tucson, AZ. HPLC grade acetonitrile
and methanol were from Burdick & Jackson (Muskegon, MI, USA). Formic acid was
from J. T. Baker (Mallinkrodt Baker, Inc., Phillipsburg, NJ, USA). Ammonium formate
was from Fisher Scientific (Fair Lawn, NJ, USA). Deionized water was re-distilled.
2.2. Plant material and sample preparation
Fresh turmeric rhizomes were collected from plants grown in a greenhouse at the
University of Arizona, frozen in liquid nitrogen, and kept at -80 °C until analyzed, as
described previously [30]. Methanolic extracts for LC-MS analysis were produced from
fresh frozen turmeric rhizome samples as previously described for ginger [30], with
overnight extraction at room temperature and shaking at 200 rpm.
2.3. LC-ESI-MS/MS analysis of turmeric extracts
LC-ESI-MS/MS analysis of turmeric extracts was performed on an Agilent 1100
HPLC system coupled to an in-line DAD detector and an Agilent LC-MSD-Trap-SL ion
trap mass spectrometer (Palo Alto, CA, USA).
232
LC separation of curcuminoids—Column: Discovery® HS C18, 3 µm, 15 cm ×
2.1 mm (Supelco, Bellefonte, PA, USA); Guard column: Discovery® HS C18, 3 µm, 2
cm × 2.1 mm (Supelco); Mobile phase: (A) buffer (5mM ammonium formate, 0.1%
formic acid, in ddH2O) and (B) acetonitrile; Gradient (in buffer A): 0-2 min, 5% B; 2-57 min, 5-100% B; 57-60 min, 100% B; 60-65 min, 100-5% B; 65-75 min, 5% B. Flow rate:
0.25 ml/min; temperature, 40°C; Injection volume, 5 µl.
Diode Array Detection—The DAD was set at 230 nm (for signal A) and 425 nm
(for signal B), at 4 nm bandwidth individually, with 550 nm reference wavelength, at 50 nm bandwidth. Full spectral scanning was also performed from 200 to 600 nm, with range step at 2 nm.
MS and MSn detection—The Agilent MSD-Trap-SL was equipped with an electrospray ionization (ESI) interface as the ion source. The acquisition parameters for
positive and negative mode were: drying N2 temperature, 350°C, 10 l/min; nebulizer
pressure 60 psi; HV capillary 4500 V; HV end plate offset -500 V; capillary current 65.9
nA (positive mode), 62.3 (negative mode); end plate current 1482.7 nA (positive mode),
1378.7 (negative mode); capillary exit RF amplitude 99.3 V (positive mode), -99.3 V
(negative mode); skimmer 40.0 V (positive mode), -40.0 V (negative mode); mass range
measured: 50-900 m/z. These were the optimized parameters for the maximum
transmission of the curcuminoid-derived ions.
233
3. Results and discussion
LC-MS/MS analysis has been shown to be a powerful tool in metabolic profiling and metabolomics research [31-33]. This technology can accurately determine the content of specific metabolites, even if these are found at low levels in the plant samples. When coupled to in-line diode array detection, the accuracy of identification and quantitation can be significantly increased. Although studies using methods to quantitate several constituents in dried turmeric samples [23,24,34] and one study using LC-MS to identify some constituents of fresh turmeric have been reported, detailed LC-MS/MS analyses of the bioactive constituents of fresh turmeric have not yet been reported. Such investigations are necessary to allow for metabolic profiling studies to be performed that could be used to ensure the authenticity of plant materials [33] or to enable biochemical investigations that seek to define the biosynthetic pathways leading to these compounds in the plant.
3.1. General LC-MS/MS approach
Crude methanolic extracts, which are commonly produced for metabolic profiling experiments, were produced from fresh-frozen turmeric rhizomes, see Experimental. The same crude extracts were then used directly for LC-MS/MS analysis, without further sample cleanup. This analysis also included in-line diode array detection. When we analyzed the resulting LC-MS/MS chromatograms, it became very clear that the three major curcuminoids were not the only diarylheptanoids present in the analyzed extracts.
234
Indeed, a total of 19 diarylheptanoids were identified in these experiments, as described
below (see Fig. 1).
Negative mode (–)ESI-LC-MS was very effective at identifying diarylheptanoids
from turmeric. An example of a Total Ion Current (TIC) chromatogram of the extract is
shown in Fig. 2a. Diarylheptanoids are suitable for detection by (–)ESI-MS because of
the presence of phenolic hydroxyl groups, which enables these compounds to be easily
ionized in negative ESI mode. However, low abundance compounds are often not easily detected by (–)ESI-MS, because of the relatively low sensitivity of (–)ESI compared to
(+)ESI. Nevertheless, the fragmentation spectra from the (–)ESI-MS/MS analysis were
structurally very informative and allowed for the identification of many diarylheptanoids
in the examined samples. (+)ESI-MS is more sensitive, especially for compounds
containing atoms such as O or N, which can often be easily protonated. However, LC-
(+)ESI-MS can generate complex TIC chromatograms due to the complex matrix of the
crude extracts (Fig. 2b). Furthermore, in addition to the [M+H]+ ion, adduct ions and daughter ions from the parent ions are easily produced and detected in (+)ESI-MS, often making it more difficult to characterize unknown compounds with the positive mode than with the negative mode. Even so, the fragmentation spectra obtained from (+)ESI-
MS/MS analysis were very useful in determining the structure of detected compounds, especially in determining the molecular weight of the parent molecules. Combining analyses from both positive and negative modes was found to be the most effective, and the complementary fragmentation spectra provided firm evidence for compound characterization and identification.
235
In-line diode array detection, which allowed for UV spectra to be obtained for eluting compounds, was also very helpful in structural confirmation (Fig. 2c and 2d).
Compounds containing at least two or more double bonds could be detected at 230 nm.
Compounds could be detected at 425 nm if they possessed a long conjugation system
with more than six double bonds in a sequence, as in the case for the three major curcuminoids [35].
3.2. Characterization of compounds 14, 16, 18, and 11
Compounds 14, 16, and 18 were identified as bisdemethoxycurcumin, demethoxycurcumin, and curcumin, respectively, the three major curcuminoids in
turmeric. Both chromatographic and spectral characteristics of these compounds were
identical to those of authentic standards (data not shown). The detailed fragmentation
behaviors of these three curcuminoids from both (–)ESI-MS/MS and (+)ESI-MS/MS
have been previously reported [29]. That study also built a foundation for us to identify
other known or unknown diarylheptanoids by comparing their chromatographic and
spectral characteristics against each other.
Compared to the three major curcuminoids, compound 11 was a minor peak in the
TIC from negative ion LC-(–)ESI-MS and in the HPLC/DAD chromatogram at 230 nm
and 425 nm. In the (–)ESI-MS/MS spectrum, ion A at both m/z 187 and 203 and ion B at
m/z 143, characteristic of the three major curcuminoids [29], were observed (Fig. 3).
Furthermore, the parent ion of compound 11 showed a difference of 16 daltons from that
of compound 14. This observation suggested that there was one additional O atom
236
(hydroxy instead of H) in compound 11 when compared to compound 14. Only if the
hydroxy is present on the benzene ring (instead of on the heptanoid chain) can the fragmentation behavior observed for compound 11 be rationalized, as we previously described for the three major curcuminoids [29]. For compounds like the diarylheptanoids that are derived from the phenylpropanoid pathway, the ring substituents are usually located at the C-3′ and/or C-4′ and/or C-5´ positions. In addition,
this would make sense biosynthetically, where compound 11 may be an intermediate in
the production of compound 16 from 14. The fact that it was possible to obtain clear
fragmentation spectra in the (–)ESI-MS/MS analysis of compound 11, even though it was
of a relatively low abundance in the matrix, also supported the presence of multiple
hydroxy groups on the aromatic ring, facilitating its ionization and making it easy to be
seen in the negative mode. The protonated ion [M+H]+ of the minor compound 11 was
also detected in LC-(+)ESI-MS analysis. However, high quality MS/MS spectra from
(+)ESI-MS/MS analysis of this compound was not available, due to its low abundance in
the complex matrix. Based on the fragment behavior in (–)ESI-MS/MS and further
confirmation of its molecular weight by (+)ESI-MS, we identified compound 11 as 1-
(3,4-dihydroxyphenyl)-7-(4-hydroxyphenyl)-1,6-heptadien-3,5-dione, or 4´-hydroxy-
bisdemethoxycurcumin (Fig. 1). This compound is reported here for the first time.
3.3. Characterization of compounds 13, 15A, 15B and 17
Compounds 13, 15A, 15B, and 17 were identified as
dihydrobisdemethoxycurcumin, dihydrodemethoxycurcumin, letestulanin B, and
237
dihydrocurcumin, respectively, by comparing their corresponding chromatographic and
spectral data to compounds 14, 16, and 18 (Fig. 2 and Fig. 3). As discussed for
compounds 14, 16, and 18, the base peaks (product ion A at m/z 189, 219, and 219,
respectively) were produced through the loss of a neutral moiety via a β-hydrogen shift in
the (–)ESI-MS/MS analysis (scheme 1a). Compared to compounds 14, 16, and 18,
compounds 13, 15, and 17 differed by saturation between carbons 6 and 7 (Fig. 1). Thus,
in both the parent ion and product ion A of compounds 13, 15, and 17, we observed a
mass shift by 2 daltons with reference to the corresponding ions from compounds 14, 16,
and 18 (Fig. 3). Due to the lack of the double bond between carbons 6 and 7 in ion A of
compounds 13, 15, and 17, ion B observed from compounds 14, 16, 18, and 11(Fig. 3),
formed from further fragmentation of ion A [29], was not observed for compounds 13, 15,
and 17. Instead ions F and G were produced individually by rearrangement and the loss
of a neutral moiety from the deprotonated compounds (Scheme 1a). In addition, two
product ions A were produced by compound 15 (m/z 219 and 189), suggesting that it
actually exists as two isomers (15A and 15B, Fig. 1).
In the (+)ESI-MS/MS spectra (Fig. 3), product ion C (m/z 147, 177 and 147, and
177, respectively) was observed as the major peak for all three compounds. Its formation is outlined in Scheme 1b. Product ions D and E, characteristic of compounds 14, 16, and
18, were not observed for compounds 13, 15, and 17 in (+)ESI-MS/MS (Fig. 3).
Therefore, the presence of a double bond between carbons 6 and 7 appears to be essential for the formation of ions D and E in compounds 14, 16, and 18 [29]. As was seen for product ion A, two product ions C were produced by compound 15 (m/z 177 and 147),
238
supporting the presence of compound 15 as a mixture of two isomers (15A and 15B).
Although, compounds 13 and 17 have been previously identified from the rhizome of turmeric [36,37], compound 15A (dihydrodemethoxycurcumin) has only been reported
from the callus culture of turmeric [38]. Compound 15B (letestulanin B) has been
identified from Aframomum letestuianum but not from turmeric [39].
3.4. Characterization of compound 12
Product ions A, F, and G, characteristic of compounds 13, 15, and 17, were also
observed in the (–) ESI-MS/MS analysis of compound 12 (Fig. 3). Product ion A formed,
however, through a neutral loss via a β-hydrogen shift, was not present as the base peak.
This observation suggested that the presence of double bond(s) between carbons 1 and 2
and/or carbons 6 and 7 may facilitate the conformation for the neutral moiety loss via a β-
hydrogen shift. Product ion F at m/z 205, produced by a different neutral moiety loss, was
the base peak in the (–) ESI-MS/MS analysis (Scheme 2a) of this compound. In (+)ESI-
MS/MS, the fragmentation behavior of compound 12 is also very different from that of
14, due to the structural difference. Product ions C, D, and E, characteristic of compound
14, were not observed in the (+)ESI-MS/MS spectra of compound 12 because their
formation would require a double bond to serve as a bridge to delocalize the positive
charge [29]. Instead, product ion H at m/z 107 was formed as the base peak by a neutral
moiety loss (Scheme 2b). The fact that no UV absorption was observed at 425 nm also
supported its identification as a diarylheptanoid lacking an extended conjugated system.
By comparing chromatographic and spectral data of compound 12 with those of
239 compound 13, compound 12 was identified as tetrahydrobisdemethoxycurcumin.
Although previously reported from Aframomum letestuianum [39], this is the first time that its presence has been reported in turmeric.
3.5. Characterization of compound 2
Compound 2 was identified as 5-hydroxy-1,7-bis(4-hydroxyphenyl)-3-heptanone by comparing its MS/MS spectra and chromatographic data with those of compounds 12,
13, and 14. Owing to the presence of β hydroxy/keto groups (at positions 5 and 3, respectively) on the alkyl chain, a McLafferty rearrangement and loss of a neutral moiety was induced in (–)ESI-MS/MS analysis (Scheme 3a). Daughter ions I at m/z 149 (base peak) and J at m/z 163 were produced by this reaction because either ring could be deprotonated during ionization and hold the negative charge. In (+)ESI-MS/MS, the loss of H2O ion K predominated because of the presence of an unconjugated hydroxy group on the alkyl chain (Scheme 3b). Owing to its saturated alkyl chain, no absorption was observed at 425 nm. Compared to compounds 12, 13, and 14, its earlier retention time in
RP-HPLC also supported its relatively hydrophilic properties, with an extra hydroxy group instead of the presence of a keto group on the alkyl chain. Although it has been reported from other plants such as Betula pendula [40], compound 2 has not previously reported from turmeric.
240
3.6. Characterization of compound 1
We tentatively identified compound 1 as 1,7-bis(4-hydroxyphenyl)-3,5-heptanediol.
Compound 1 showed a postitive mass shift of 2 daltons compared to compound 2 in the
(–)ESI-MS analysis. In addition, its retention time was slightly earlier than that for
compound 2. Although it was detected by (–)ESI-MS, product ions were difficult to
observe with this method. The lack of a keto group in the alkyl chain made it difficult to fragment in the (–)ESI-MS/MS analysis, thus the parent ion at m/z 315 was still the base
peak (Fig. 3). Only one product ion I was observed as a minor peak at m/z 149, and it was
formed by the loss of a neutral moiety (Scheme 4). Similarly, the lack of a keto group
(for protonation) and the low content of compound 1 in the plant sample made it
undetectable in (+)ESI-MS analysis from the complex matrix. In addition, no absorption
at 425 nm was observed. Altogether, these results supported our tentative identification of
compound 1, which has been reported from Alpinia blepharocalyx [41], but not
previously from turmeric.
3.7. Characterization of compounds 6 and 8
Compounds 6 and 8 showed strong absorption at 425 nm, suggesting the presence
of a long conjugated system. The parent ions from both negative and positive ESI-MS
and ESI-MS/MS analyses suggested molecular weights of 322 and 352, respectively.
Fragment ions were difficult to observe from either negative or positive ESI-MS/MS (Fig.
3), suggesting very stable chemical structures that were difficult to break apart.
Compared to compound 8, the mass of compound 6 was 30 daltons less, suggesting the
241
lack of a methoxy group. Also, the retention time from RP-HPLC was very close for
these two compounds. In the (–)ESI-MS/MS analysis of compounds 6 and 8, ion [M-CO-
H]- was observed at m/z 293 and 323, respectively, as the major characteristic fragment,
even though it was present as a small peak in the spectrum (Fig. 3). Based on this
information, compound 8 was tentatively identified as 1,7-bis(4-hydroxy-3- methoxyphenyl)-1,4,6-heptatrien-3-one, which has been previously reported from
turmeric [42]. By comparison to compound 8, compound 6 appeared to be 1-(4-
hydroxyphenyl)-7-(4-hydroxy-3-methoxyphenyl)-1,4,6-heptatrien-3-one (Fig. 1). This
compound has not been previously reported.
3.8. Characterization of compounds 3, 4, and 5
Compound 3 showed very simple and clear fragmentation patterns in both (–)ESI-
MS/MS and (+)ESI-MS/MS (Fig.3), where the base peak was produced by a neutral
moiety loss of 106 daltons. In addition, compared to compound 6, compound 3 showed
an increase of 2 daltons, suggesting a reduction of a double bond. This observation also
supported the proposed structure for compound 6 as discussed above. The fragmentation
behavior of compound 3 was rationalized as shown in Scheme 5a and 5b for (–)ESI-
MS/MS and (+)ESI-MS/MS. Based on this information and comparison to the other
diarylheptanoids observed in this study, we identified compound 3 as 1-(4-
hydroxyphenyl)-7-(4-hydroxy-3-methoxyphenyl)-4,6-heptadien-3-one, which has only
been reported from the roots of Juglans mandshurica [43].
242
Compared to compound 3, compounds 4 and 5 demonstrated a similar
fragmentation behavior in (+)ESI-MS/MS (Fig. 3 and Scheme 5b) and showed an
increase by 30 and 60 daltons, indicating the presence of one and two additional methoxy
groups. Although deprotonated compounds 4 and 5 were also detected in the (–)ESI-MS
analysis, good quality (–)ESI-MS/MS spectra were not obtained under the conditions
used for this study, due to their low abundance in the crude methanolic extract. Based on
this information, compound 4 was tentatively identified as 1,7-bis(4-hydroxy-3-
methoxyphenyl)-4,6-heptadien-3-one, which has been previously identified from ginger
[44], and compound 5 was tentatively identified as 1-(4-hydroxy-3-methoxyphenyl)-7-(4-
hydroxy-3,5-dimethoxyphenyl)-4,6-heptadien-3-one. Compound 5 was identified as a new compound.
3.9. Characterization of compounds 9 and 10
Both compounds 9 and 10 showed a base peak at m/z 177 (ion C) in the (+)ESI-
MS/MS analysis (Fig. 3), which is characteristic of some diarylheptanoids and suggests a
+ common structural moiety for these two compounds. In addition, ion N [M-H2O+H] was
another major characteristic peak in the (+)ESI-MS/MS spectra, suggesting the presence
of an hydroxy group on the chain instead of being on the benzene ring. Based on this
information, we proposed the chemical structures of compounds 9 and 10 given in Fig. 1
and named them 1-hydroxy-1-(4-hydroxyphenyl)-7-(4-hydroxy-3-methoxyphenyl)-6-
hepten-3,5-dione and 1-(3,4-dihydroxyphenyl)-7-(4-hydroxy-3-methoxyphenyl)-6-
hepten-3,5-dione, respectively. The proposed structures of compounds 9 and 10 were
243
confirmed by their fragmentation behavior in (+)ESI-MS/MS (Scheme 6). However, further support from (–)ESI-MS/MS analysis of these compounds was not available, due to their low abundance in the plant material. These two compounds were tentatively identified as new compounds.
3.10. Characterization of compound 7
Although compound 7 was detected in both (–)ESI-MS and (+)ESI-MS analysis, only the (+)ESI produced a MS/MS spectra for this compound. Ion O, obtained by the loss of one H2O, served as the base peak in the (+)ESI-MS/MS spectrum, suggesting the
presence of an hydroxy group on the alkyl chain. Another characteristic ion C was
obtained by further fragmentation of ion P (Scheme 7). We proposed the structure of
compound 7 given in Fig. 1 under the assumption that it is a diarylheptanoid.
Furthermore, another two ions at m/z 137 and m/z 221 (Fig. 3) were produced by
breaking the bond between carbons 5 and 6 and the bond between carbons 6 and 7,
respectively. Thus, the fragmentation behavior of compound 7 confirmed the proposed
structure of this compound. Compound 7, named 5-hydroxy-1,7-bis(3,4-
dihydroxyphenyl)-1-hepten-3-one, was tentatively identified as new compound.
Conclusions
LC-ESI-MS/MS coupled to diode array detection served as a powerful and rapid
tool to identify diarylheptanoids in crude methanolic extracts from fresh turmeric
244 rhizomes. The identified compounds differed by substituent groups on the heptanoid skeleton and/or on the aromatic rings. These differences were ascertained by comparing the chromatographic and mass spectral data from these compounds with those from standard compounds, if available, and against each other. The 19 diarylheptanoids identified from fresh turmeric rhizome in this investigation could also be used as chemical evidence in turmeric authentication investigations, especially if these compounds, or derivatives of them, are found to be specific to turmeric. In addition, metabolic profiling information about these diarylheptanoids will be useful in providing evidence for or against specific pathways in efforts to elucidate the biosynthetic pathway to curcumin and related compounds.
Acknowledgements
The authors acknowledge financial assistance from the National Science
Foundation Plant Genome Program, grant DBI-0227618 to D.R.G., and the National
Institutes of Health NCCAM/ODS, grants #5 P50 AT 000474-05 and 3 P50 AT 000474-
03 S1 to B.N.T. We also thank Veronica Rodriguez, Hyunjo Koo, and Brenda Jackson for assistance with chemical analyses. The contents of this publication are solely the responsibility of the authors and do not necessarily represent the official views of
NCCAM, ODS, or the National Institutes of Health.
245
Figure legends
Fig. 1. Chemical structures and molecular weights of diarylheptanoids identified in turmeric.
Fig.2. (a) and (b) are total ion current (TIC) chromatograms from negative ion (–)ESI-
HPLC-MS and positive ion (+)ESI-HPLC-MS individually; (c) and (d) are HPLC-DAD chromatogram set at 230 nm and 425 nm, respectively, of the crude fresh rhizome extract of turmeric. Note: IND is internal standard compound.
Fig. 3. Negative ion (–)ESI/MS/MS (left) and positive ion (+)ESI/MS/MS (right) of diarylheptanoids. Product ion labels correspond to fragments depicted in Schemes.
Fig. 3. (Cont.) Negative ion (–)ESI/MS/MS (left) and positive ion (+)ESI/MS/MS (right) of diarylheptanoids. Product ion labels correspond to fragments depicted in Schemes.
Scheme 1. a): (–)ESI-MS/MS fragmentation of diarylheptanoids 13, 15A, 15B, and 17
b): (+)ESI-MS/MS fragmentation of diarylheptanoids 13, 15A, 15B, and 17
Scheme 2. a): (–)ESI-MS/MS fragmentation of diarylheptanoid 12
b): (+)ESI-MS/MS fragmentation of diarylheptanoid 12
246
Scheme 3. a): (–)ESI-MS/MS fragmentation of diarylheptanoid 2
b): (+)ESI-MS/MS fragmentation of diarylheptanoid 2
Scheme 4. (–)ESI-MS/MS fragmentation of diarylheptanoid 1
Scheme 5. a): (–)ESI-MS/MS fragmentation of diarylheptanoid 3
b): (+)ESI-MS/MS fragmentation of diarylheptanoids 3, 4, and 5
Scheme 6. (+)ESI-MS/MS fragmentation of diarylheptanoids 9 and 10
Scheme 7. (+)ESI-MS/MS fragmentation of diarylheptanoid 7
247
References
[1] J.C. Tilak, M. Banerjee, H. Mohan, T.P. Devasagayam, Phytother. Res. 18 (2004)
798.
[2] T. Nishiyama, T. Mae, H. Kishida, M. Tsukagawa, Y. Mimaki, M. Kuroda, Y.
Sashida, K. Takahashi, T. Kawada, K. Nakagawa, M. Kitahara, J. Agric. Food
Chem. 53 (2005) 959.
[3] B.B. Aggarwal, A. Kumar, A.C. Bharti, Anticancer Res. 23 (2003) 363.
[4] H. Ahsan, N. Parveen, N.U. Khan, S.M. Hadi, Chem. Biol. Interact. 121 (1999)
161.
[5] A. Asai, T. Miyazawa, J Nutr. 131 (2001) 2932.
[6] W. Chearwae, S. Anuchapreeda, K. Nandigama, S.V. Ambudkar, P. Limtrakul,
Biochem. Pharmacol. 68 (2004) 2043.
[7] S.D. Deodhar, R. Sethi, R.C. Srimal, Indian J. Med. Res. 71 (1980) 632.
[8] L.N. Grinberg, O. Shalev, H.H. Tonnesen, E.A. Rachmilewitz, Int. J. Pharm. 132
(1996) 251.
[9] B. Joe, M. Vijaykumar, B.R. Lokesh, Crit. Rev. Food Sci. Nutr. 44 (2004) 97.
[10] W.C. Jordan, C.R. Drew, J Natl. Med. Assoc. 88 (1996) 333.
[11] G.J. Kelloff, J.A. Crowell, E.T. Hawk, V.E. Steele, R.A. Lubet, C.W. Boone, J.M.
Covey, L.A. Doody, G.S. Omenn, P. Greenwald, W.K. Hong, D.R. Parkinson, D.
Bagheri, G.T. Baxter, M. Blunden, M.K. Doeltz, K.M. Eisenhauer, K. Johnson,
248
G.G. Knapp, D.G. Longfellow, W.F. Malone, S.G. Nayfield, H.E. Seifried, L.M.
Swall, C.C. Sigman, J. Cell Biochem. Suppl. 26 (1996) 54.
[12] T.H. Leu, M.C. Maa, Curr. Med. Chem. 2 (2002) 357.
[13] A. Mazumder, K. Raghavan, J. Weinstein, K.W. Kohn, Y. Pommier, Biochem.
Pharmacol. 49 (1995) 1165.
[14] R.R. Satoskar, S.J. Shah, S.G. Shenoy, Int. J. Clin. Pharmacol. Ther. Toxicol. 24
(1986) 651.
[15] H.H. Toennesen, J.V. Greenhill, Int. J. Pharm. 87 (1992) 79.
[16] H.H. Toennesen, G. Smistad, T. Aagren, J. Karlsen, Int. J. Pharm. 90 (1993) 221.
[17] H.H. Tonnesen, S. Kristensen, L.N. Grinberg, Int. J. Pharm. 110 (1994) 161.
[18] R.P. Bansal, J.R. Bahl, S.N. Garg, A.A. Naqvi, S. Kumar, Pharm. Biol. 40 (2002)
384.
[19] A.P. Gupta, M.M. Gupta, S. Kumar, J. Liq Chromatogr. R. T. 22 (1999) 1561.
[20] X.G. He, L.Z. Lin, L.Z. Lian, M. Lindenmaier, J. Chromatogr. A 818 (1998) 127.
[21] R. Hiserodt, T.G. Hartman, C.T. Ho, R.T. Rosen, J. Chromatogr. A 740 (1996) 51.
[22] K. Inoue, S. Hamasaki, Y. Yoshimura, M. Yamada, M. Nakamura, Y. Ito, H.
Nakazawa, J. Liq. Chromatogr. R. T. 26 (2003) 53.
[23] G.K. Jayaprakasha, L.J.M. Rao, K.K. Sakariah, J. Agr. Food Chem. 50 (2002)
3668.
[24] M. Lechtenberg, B. Quandt, A. Nahrstedt, Phytochem. Analysis 15 (2004) 152.
[25] Y. Pak, R. Patek, M. Mayersohn, J. Chromatogr. B 796 (2003) 339.
249
[26] V.K. Raina, S.K. Srivastava, N. Jain, A. Ahmad, K.V. Syamasundar, K.K.
Aggarwal, Flavour Frag. J. 17 (2002) 99.
[27] R. Richmond, E. PomboVillar, J. Chromatogr. A 760 (1997) 303.
[28] X.H. Sun, C.L. Gao, W.D. Cao, X.R. Yang, E.K. Wang, J. Chromatogr. A 962
(2002) 117.
[29] H. Jiang, A. Somogyi, B.N. Timmermann, D.R. Gang, Submitted to Rapid
Commun. Mass Spectrom. (2005).
[30] H. Jiang, Z. Xie, H. Koo, S.P. McLaughlin, B.N. Timmermann, D.R. Gang,
Submitted to Phytochemistry (2005).
[31] F. Beaudry, J.C.Y. Le Blanc, M. Coutu, I. Ramier, J.P. Moreau, N.K. Brown,
Biomed. Chromatogr. 13 (1999) 363.
[32] J.M. Halket, D. Waterman, A.M. Przyborowska, R.K. Patel, P.D. Fraser, P.M.
Bramley, J. Exp. Bot. 56 (2005) 219.
[33] G.C. Kite, M.J.R. Howes, M.S.J. Simmonds, Rapid Commun. Mass Sp. 18 (2004)
2859.
[34] K. Patel, G. Krishna, E. Sokoloski, Y. Ito, J. Liq. Chromatogr. R. T. 23 (2000)
2209.
[35] F. Zsila, Z. Bikadi, M. Simonyi, Biochem. Bioph. Res. Co. 301 (2003) 776.
[36] S. Park, D. Kim, J. Nat. Prod. 65 (2002) 1227.
[37] V. Ravindranath, M. Satyanarayana, Phytochemistry 19 (1980) 2031.
[38] T. Kita, S. Imai, N. Kamata, N. Tsuge, in, (House Food Industrial Co., Ltd.,
Japan). Application: JP, 2002, p. 14 pp.
250
[39] P. Kamnaing, A. Tsopmo, E. Tanifum, M. Tchuendem, P. Tane, J. Ayafor, O.
Sterner, D. Rattendi, M. Iwu, B. Schuster, C. Bacchi, J. Nat. Prod. 66 (2003) 364.
[40] E. Smite, L. Lundgren, R. Andersson, Phytochemistry 32 (1993) 365.
[41] M. Ali, Y. Tezuka, S. Awale, A. Banskota, S. Kadota, J. Nat. Prod. 64 (2001) 289.
[42] S.Y. Park, D.S.H.L. Kim, J. Nat. Prod. 65 (2002) 1227.
[43] G. Li, C.S. Seo, S.H. Lee, Y. Jahng, H.W. Chang, C.S. Lee, M.H. Woo, J.K. Son,
B. Kor. Chem. Soc. 25 (2004) 397.
[44] E. Nurtjahja-Tjendraputra, A. Ammit, B. Roufogalis, V. Tran, C. Duke, Thromb.
Res. 111 (2003) 259.
251
Fig. 1. Chemical structures and molecular weights of diarylheptanoids identified in turmeric.
252
Fig.2. (a) and (b) are total ion current (TIC) chromatograms from negative ion (–)ESI-
HPLC-MS and positive ion (+)ESI-HPLC-MS individually; (c) and (d) are HPLC-DAD chromatogram set at 230 nm and 425 nm, respectively, of the crude fresh rhizome extract of turmeric. Note: IND is internal standard compound.
253
Fig. 3. Negative ion (–)ESI/MS/MS (left) and positive ion (+)ESI/MS/MS (right) of diarylheptanoids. Product ion labels correspond to fragments depicted in Schemes.
254
Fig. 3. (Cont.) Negative ion (–)ESI/MS/MS (left) and positive ion (+)ESI/MS/MS (right) of diarylheptanoids. Product ion labels correspond to fragments depicted in Schemes.
255
Scheme 1. a): (–)ESI-MS/MS fragmentation of diarylheptanoids 13, 15A, 15B, and 17
b): (+)ESI-MS/MS fragmentation of diarylheptanoids 13, 15A, 15B, and 17
256
Scheme 2. a): (–)ESI-MS/MS fragmentation of diarylheptanoid 12
b): (+)ESI-MS/MS fragmentation of diarylheptanoid 12
257
Scheme 3. a): (–)ESI-MS/MS fragmentation of diarylheptanoid 2
b): (+)ESI-MS/MS fragmentation of diarylheptanoid 2
258
Scheme 4. (–)ESI-MS/MS fragmentation of diarylheptanoid 1
259
Scheme 5. a): (–)ESI-MS/MS fragmentation of diarylheptanoid 3
b): (+)ESI-MS/MS fragmentation of diarylheptanoids 3, 4, and 5
260
Scheme 6. (+)ESI-MS/MS fragmentation of diarylheptanoids 9 and 10
261
Scheme 7. (+)ESI-MS/MS fragmentation of diarylheptanoid 7
262
APPENDIX F
Manuscript “Identification of diarylheptanoids in ginger (Zingiber officinale Rosc.) by
LC-ESI-MS/MS”, submitted to the Journal of Agricultural and Food Chemistry.
263
Title: Identification of diarylheptanoids in ginger (Zingiber officinale Rosc.) by LC-ESI-
MS/MS
Hongliang Jianga,b,c,d, Barbara N. Timmermanna,e, and David R. Ganga,b,c*
aArizona Center for Phytomedicine Research, College of Pharmacy, University of
Arizona, Tucson, AZ 85721, USA
bDepartment of Plant Sciences, College of Agriculture and Life Sciences, University of
Arizona, Tucson, AZ 85721, USA
cBIO5 Institute, University of Arizona, Tucson, AZ 85721, USA
dDepartment of Pharmaceutical Sciences, College of Pharmacy, University of Arizona,
Tucson, AZ 85721, USA
eDepartment of Pharmacology and Toxicology, College of Pharmacy, University of
Arizona, Tucson, AZ 85721, USA
*Correspondence should be addressed to David R. Gang, Department of Plant Sciences and BIO5 Institute, University of Arizona, Tucson, AZ 85721-0036, USA
Tel: 520-621-7154
Fax: 520-621-7186
email: [email protected]
264
ABSTRACT
In our continuing investigation of diarylheptanoids in Zingiberaceae plants using LC-
ESI-MS/MS, 26 diarylheptanoids were identified from fresh ginger rhizome. Of the 26
compounds 15 diarylheptanoids appear to be new compounds. In addition, 18 were
acetylated, which is different from our investigation of diarylheptanoids from turmeric,
another member of the Zingiberaceae, which did not possess any acetylated
diarylheptanoids. Distinct groups (homologous series) of diarylheptanoids were found in
extracts from ginger rhizome. These groups were differentiated by structural differences
on the heptane skeletons, whereas homologs within each group differed by substitution
patterns on the aromatic rings. Diagnostic fragmentation behavior in (+) and (−)ESI-
MS/MS analyses for each group of homologs, as well as information regarding polarity
obtained from retention time data, allowed us to classify compounds by group and
identify them based on key structural features.
KEYWORDS: Zingiber officinale; ginger; diarylheptanoids; LC-ESI-MS/MS
265
INTRODUCTION
Diarylheptanoids belong to a class of natural products with a 1,7-diarylheptane skeleton
(1) that include the curcuminoids, characteristic medicinal compounds of turmeric.
Diarylheptanoids have been found to possess a variety of biological and pharmacological
activities including antioxidant, antihepatotoxic, anti-inflammatory, antiproliferative,
antiemetic, chemopreventive, and antitumor activities (2;3;4;5;6;7;8), leading to
increased interest in recent years in this group of compounds (9;10).
Zingiberaceae plants appear to be the major producers of diarylheptanoids
(2;7;8;11;12;13;14;15;16;17;18;19). Medicinal plants such as turmeric (Curcuma longa
L.) and ginger (Zingiber officinale Rosc.) in the Zingiberaceae are known to produce a
variety of diarylheptanoids (17;20). Numerous analytical methods including high
performance liquid chromatography (HPLC) and its coupling to mass spectrometry (LC-
MS), thin layer chromatography (TLC), and capillary electrophoresis (CE) have been
applied to characterize the major diarylheptanoids (curcuminoids) in turmeric and related
compounds (gingerols) in ginger (11;20;21;22;23;24;25;26). Diarylheptanoids as well as gingerol-related compounds are major classes of biologically active natural products in ginger. However, no on-line analytical method has been reported to characterize and measure these compounds in ginger.
In this investigation, we used both (+) and (−)-ESI-MS/MS coupled to LC in our continuing work to identify diarylheptanoids in Zingiberaceae plants. LC-MS instruments from two different manufacturers were used. Based on our previous work, which
266
investigated the fragmentation mechanisms of standard diarylheptanoids and gingerols in these two instruments (27;28), we were able to identify 26 compounds in ginger extracts as diarylheptanoids. Fifteen of these are new compounds.
MATERIALS AND METHODS
Chemicals and Reagents. HPLC grade acetonitrile and methanol were from Burdick &
Jackson (Muskegon, MI, USA). Formic acid was from J. T. Baker (Mallinkrodt Baker,
Inc., Phillipsburg, NJ, USA). Ammonium formate was from Fisher Scientific (Fair Lawn,
NJ, USA). Deionized water was re-distilled.
Plant Material and Sample Preparation. Fresh ginger rhizomes were collected from plants grown in a greenhouse at the University of Arizona, frozen in liquid nitrogen, and kept at -80 °C until analyzed. Methanolic extracts to be used in LC-MS analysis were produced from fresh frozen ginger rhizome samples essentially as described previously for ginger samples (29), with overnight extraction at room temperature and shaking at
200 rpm.
LC-ESI-MS/MS Analysis of Ginger Extracts. Analysis of ginger extracts was performed on two separate ion trap mass spectrometer systems: i) a ThermoFinnigan
Surveyor MS HPLC coupled to an in-line PDA detector and a ThermoFinnigan LCQ
Advantage ion trap (San Jose, CA, USA), and ii) an Agilent 1100 HPLC system coupled
267
to an in-line DAD detector and an Agilent LC-MSD-Trap-SL ion trap (Palo Alto, CA,
USA).
LC separation of diarylheptanoids—For both instruments, the same column and elution parameters were used. Column: Discovery® HS C18, 3 µm, 15 cm × 2.1 mm (Supelco,
Bellefonte, PA, USA); Guard column: Discovery® HS C18, 3 µm, 2 cm × 2.1 mm
(Supelco); Mobile phase: (A) buffer (5mM ammonium formate, 0.1% formic acid, in
ddH2O) and (B) acetonitrile; Gradient (in buffer A): 0-2 min, 5% B; 2-57 min, 5-100% B;
57-60 min, 100% B; 60-65 min, 100-5% B; 65-75 min, 5% B. Flow rate: 0.25 ml·min-1; temperature, 40 °C; Injection volume, 5 µl.
MS and MS2 Parameters for ThermoFinnigan LCQ Advantage—The acquisition
parameters for positive and negative mode were: sheath gas flow 29 (positive), 36
(negative) (in arbitrary units); aux/sweep gas flow 6 (positive), 0 (negative); source
voltage 5 kV (positive), 4.5 kV (negative); source current 80 µA (positive and negative); capillary voltage 36 V (positive), -32 V (negative); capillary temperature 270 °C; tube lens offset 45 V (positive), -25 V (negative); collision gas pressure ca. 10-5 torr; Q-value
25; mass range measured: 100-1000 m/z.
MS and MS2 Parameters for Agilent LC-MSD-Trap-SL—The acquisition parameters
for positive and negative mode were: drying N2 temperature, 350°C, 10 l/min; nebulizer
pressure 60 psi; HV capillary 4500 V; HV end plate offset -500 V; capillary current 65.9
nA (positive mode), 62.3 (negative mode); end plate current 1482.7 nA (positive mode),
268
1378.7 (negative mode); capillary exit RF amplitude 99.3 V (positive mode), -99.3 V
(negative mode); skimmer 40.0 V (positive mode), -40.0 V (negative mode); mass range
measured: 50-900 m/z.
RESULTS AND DISCUSSION
General LC-MS/MS Approach. LC-ESI-MS/MS has been shown to be a powerful tool for on-line identification of acyclic diarylheptanoids, providing, for example, characteristic fragmentation behaviors for each specific group of diarylheptanoids in
turmeric (30). Compared to turmeric, which is rich in acyclic diarylheptanoids, ginger
produces both acyclic and cyclic diarylheptanoids (4;14;15;16;17;31). Therefore, LC-
ESI-MS/MS can also be used to identify acyclic diarylheptanoids in ginger. In this study,
two LC-ESI-MS/MS instruments from different manufactures were used to characterize
diarylheptanoids in crude methanolic extracts of fresh ginger rhizome, providing
complementary information on detection and determination of these compounds. The
Agilent LC-MSD-Trap-SL ion trap demonstrated better sensitivity in (+)ESI-MS and
provided diagnostic fragments for diarylheptanoids from several groups shown in Figure
1. On the other hand, some diarylheptanoids were only identified in the ThermoFinnigan
LCQ Advantage in (−)-ESI-MS analysis. Most diarylheptanoids identified were detected
by both instruments (Table 1) and both instruments demonstrated the same ionization and
fragmentation behaviors for the diarylheptanoids analyzed. The retention time from the
Agilent instrument showed an approximately 2.5 min delay for all compounds compared
269
to the ThermoFinnigan instrument, due to differences in the pump systems in these two
instruments. However, it was a straight forward task to align and confirm the compounds
detected from both instruments according to both the m/z of the parent ions and relative
retention time in the TIC chromatograms (Table 1). In order to distinguish the retention
a b time from the two instruments in our following description, we assigned tR and tR , respectively, to be the retention time from the ThermoFinnigan and the Agilent instruments. The chromatographic and spectral data for all of the compounds identified are summarized in Table 1.
Characterization of Compounds 1,2, 9 and 10. This group of diarylheptanoids was identified by their fragmentation behavior in both (+) and (−)-MS/MS. All four of these diarylheptanoids possess a common structural moiety, consisting of 5-hydroxy and 3-oxo groups on the heptane skeleton (Figure 1). This structural moiety is also present in the gingerols, whose fragmentation mechanisms in ESI-MS/MS have been studied by our group (27). Compounds 1, 9, and 10 have been previously reported from ginger (14;17), whereas compound 2 was identified as a new compound.
a b Compound 1 with a retention time of 19.9 min (tR ) and 21.9 (tR ) was detected at
− + m/z 389 [M-H] in (−)ESI-MS and at m/z 408 [M+NH4] in (+)ESI-MS, suggesting a
molecular weight of 390 Da. Several diarylheptanoids with a molecular weight of 390 Da have been previously isolated from ginger and compound 1 could be one of these
+ compounds (17). From the (+)ESI-MS/MS spectrum of the [M+NH4] ion, three major
daughter ions at m/z 373 (base peak), 193 and 391 were observed (Table 1). Daughter ion
270
+ + at m/z 391 [M+H] was produced by the loss of NH3. Ion A at m/z 373 [M-H2O+H]
+ (base peak) was obtained by further loss of one H2O from [M+H] , suggesting the presence of an hydroxy group on the heptane skeleton. Among the diarylheptanoids isolated from ginger that have a molecular weight of 390 Da, 5-hydroxy-1-(3,4- dihydroxy-5-methoxyphenyl)-7-(4-hydroxy-3-methoxyphenyl)-3-heptanone (17), possessing a structural moiety of 5-hydroxy and 3-oxo group on the heptane skeleton, was suggested for compound 1 (Figure 1). Another characteristic daughter ion B at m/z
193 was formed by a rearrangement and loss of a neutral moiety from ion A [M-
+ H2O+H] in (+)ESI-MS/MS (Scheme 1). This reaction occurred similarly in (+)ESI-
MS/MS analysis of standard gingerols (27). In (−)ESI-MS/MS, the base peak at m/z 209
(ion D) and another peak at m/z 179 (ion C) can be produced by McLafferty
rearrangement and loss of a neutral moiety from the parent ion [M-H]− with
deprotonation occurring on either aromatic ring (Scheme 1). The McLafferty
rearrangement was also observed for standard gingerols with the same 5-hydroxy and 3- oxo moiety (27). Daughter ion at m/z 371 Da was formed by loss of one H2O in (−)ESI-
MS/MS from the parent ion [M-H]−, supporting the presence of the hydroxy group on the heptane skeleton. Thus, the fragmentation behavior of compound 1 in both (+)-and
(−)ESI-MS/MS confirmed its proposed structure as shown in Figure 1.
a Compound 10 was also detected by both (+) and (−)ESI-MS at 22.7 (tR ) and 25.1
b − (tR ) min in both instruments (Table 1). Parent ions at m/z 373 [M-H] in (−)ESI-MS and
+ at m/z 392 [M+NH4] in (+)ESI-MS were observed, suggesting a molecular weight of 374
Da for this compound. Compared to compound 1, a decrease of 16 Da (H instead of OH)
271
for the parent and some daughter ions in both (+) and (−)ESI-MS/MS, along with the
same fragmentation behavior (Table 1 and Scheme 1), was observed, suggesting that
these compounds are homologs. Therefore, the structure of compound 10, lacking one
hydroxy group on a benzene ring when compared to compound 1, was proposed (Figure
+ + 1). The production of ions at m/z 357 [M-H2O+H] (ion A), 375 [M+H] and 177 (ion B)
− in (+)ESI-MS/MS and at m/z 179 (ion D), 193 (ion C), 355 [M-H2O-H] in (−)ESI-
MS/MS supported the proposed structure of this compound (Scheme 1). Thus, compound
10 was identified as 5-hydroxy-1,7-bis(4-hydroxy-3-methoxyphenyl)-3-heptanone
(Figure 1 and Table 1), which has been previously isolated from ginger (17).
Compound 9, however, was only detected in (+)ESI-MS in the two instruments at
a b 22.3 min (tR ) and 24.6 (tR ) (Table 1), presenting an ion at m/z 422. The retention time
was very close to that for compound 10. In addition, a mass shift of 30 Da (OMe instead
of H) was observed for the parent ion and corresponding daughter ions in (+)ESI-MS/MS
of compound 9 when compared to compound 10, which is a homolog differentiated by
one methoxy group. The structure of compound 9 was proposed as shown in Figure 1.
The formation of its daughter ions at m/z 387 (ion A), 405 [M+H]+, and 207 (ion B) from
+ [M+NH4] support the proposed structure of compound 9 (Scheme 1). Therefore,
compound 9 was identified as 5-hydroxy-1-(4-hydroxy-3,5-dimethoxyphenyl)-7-(4-
hydroxy-3-methoxyphenyl)-3-heptanone (Figure 1 and Table 1), which has been previously isolated from ginger (14).
a Compound 2 (at 20 min (tR )) was only detected by the ThermoFinnigan LCQ
+ Advantage ion trap in both (+)ESI-MS at m/z 378 [M+NH4] and (−)ESI-MS at m/z 359
272
[M-H]−, however, no MS/MS spectra in the positive ion mode were obtained, due to very
low abundance. Compared to compound 1, this compound showed a decrease of 30 Da
(H instead of OMe) for both its parent and daughter ions in the (−)ESI-MS/MS analysis.
In addition, compounds 1 and 2 demonstrated very similar chromatographic behavior,
suggesting that compound 2 is a homolog of compound 1, differentiated by one methoxy
group. A structure of compound 2 was thereby suggested. This proposed structure of
compound 2 (Figure 1) was confirmed by the fragmentation behavior of its parent ion
[M-H]− in (−)ESI-MS/MS (Scheme 1). Therefore, compound 2 was tentatively identified as 5-hydroxy-1-(4-dihydroxy-3-methoxyphenyl)-7-(3,4-dihydroxyphenyl)-3-heptanone
(Figure 1 and Table 1), which has not been previously reported.
Characterization of Compounds 4, 5 and 8. This group of diarylheptanoids was only detected by the Agilent LC-MSD-Trap-SL ion trap. The parent ions were observed in (+) and (−)ESI-MS, however, no MS/MS spectra could be obtained in the negative ion mode.
This group of compounds possesses a common structural moiety of 3,5-dihydroxy on the heptane skeleton (Figure 1). Compounds 5 and 8 have been previously reported from ginger (17). Compound 4 was tentatively identified as a new compound.
b − Compound 5, detected at 23.1 min (tR ), showed a parent ion at m/z 405 [M-H] in
(−)ESI-MS and m/z 407 [M+H]+ in the (+)ESI-MS. These observation suggested a
+ molecular weight of 406 Da. The base peak at m/z 371 [M-2H2O+H] (ion E) and another
+ major peak at m/z 389 [M-H2O+H] in the (+)ESI-MS/MS were produced by the loss of
one or two water molecules from its parent ion [M+H]+, suggesting the presence of two
273
hydroxy groups on the alkyl chain instead of on the aromatic ring. A diarylheptanoid with
a molecular weight of 406 and possessing the two 3,5-dihydroxy moieties on the heptane skeleton has been isolated from ginger (17). The production of another three characteristic daughter ions at m/z 193 (ion G), 167 (ion H), and 217 (ion F) in (+)ESI-
MS/MS confirmed our proposed structure for compound 5 (Figure 1 and Scheme 2) and
its identification as 3,5-dihydroxy-1-(4-hydroxy-3,5-dimethoxyphenyl)-7-(4-hydroxy-3- methoxyphenyl)heptane.
The parent ions of compound 8 were observed at m/z 375 [M-H]− in (−)ESI-MS
and m/z 377 [M+H]+ in the (+)ESI-MS, indicating a molecular weight of 376. When
compared to compound 5, compound 8 showed a decrease of 30 Da (H instead of OMe)
for its parent ion and some of its corresponding daughter ions (Table 1), indicating that it
may be a homolog of compound 5, differing by lack of a methoxy group. In addition, the
similar chromatographic behavior of compound 8 when compared to compound 5, with
b retention time at 23.8 min (tR ), supported the hypothesis that it was a homolog. Therefore, compound 8 was proposed to be 3,5-dihydroxy-1,7-bis(4-hydroxy-3- methoxyphenyl)heptane (Figure 1 and Table 1), which has been previously isolated from ginger (17). The fragmentation behavior of this compound (Scheme 2) in (+)ESI-MS/MS confirmed this suggested structure.
Compared to compound 5, compound 4 demonstrated an increase of 30 Da (+
OMe) for its parent ion and some of its corresponding daughter ions (Table 1), suggesting that its molecular weight was 436 Da and that it could also be a homolog of compound 5.
In addition, its similar retention time (Table 1) also supported this hypothesis. Therefore,
274 compound 4 appeared to be 3,5-dihydroxy-1,7-bis(4-hydroxy-3,5- dimethoxyphenyl)heptane, which has not been previously reported. The formation of its daughter ions in the (+)ESI-MS/MS supported the tentative identification of compound 4
(Scheme 2).
Characterization of Compounds 14, 15, 16, 19, 20, 21, 22, 23, 24, and 25. Except for compound 19, all compounds in this group were detected by the two instruments in both
(+) and (−)ESI-MS analyses. A 3,5-diacetoxy moiety instead of a 3,5-dihydroxy on the heptane skeleton was characteristic of these 10 compounds (Figure 1). The structural differences within this group lie in the pattern of substitution on the aromatic rings. In order to clearly describe the identification of these 10 compounds, we classified them into three subgroups, according to their approximate retention time (Table 1). Subgroup 1 includes compounds 14, 15, and 16, which possess four hydroxy groups on the aromatic rings. Subgroup 2 is composed of compounds 19, 20, 21, and 22, which possess three hydroxy groups on the aromatic rings. Only two hydroxy groups are present on the aromatic rings of compounds 23, 24, and 25, subgroup 3. Among these 10 diarylheptanoids, compounds 14, 15, 19, 20, 21, and 24 have not yet been reported.
Compounds 14, 15, and 16 in subgroup 1 were detected in the two instruments at
a b about 25.0 min (tR ) and 27.3 min (tR ) (Figure 1 and Table 1). Parent ions at m/z 491 [M-
− − − + H] , 461 [M-H] , and 431 [M-H] in (−)ESI-MS and at m/z 510 [M+NH4] , 480
+ + [M+NH4] , and 450 [M+NH4] in (+)ESI-MS were observed for compounds 14, 15, and
16, respectively, suggesting molecular weight of 492, 462, and 432. Compared to
275
compound 16, compounds 14 and 15 showed an increase of 60 Da (+2 OMe) and 30 Da
+ (+ OMe), respectively. In (+)ESI-MS/MS analysis of parent ion m/z 450 [M+NH4] from
compound 16, daughter ions including 433 [M+H]+, 373 [M-AcOH+H]+, and 313 [M-
2AcOH+H]+ were observed, suggesting the presence of two acetyl groups. It appeared
that compound 16 was 3,5-diacetoxy-1,7-bis(3,4-dihydroxyphenyl)heptane, a
diarylheptanoid with a molecular weight of 432 and two acetyl groups, and which has
been previously isolated from ginger (14) (Figure 1). Daughter ion E at m/z 313 [M-
+ 2AcOH+H] from compound 16, produced by the loss of one NH3 and two CH3COOH molecules from the parent ion in (+)ESI-MS/MS (Scheme 3), possessed the same structural skeleton as ion E from compounds 4, 5 and 8 (Scheme 2) which was formed by loss of one NH3 and two H2O instead of two CH3COOH molecules from the
corresponding parent ions. Another diagnostic ion G at m/z 149 was also observed for
compound 16 by further fragmentation of ion E (Table 1 and Scheme 3), which is the
same mechanism as in Scheme 2. In addition, daughter ions at m/z 389 [M-CH2=C=O-
H]−, 371 [M-AcOH-H]−, and 311 [M-2AcOH-H]− produced in the (−)ESI-MS/MS
analysis also confirmed the presence of two acetyl groups in compound 16. Therefore,
the proposed structure of compound 16 was confirmed.
Compared to compound 16, a diagnostic ion G from the parent ion of compound
15 in (+)-ESI-MS/MS, which was present at both m/z 149 and 179, suggested that the
additional methoxy group is present on one aromatic ring (Scheme 3). In (−)ESI-MS/MS
analysis of the parent ion from compound 15, some of the daughter ions showed an
increase of 30 Da compared to the corresponding daughter ions from compound 16. In
276
− − addition, daughter ions at m/z 446 [M-CH3-H] , 386 [M-CH3-AcOH-H] , and 326 [M-
− CH3-2AcOH-H] were also observed (Table 1). This also supported the presence of an
additional methoxy group on the aromatic ring, because these daughter ions produced by
loss of a methyl group could be stabilized by the aromatic ring. Therefore, compound 15
was tentatively identified as 3,5-diacetoxy-1-(3,4-dihydroxyphenyl)-7-(3,4-dihydroxy-5-
methoxyphenyl)heptane (Figure 1 and Table 1). Similarly, compound 14 was tentatively
identified as 3,5-diacetoxy-1,7-bis(3,4-hydroxy-5-methoxyphenyl)heptane by comparing
the m/z of its daughter ions with that of corresponding ions from compound 15 in both (−)
and (+)ESI-MS/MS, showing an increase of 30 Da (+OMe) for all other daughter ions
except for ion G (Table 1). This characteristic ion G was present at only m/z 179,
suggesting that both aromatic rings contained a methoxy group (Scheme 3), and
confirming the proposed structure for compound 14 as shown in Figure 1.
Compounds 19, 20, 21 and 22 in subgroup 2 were found at retention times of
a b approximately 27.8 min (tR ) and 30.2 min (tR ) (Figure 2 and Table 1). Compared to the
compounds in subgroup 1, compounds in subgroup 2 eluted about 2.8 min later,
suggesting lower polarity for these compounds. Parent ions at m/z 505 [M-H]−, 415 [M-
− − − + H] , 475 [M-H] , and 445 [M-H] in the (−)ESI-MS and at m/z 524 [M+NH4] , 434
+ + + [M+NH4] , 494 [M+NH4] , and 464 [M+NH4] in the (+)ESI-MS analyses were detected
for compounds 19, 20, 21, and 22, respectively, suggesting molecular weights of 506,
416, 476, and 446 Da. Compared to compound 20, compounds 19, 21, and 22 showed an increase of 90 Da (+3 OMe), 60 Da (+2 OMe), and 30 Da (+ OMe), respectively. This suggested that, as for compounds in subgroup 1, compounds in subgroup 2 may also
277
belong to a homologous series, differentiated by the number of methoxy groups on the aromatic rings. By comparing the parent and their daughters ions for compounds in subgroup 2 with those for compounds in subgroup 1, a difference of 16 Da (H instead of
OH, respectively) was observed between the three pairs of compounds, including compounds 14 and 21, compounds 15 and 22, and compounds 16 and 20 (Table 1). This suggested that compounds in subgroup 1 and compounds in subgroup 2 were also homologs, differentiated by lack of one hydroxy moiety in subgroup 2 when compared to subgroup 1. Meanwhile, the lower polarity of compounds in subgroup 2, which possess three hydroxy groups instead of four, was also thereby explained. Based on this
information, compounds 19, 20, 21, and 22 were tentatively identified as 3,5-diacetoxy-7-
(3,4-dihydroxy-5-methoxyphenyl)-1-(4-hydroxy-3,5-dimethoxyphenyl)heptane; 3,5-
diacetoxy-7-(4-hydroxyphenyl)-1-(3,4-dihydroxyphenyl)heptane; 3,5-diacetoxy-7-(4-
hydroxy-3-methoxyphenyl)-1-(3,4-dihydroxy-5-methoxyphenyl)heptane; and 3,5-
diacetoxy-7-(3,4-dihydroxyphenyl)-1-(4-hydroxy-3-methoxyphenyl)heptane, respectively.
a Compounds 23, 24, and 25 in subgroup 3 were detected at about 31.0 min (tR )
b and 33.3 min (tR ), which was about 3.2 min later than compounds in subgroup 2 (Figure
2 and Table 1). This suggested lower polarity for this subgroup. The parent ions of compounds 23, 24, and 25 were observed at m/z 489 [M-H]−, 429 [M-H]−, and 459 [M-
− + + + H] in the (−)ESI-MS and at m/z 508 [M+NH4] , 448 [M+NH4] , and 478 [M+NH4] in
(+)ESI-MS analyses (Table 1), suggesting molecular weights of 490, 430, and 460 Da,
respectively. Compared to compound 24, compounds 23 and 25 showed an increase of 60
Da (+2 OMe) and 30 Da (+ OMe), respectively. As for compounds in subgroup 2, these
278
three compounds were characterized by comparing their fragmentation behavior with
compounds in subgroups 1 and 2. Compounds 23, 24, and 25 demonstrated the same
fragmentation pattern as compounds 19, 22, and 21, respectively, with a difference of 16
Da (H instead of OH, respectively) for their corresponding parent ions and daughter ions,
and suggesting that compounds in subgroup 3 with two hydroxy groups and compounds
in subgroup 2 with three hydroxy groups were also homologs. Compounds 23, 24, and 25
were thereby identified as 3,5-diacetoxy-1-(4-hydroxy-3,5-dimethoxyphenyl)-7-(4-
hydroxy-3-methoxyphenyl)heptane; 3,5-diacetoxy-7-(4-hydroxyphenyl)-1-(4-hydroxy-3-
methoxyphenyl)heptane; and 3,5-diacetoxy-1,7-bis(4-hydroxy-3-methoxyphenyl)heptane,
respectively.
Characterization of Compounds 3, 6, 7, 11, 12, 13, 17, and 18. Due to low abundance
for compounds in this group and/or differences in instrument sensitivity, only compounds
7 and 13 were detected by both instruments. Compounds 17 and 18 were only seen when
analyzed by in the Agilent instrument. On the other hand, compounds 3, 6, 11, and 12
were only observed in the ThermoFinnigan instrument. Although all compounds were
detected by both (+) and (−)ESI-MS analysis, only compounds 7 and 13 offered good
quality MS/MS spectra in both positive and negative ionization modes (Table 1). As a
result, characterization of this group is based mostly on compounds 7 and 13. This group
possesses 3-acetoxy and 5-hydroxy moieties on the heptane skeleton (Figure 1).
Compound 18 has been previously isolated from ginger (17)and the other seven
compounds have not been previously reported.
279
− + Based on the m/z of the parent [M-H] and [M+NH4] ions of these compounds,
molecular weights of 420, 450, 390, 404, 374, 434, 448, and 418 Da for compounds 3, 6,
7, 11, 12, 13, 17, and 18, respectively, were determined (Figure 1 and Table 1). In
(−)ESI-MS/MS analysis, fragmentation spectra of the parent ions [M-H]− showed a base peak of [M-AcOH-H]−, suggesting the presence of an acetoxy group. Ions of [M+H]+,
+ + [M-H2O+H] , and [M-H2O-AcOH+H] were observed in (+)ESI-MS/MS analysis of
+ parent ion [M+NH4] , suggesting the presence of one hydroxy and one acetoxy on the
alkyl chain of these compounds. In addition, the corresponding diarylheptanoids with an
addition of 42 Da (CH3COO instead of OH), possessing a 3,5-diacetoxy moiety on the heptane skeleton, were described in the previous subsection for all of the compounds in this group. This group of compounds was therefore proposed to be the mono-deacetylated homologs of compounds described in the previous subsection (Figure 1).
The structures of compounds 7, 13, 17, and 18 were confirmed by the
+ fragmentation behavior of parent ion [M+NH4] in (+)ESI-MS/MS analysis (Table 1 and
Scheme 3). Therefore, they were tentatively identified as 3-acetoxy-5-hydroxy-1,7-
bis(3,4-dihydroxyphenyl)heptane; 3-acetoxy-5-hydroxy-1-(4-hydroxy-3-
methoxyphenyl)-7-(3,4-dihydroxy-5-methoxyphenyl)heptane; 3-acetoxy-5-hydroxy-1-(4-
hydroxy-3-methoxyphenyl)-7-(4-hydroxy-3,5-dimethoxyphenyl)heptane; and 3-acetoxy-
5-hydroxy-1,7-bis(4-hydroxy-3-methoxyphenyl)heptane, respectively.
Compounds 3 and 6, lacking (+)-ESI-MS/MS data, were tentative identified by
comparing their molecular weights, mass spectral data from (−)ESI-MS/MS, and
chromatographic behavior with those from compound 7 (Table 1). These three
280
compounds, differentiated by 30 Da (+ OMe) and 60 Da (+2 OMe), respectively, showed
a close retention at about 20.8 min (tR ) and similar fragmentation behavior in (−)ESI-
− MS/MS analysis. Ion [M-CH3-H] from compounds 3 and 6 was also observed in (−)ESI-
MS/MS, supporting the presence of methoxy group on the aromatic ring. However, for
compound 3, assignment of the methoxy group on the appropriate aromatic ring was not
possible and two possible names (Table 1) and structures (Figure 1) are possible for
compound 3. Compound 6 were tentatively identified as 3-acetoxy-5-hydroxy-1,7-
bis(3,4-dihydroxy-5-methoxyphenyl)-heptane.
Similarly, compounds 11 and 12 were tentatively identified by comparing their
chromatographic behavior and mass spectral data from (−)ESI-MS/MS analysis with
those from compound 13. These three homologs, differentiated by 30 Da (+ OMe) and 60
a Da (+2 OMe), respectively, demonstrated similar retention times at about 23.5 min (tR )
and similar fragmentation patterns in the (−)ESI-MS/MS analysis (Table 1). Compounds
11 and 13, with the presence of methoxy groups on the aromatic rings, also showed
− daughter ions [M-CH3-H] . Based on the structure of compound 13, two possible
structures were also suggested for both compounds 11 and 12 due to the lack of firm evidence of assignment of the methoxy or hydroxy group on one of the two aromatic rings (Figure 1 and Table 1).
a Characterization of Compound 26. Compound 26 (at 31.9 min (tR )) was only detected
in the (−)ESI-MS analysis. This compound was identified by its fragmentation behavior in (−)ESI-MS/MS and its retention time. This compound has also been identified in
281 extracts from turmeric rhizome samples and its fragmentation mechanism in (−)ESI-
MS/MS has been described (30). Therefore, compound 26, identified as dihydrocurcumin, is the only diarylheptanoid identified in our investigations by LC-ESI-MS/MS of diarylheptanoids that accumulates in both ginger and turmeric rhizomes.
ACKNOWLEDGMENTS
The authors acknowledge financial assistance from the National Science Foundation
Plant Genome Program, grant DBI-0227618 to D.R.G., and the National Institutes of
Health NCCAM/ODS, grants #5 P50 AT 000474-05 and 3 P50 AT 000474-03 S1 to
B.N.T. We also thank Veronica Rodriguez, Hyunjo Koo, and Brenda Jackson for assistance with chemical analysis. The contents of this publication are solely the responsibility of the authors and do not necessarily represent the official views of
NCCAM, ODS, or the NIH.
282
LITERATURE CITED
(1) Roughley, P. J.; Whiting, D. A. Diarylheptanoids. problems of the biosynthesis.
Tetrahedron Lett. 1971, 3741-6.
(2) Ali, M. S.; Banskota, A. H.; Tezuka, Y.; Saiki, I.; Kadota, S. Antiproliferative
activity of diarylheptanoids from the seeds of Alpinia blepharocalyx. Biol. Pharm.
Bull. 2001, 24, 525-8.
(3) Flynn, D. L.; Rafferty, M. F.; Boctor, A. M. Inhibition of 5-
hydroxyeicosatetraenoic acid (5-HETE) formation in intact human neutrophils by
naturally occurring diarylheptanoids: inhibitory activities of curcuminoids and
yakuchinones. Prostag. Leuk. Med. 1986, 22, 357-60.
(4) Hikino, H.; Kiso, Y.; Kato, N.; Hamada, Y.; Shioiri, T.; Aiyama, R.; Itokawa, H.;
Kiuchi, F.; Sankawa, U. Series on liver-protective drugs. 25. antiheptatotoxic
actions of gingerols and diarylheptanoids. J. Ethnopharmacol. 1985, 14, 31-39.
(5) Ishida, J.; Kozuka, M.; Tokuda, H.; Nishino, H.; Nagumo, S.; Lee, K. H.; Nagai,
M. Chemopreventive potential of cyclic diarylheptanoids. Bioorgan. Med. Chem.
2002, 10, 3361-3365.
(6) Ishida, J.; Kozuka, M.; Wang, H.; Konoshima, T.; Tokuda, H.; Okuda, M.; Yang
Mou, X.; Nishino, H.; Sakurai, N.; Lee, K. H.; Nagai, M. Antitumor-promoting
effects of cyclic diarylheptanoids on Epstein-Barr virus activation and two-stage
mouse skin carcinogenesis. Cancer Lett. 2000, 159, 135-40.
(7) Masuda, Y.; Kikuzaki, H.; Hisamoto, M.; Nakatani, N. Antioxidant properties of
283
gingerol related compounds from ginger. Biofactors 2004, 21, 293-296.
(8) Shin, D.; Kinoshita, K.; Koyama, K.; Takahashi, K. Antiemetic principles of
Alpinia officinarum. J. Nat. Prod. 2002, 65, 1315-8.
(9) Egan, M. E.; Pearson, M.; Weiner, S. A.; Rajendran, V.; Rubin, D.; Glockner-
Pagel, J.; Canny, S.; Du, K.; Lukacs, G. L.; Caplan, M. J. Curcumin, a major
constituent of turmeric, corrects cystic fibrosis defects. Science 2004, 304, 600-
602.
(10) Sharma, R. A.; Euden, S. A.; Platton, S. L.; Cooke, D. N.; Shafayat, A.; Hewitt, H.
R.; Marczylo, T. H.; Morgan, B.; Hemingway, D.; Plummer, S. M.; Pirmohamed,
M.; Gescher, A. J.; Steward, W. P. Phase I clinical trial of oral curcumin:
biomarkers of systemic activity and compliance. Clin. Cancer Res. 2004, 10,
6847-6854.
(11) He, X. G.; Lin, L. Z.; Lian, L. Z.; Lindenmaier, M. Liquid chromatography
electrospray mass spectrometric analysis of curcuminoids and sesquiterpenoids in
turmeric (Curcuma longa). J. Chromatogr. A 1998, 818, 127-132.
(12) Kadota, S.; Tezuka, Y.; Prasain, J. K.; Ali, M. S.; Banskota, A. H. Novel
diarylheptanoids of Alpinia blepharocalyx. Curr. Top. Med. Chem. 2003, 3, 203-
225.
(13) Kamnaing, P.; Tsopmo, A.; Tanifum, E. A.; Tchuendem, M. H. K.; Tane, P.;
Ayafor, J. F.; Sterner, O.; Rattendi, D.; Iwu, M. M.; Schuster, B.; Bacchi, C.
Trypanocidal diarylheptanoids from Aframomum letestuianum. J. Nat. Prod. 2003,
66, 364-367.
284
(14) Kikuzaki, H.; Kobayashi, M.; Nakatani, N. Constituents of Zingiberaceae. Part 4.
Diarylheptanoids from rhizomes of Zingiber officinale. Phytochemistry 1991, 30,
3647-3651.
(15) Kikuzaki, H.; Usuguchi, J.; Nakatani, N. Constituents of Zingiberaceae. Part 1.
Diarylheptanoids from the rhizomes of ginger (Zingiber officinale Roscoe). Chem.
Pharm. Bull. 1991, 39, 120-122.
(16) Kikuzaki, H.; Nakatani, N. Cyclic diarylheptanoids from rhizomes of Zingiber
officinale. Phytochemistry 1996, 43, 273-277.
(17) Ma, J. P.; Jin, X. L.; Yang, L.; Liu, Z. L. Diarylheptanoids from the rhizomes of
Zingiber officinale. Phytochemistry 2004, 65, 1137-1143.
(18) Matsuda, H.; Morikawa, T.; Ninomiya, K.; Yoshikawa, M. Hepatoprotective
constituents from Zedoariae rhizoma: absolute stereostructures of three new
carabrane-type sesquiterpenes, curcumenolactones A, B, and C. Bioorgan. Med.
Chem. 2001, 9, 909-916.
(19) Suksamrarn, A.; Eiamong, S.; P., P.; Charoenpiboonsin, J. Phenolic
diarylheptanoids from Curcuma xanthorrhiza. Phytochemistry 1994, 36, 1505-
1508.
(20) He, X. G.; Bernart, M. W.; Lian, L. Z.; Lin, L. Z. High-performance liquid
chromatography electrospray mass spectrometric analysis of pungent constituents
of ginger. J. Chromatogr. A 1998, 796, 327-334.
(21) Gong, F.; Fung, Y. S.; Liang, Y. Z. Determination of volatile components in
ginger using gas chromatography-mass spectrometry with resolution improved by
285
data processing techniques. J. Agr. Food Chem. 2004, 52, 6378-6383.
(22) Gupta, A. P.; Gupta, M. M.; Kumar, S. Simultaneous determination of
curcuminoids in Curcuma samples using high performance thin layer
chromatography. J. Liq. Chromatogr. R. T. 1999, 22, 1561-1569.
(23) Huang, H. Y.; Kuo, K. L.; Hsieh, Y. Z. Determination of cinnamaldehyde,
cinnamic acid, paeoniflorin, glycyrrhizin and [6]-gingerol in the traditional
Chinese medicinal preparation Kuei-chih-tang by cyclodextrin-modified micellar
electrokinetic chromatography. J. Chromatogr. A 1997, 771, 267-274.
(24) Jayaprakasha, G. K.; Rao, L. J. M.; Sakariah, K. K. Improved HPLC method for
the determination of curcumin, demethoxycurcumin, and bisdemethoxycurcumin.
J. Agr. Food Chem. 2002, 50, 3668-3672.
(25) Jolad, S. D.; Lantz, R. C.; Solyom, A. M.; Chen, G. J.; Bates, R. B.; Timmermann,
B. N. Fresh organically grown ginger (Zingiber officinale): composition and
effects on LPS-induced PGE(2) production. Phytochemistry 2004, 65, 1937-1954.
(26) Lechtenberg, M.; Quandt, B.; Nahrstedt, A. Quantitative determination of
curcuminoids in Curcuma rhizomes and rapid differentiation of Curcuma
domestica Val. and Curcuma xanthorrhiza Roxb. by capillary electrophoresis.
Phytochem. Analysis 2004, 15, 152-158.
(27) Jiang, H.; Somogyi, A.; Timmermann, B. N.; Gang, D. R. Instrument dependence
of ESI ionization and MS/MS fragmentation of the gingerols: a cautionary tale for
metabolomics investigations. Submitted to J. Am. Soc. Mass Spectr. 2005.
(28) Jiang, H.; Somogyi, A.; Timmermann, B. N.; Gang, D. R. Analysis of
286
curcuminoids by positive and negative electrospray ionization and tandem mass
spectrometry. Submitted to Rapid Commun. Mass Spectrom. 2005.
(29) Jiang, H.; Xie, Z.; Koo, H.; McLaughlin, S. P.; Timmermann, B. N.; Gang, D. R.
Metabolic profiling and phylogenetic analysis of medicinal Zingiber species: tools
for authentication of ginger (Zingiber officinale Rosc.). Phytochemistry 2005, In
Press.
(30) Jiang, H.; Timmermann, B. N.; Gang, D. R. Use of LC-ESI-MS/MS to identify
diarylheptanoids in turmeric (Curcuma longa L.) rhizome. Submitted to J.
Chromatogr. A 2005.
(31) Ma, J. P.; Jin, X. L.; Yang, L.; Liu, Z. L. Two new diarylheptanoids from the
rhizomes of Zingiber officinale. Chinese Chem. Lett. 2004, 15, 1306-1308.
Table 1. Chromatographic and mass spectral characteristics of compounds detected by LC-ESI-MS in extracts from ginger rhizome
a b Negative ESI Positive ESI tR tR 2 Compound - Fragment ions in + + Fragment ions in MS (min) (min) [M-H] 2 c [M+H] [M+NH3+H] c number Compound name MS (m/z) (m/z) 5-hydroxy - 1- (3,4 - dihydroxy - 5 - methoxyphenyl) - 7 - (4 - 19.9 21.9 389 209, 195, 371, 179 408 373, 193, 391 1 hydroxy-3-methoxyphenyl)-3-heptanone
d 5-hydroxy-1-(4-dihydroxy-3-methoxyphenyl)-7-(3,4- 20.0 N/A 359 179, 165, 341 378 N/A 2 dihydroxyphenyl)-3-heptanone 3-acetoxy-5-hydroxy-1-(3,4-dihydroxyphenyl)-7-(3,4-dihydroxy-5- d methoxyphenyl)heptane; 20.5 N/A 419 359, 404, 344, 377 438 N/A 3 or 3-acetoxy-5-hydroxy-1-(3,4-dihydroxy-5-methoxyphenyl)-7- (3,4-dihydroxyphenyl)heptane
N/A 22.7 435 N/A 437 401, 419, 193, 167, 247 4d 3,5-dihydroxy-1,7-bis(4-hydroxy-3,5-dimethoxyphenyl)
3,5-dihydroxy-1-(4-hydroxy-3,5-dimethoxyphenyl)-7-(4-hydroxy- N/A 23.1 405 N/A 407 371, 389, 193, 167, 217 5 3-methoxyphenyl)heptane
d 3-acetoxy-5-hydroxy-1,7-bis(3,4-dihydroxy-5-methoxyphenyl)- 21.1 N/A 449 389, 434, 374 468 N/A 6 heptane
21.3 23.6 389 329, 347 408 373, 391, 313, 331, 149 7d 3-acetoxy-5-hydroxy-1,7-bis(3,4-dihydroxyphenyl)-heptane
N/A 23.8 375 N/A 377 341, 359, 163, 217, 137 8 3,5-dihydroxy-1,7-bis(4-hydroxy-3-methoxyphenyl)heptane 5-hydroxy-l-(4-hydroxy-3,5-dimethoxyphenyl)-7-(4-hydroxy-3- 22.3 24.6 403 N/A 422 387, 405, 207 9 methoxyphenyl)-3-heptanone
22.7 25.1 373 179, 165, 193, 355 392 357, 375, 177 10 5-hydroxy-l,7-bis(4-hydroxy-3-methoxyphenyl)-3-heptanone
3-acetoxy-5-hydroxy-1-(4-hydroxy-3-methoxyphenyl)-7-(3,4- d dihydroxyphenyl)heptane; 23.2 N/A 403 343, 388, 361 422 N/A 11 or 3-acetoxy-5-hydroxy-1-(3,4-dihydroxyphenyl)-7-(4-hydroxy-3- methoxyphenyl)heptane 3-acetoxy-5-hydroxy-1-(4-hydroxyphenyl)-7-(3,4- d dihydroxyphenyl)heptane; 23.6 N/A 373 313 392 N/A 12 or 3-acetoxy-5-hydroxy-1-(3,4-dihydroxyphenyl)-7-(4- hydroxyphenyl)heptane
d 3-acetoxy-5-hydroxy-1-(4-hydroxy-3-methoxyphenyl)-7-(3,4- 23.8 26.2 433 373, 358, 418, 391 452 417, 435, 357, 179 13 dihydroxy-5-methoxyphenyl)heptane 287 431, 371, 476, 416, d 3,5-diacetoxy-1,7-bis(3,4-dihydroxy-5-methoxyphenyl)heptane 24.8 27.0 491 510 433, 373, 179, 493 14 356, 449, 389
401, 341, 446, 386, d 3,5-diacetoxy-1-(3,4-dihydroxyphenyl)-7-(3,4-dihydroxy-5- 25.0 27.3 461 480 403, 343, 179, 463, 149 15 419, 326 methoxyphenyl)heptane 25.1 27.4 431 371, 311, 389 450 373, 313, 433, 149 16 3,5-diacetoxy-1,7-bis(3,4-dihydroxyphenyl)heptane
d 3-acetoxy-5-hydroxy-1-(4-hydroxy-3-methoxyphenyl)-7-(4- N/A 27.6 447 N/A 466 431, 449, 371, 389, 193 17 hydroxy-3,5-dimethoxyphenyl)heptane
3-acetoxy-5-hydroxy-1,7-bis(4-hydroxy-3- N/A 28.3 417 N/A 436 401, 419, 341, 163 18 methoxyphenyl)heptane 445, 413, 385, 490, d 3,5-diacetoxy-7-(3,4-dihydroxy-5-methoxyphenyl)-1-(4-hydroxy- 27.3 N/A 505 430, 473, 463, 353, 524 447, 387, 507, 179, 193 19 3,5-dimethoxyphenyl)heptane, 403, 370
d 3,5-diacetoxy-7-(4-hydroxyphenyl)-1-(3,4- 27.6 30.0 415 355, 373, 295, 313 434 357, 417, 149 20 dihydroxyphenyl)heptane
415, 355, 460, 433, d 3,5-diacetoxy-7-(4-dihydroxy-3-methoxyphenyl)-1-(3,4- 28.1 30.5 475 494 417, 357, 477, 179,163 21 400, 340, 373 dihydroxy-5-methoxyphenyl)heptane
385, 325, 403, 430, 3,5-diacetoxy-7-(3,4-dihydroxyphenyl)-1-(4-hydroxy-3- 28.3 30.7 445 464 387, 327, 447, 163, 149 22 370, 343 methoxyphenyl)heptane
3,5-diacetoxy-1-(4-hydroxy-3,5-dimethoxyphenyl)-7-(4-hydroxy- 30.7 33.1 489 N/A 508 431, 371,491, 193, 163 23 3-methoxyphenyl)heptane
d 3,5-diacetoxy-7-(4-hydroxyphenyl)-1-(4-hydroxy-3- 30.9 33.3 429 N/A 448 371, 311, 163, 431 24 methoxyphenyl)heptane 31.3 33.6 459 N/A 478 401, 341, 163, 461 25 3,5-diacetoxy-1,7-bis(4-hydroxy-3-methoxyphenyl)heptane 31.9 N/A 369 219, 233, 149, 175 N/A 26 dihydrocurcumin Note: a Retention time in the TIC chromatogram obtained from ThermoFinnigan LCQ Advantage ion trap mass spectrometer coupled with HPLC; b Retention time in the TIC chromatogram obtained from Agilent LC-MSD-Trap-SL coupled with HPLC; c Daughter ions shown in each row are given in the order of their relative abundance: the first ion, in each case, is the most abundant; 288 d Compound was tentatively identified as the proposed chemical structure shown in Figure 1 and has not yet been reported.
289
O OH OH OH
R1 R2 R1 R2 MW=390 (1, R1=Ar4, R2=Ar3) MW=436 (4*, R1=R2=Ar5) MW=360 (2*, R1=Ar2, R2=Ar3) MW=406 (5, R1=Ar5, R2=Ar3) MW=404 (9, R1=Ar5, R2=Ar3) MW=376 (8, R1= R2=Ar3) MW=374 (10, R1=R2=Ar3)
OAc OAc OH OAc
R1 R2 R1 R2
MW=492 (14*, R1=R2=Ar4) MW=420 (3*, R1=Ar4, R2=Ar2; or R1=Ar2, R2=Ar4) MW=462 (15*, R1=Ar4, R2=Ar2) MW=450 (5*, R1= R2=Ar4) MW=432 (16, R1= R2=Ar2) MW=390 (6*, R1=R2=Ar2) MW=506 (19*, R1=Ar4, R2=Ar5) MW=404 (11*, R1=Ar2, R2=Ar3; or R1=Ar3, R2=Ar2) MW=416 (20*, R1=Ar2, R2=Ar1) MW=374 (12*, R1=Ar2, R2=Ar1; or R1=Ar2, R2=Ar1) MW=476 (21*, R1=Ar4, R2=Ar3) MW=434 (13*, R1=Ar4, R2=Ar3) MW=446 (22, R1=Ar2, R2=Ar3) MW=448 (17*, R1=Ar5, R2=Ar3) MW=490 (23, R1=Ar5, R2=Ar3) MW=418 (18, R1=R2=Ar3) MW=430 (24*, R1=Ar3, R2=Ar1) MW=460 (25, R1=R2=Ar3) O O
R1 R2 MW=370 (26, R1= R2=Ar3)
MeO MeO
HO HO HO HO HO
OH OMe OH OMe
Ar1 Ar2 Ar3 Ar4 Ar5
Figure 1. Chemical structures and molecular weights of diarylheptanoids identified in ginger rhizome. Note: the corresponding names of these 26 compounds are provided in
Table 1; * indicates new compound.
290
A Total Ion Current (TIC) Chromatogram from negative ion (-)ESI-HPLC-MS 7 X10 20,21,22 14,15,16 1.6
1.2 6,7 13 11,12 26 0.8 1,2,3 19
0.4 0.0
Abundance B Total Ion Current (TIC) Chromatogram from negative ion (-)ESI-HPLC-MS 7 23, 24 X10 20,21,22 4.0 25 6,7 10 14,15,16
Absolute 3.0 11,12 3 9 13 2.0 19 1,2 1.0
0.0
6 C Total Ion Current (TIC) Chromatogram from negative ion (-)ESI-HPLC-MS X10 20,21,22 1.5 9 13 7,8 10 15,16 23 1.0 5 1 18 24,25 0.5 4
0.0
Abundance D Total Ion Current (TIC) Chromatogram from negative ion (-)ESI-HPLC-MS 8 X10 15,16,17 1.0 9 20,21,22 10 7,8 24,25
Absolute 5 14 18 23 0.5 1 4 13
0.0 010203040506070 Time
Figure 2. LC-MS analysis of the crude ginger rhizome extract reveals the presence of a
large number of diarylheptanoids. A and B are total ion current (TIC) chromatograms
from negative ion (-)ESI-HPLC-MS and positive ion (+)ESI-HPLC-MS analysis,
respectively, in a ThermoFinnigan LCQ Advantage ion trap coupled to HPLC; C and D are TIC chromatograms from negative ion (-)ESI-HPLC-MS and positive ion (+)ESI-
HPLC-MS analysis, respectively, in an Agilent LC-MSD-Trap-SL ion trap coupled to
HPLC.
291
A O O H R2 + O + [M+NH ]+ + 4 [M+H] R1 R2 R1 + R2 R1 408 (1) 391 (1) [M+H-H O]+ 422 (9) 405 (9) 2 392 (10) 375 (10) 373 (1) 193 (1) 387 (9) A 207 (9) B 357 (10) 177 (10)
B O OH R3 OMe
390 (1, R3=OH, R4=OMe) HO 360 (2, R3=OH, R4=H) OH 374 (10, R3=OMe, R4=H) R4
H H O O O O
R3 OMe MeO R3
389 (1, R3=OH, R4=OMe) 389 (1, R3=OH, R4=OMe) OH 359 (2, R3=OH, R4=H) O HO 359 (2, R3=OH, R4=H) O 373 (10, R3=OMe, R4=H) 373 (10, R3=OMe, R4=H) R4 R4
OH O R3 OMe H 179 (1) 209 (1) O 179 (2) C O 179 (2) D 179 (10) 193 (10) R4
Scheme 1. A): (+)ESI fragmentation of diarylheptanoids 1, 9, and 10
B): (−)ESI fragmentation of diarylheptanoids 1, 2, and 10
292
+ OH2 OH OH + -H2O + CH2 R1 R2 R1 R2 R1 437(4) 419(4) 167(4) 407(5) 389(5) 167(5) H 137(8) 377(8) 359(8) -H2O
+ R1 + + H R1 R2 401(4) 371(5) E 341(8) R2 R2 247(4) 217(5) F 217(8) + R2 + R2 + R1 R1 R1 H 193(4) 193(5) G 163(8)
Scheme 2. (+)ESI fragmentation of diarylheptanoids 4, 5, and 8
293
+ AcOH [M-AcOH+H] AcOH + + + 1 A [M+NH4] [M+H] R1 R2 2 + [M-H2O+H] + + H2O AcOH [M-2AcOH+H] OR [M-AcOH-H2O+H] E
R2 +
+ + R2 R1 R1 G R1 H
[M-H]- B CH3 CH2C=O H O - - 2 - [M-CH3-H] [M-AcOH-H] [M-CH3C=O] CH2C=O AcOH AcOH AcOH - - - [M-CH3-AcOH-H] [M-2AcOH-H] [M-CH3C=O-AcOH] H2O AcOH CH3 [M-CH -2AcOH-H]- 3
Scheme 3. A): (+)ESI fragmentation of diarylheptanoids (14, 15, 16, 19, 20, 21, and 22 fragmented via reaction 1; 7, 13, 17, and 18 fragmented via reaction 2)
B): (−)ESI fragmentation of diarylheptanoids (the fragmentation of 3, 6, 7, 11,
12, and 13 were indicated using only solid arrows; the fragmentation of 14, 15, 16, 19, 20,
21, 22, 23, 24, and 25 were indicated using both solid and dashed arrows). Note: the m/z of each ion is shown in Table 1.