UNIVERSITY OF CINCINNATI
Date:______
I, ______, hereby submit this work as part of the requirements for the degree of: in:
It is entitled:
This work and its defense approved by:
Chair: ______
Characterization of Biologically Important Volatile and Non- Volatile Molecules via Heteroatom Determination Using Chromatography and Mass Spectrometry
A dissertation submitted to the
Division of Research and Advanced Studies of the University of Cincinnati
in partial fulfillment of the requirements for the degree of
Doctor of Philosophy
in the Department of Chemistry of the College of Arts and Sciences
2006
by
Monika Shah
B.Sc., Chemistry, Banaras Hindu University, 1999 M.Sc., Chemistry, Banaras Hindu University, 2001
Committee Chair: Professor Joseph A. Caruso ABSTRACT OF DISSERTATION
Interest in molecule specific characterization of metal and metalloid containing species within a biological system is gaining importance in recent years. This not only facilitates the establishment of toxic or beneficial effects of certain form of elements, but in certain cases paves
way to understanding of biotransformation of elements and also helps to assess the fate of the
environmental contaminants within a biological system. Coupled techniques involving
hyphenation of separation methods such as high performance liquid chromatography and gas
chromatography with element/molecule specific detection has become the most powerful tool in
determination of metal(loid) containing species. Heteroatom determination using mass
spectrometric techniques such as inductively coupled plasma mass spectrometry (based on
element selective detection) and others such as ESI, EI and MALDI (based on element specific
isotope patterns) have been used for the complete characterization of the species of interest.
Heteroelements such as selenium, iodine and phosphorus are known to be important atoms in a
wide variety of molecules of biological interest ranging from small volatiles and non-volatile
amino acids to the macromolecules such as proteins. The aim of this dissertation is characterization of such species in various complex matrices using the above techniques. These include determination and identification of selenium containing amino acids in green onions, characterization of iodine containing species in commercially available seaweeds, studying the novel selenium volatiles in the green onions and localization of the high molecular weight selenium containing species in the mung bean sprouts. On the basis of the results, conclusions are drawn about their chemical association with biological molecules, both micro (volatile and non-volatile forms), macro molecules and their metabolites in the studied system. Another aspect of this dissertation is method development for the fast and sensitive determination of anthropogenic molecules such as phosphorus containing fire retardants/plasticizer so that their fate from the plastic material to the blood plasma can be studied. ACKNOWLEDGEMENTS
The course of my dissertation work at University of Cincinnati (summer of 2002 - spring of
2006) has been nothing but a path to constant learning and growing, not only as a better researcher, but also as a better person. There are many people who I think contributed toward it to make it a very memorable ride. It is a pleasant task that I have now the opportunity to express my gratitude for all of them.
The first person I would like to acknowledge is my husband, Arpit Das. He is the one
without whom this journey had never been started and also never had come to an end. He
strongly supported my decision to join graduate school in the first place and since then he has been nothing but a constant source of inspiration and enthusiasm for me. I learned from him ability to be patient during tough times. I want to thank him for believing in me and constantly
stimulating the spark in me to help me accomplish my goals.
The second person without whom, I think, I would have never been able to fulfill my
dreams is my research advisor, Dr. Joseph A. Caruso. He is a greatest advisor, not only because
of his science, but also he is one of the greatest people I have ever met. I acknowledge him for
accepting me in his group and giving me opportunities to work on various projects that have
become a part of my dissertation. He has always appreciated my contributions and his efforts
have been very supportive and kept me challenged enough to get best out of the graduate school
experience. Under his guidance, I learned to become an independent researcher and also developed a competitive portfolio, which will help me to become professional chemist. Not only that, he has a great impact on my attitude towards life. One of the most important things he has
taught me is to believe in myself and to believe in best of others. He has always tried to instill great deal of confidence in me, which has helped me to make right decisions for my future. I will always try to emulate his working style and consider him as my role model for rest of my life.
I would like to thank Dr. Patrick Limbach and Dr. Theresa Reineke for serving on my
committee. I want to thank them for their insightful suggestions during presentations I have
given. Not only that, I am fascinated by their accomplishments and achievements in their career
and am constantly motivated by these. Another faculty member I would like to acknowledge is
Dr. Thomas Ridgway for giving me opportunities to teach various day and evening classes.
These assignments have helped me to improve my teaching as well as communication skills. I
would also like to pay many thanks to many colleagues past and present. I want to thank Dr. Sasi
Kannamkumarath, Dr. Rodolfo Wuilloud, Dr. Sandra Mounicou, Dr. Juris Meija for many
discussions and giving me valuable advice on various research projects I have worked on and on
which we have collaborated. I would like to thank Douglas Richardson, Kevin Kubachaka, Scott
Afton, Jenny Ellis, Allison Krentz, Santha Ketaverapu and Heather Ternary for their constant
support and also keeping very friendly environment in the lab which I think is one of the most
important things to remain productive in any field. I would also like to thank University of
Cincinnati - Department of Chemistry and the National Institute of Environmental Health
Sciences for financial support of my work. Agilent Technologies is to be recognized for their
continuing support with the instrumentation central to my dissertation studies.
I cannot end this without paying my gratitude to my parents, Ashok and Abha Shah, my
brothers Avanish and Ankit Shah and my grandfather, the Late Shri Madhusudan Das Shah.
Without them and their support, I would have not been able to come so far in my life and also in
my career.
--- Monika Shah
TABLE OF CONTENTS
CHAPTER 1: INTRODUCTION…………………………………………………………4
1.1. Heteroatom containing molecules 1.2. Characterization of heteroatom containing molecules 1.2.1. ICP-MS for heteroatom determination 1.2.2. Coupling of chromatography and ICP-MS 1.2.2.1. LC-ICP-MS 1.2.2.2. GC-ICP-MS 1.2.3. Molecular mass spectrometry in speciation analysis 1.3 References
CHAPTER 2: IDENTIFICATION AND CHARACTERIZATION OF SELENIUM SPECIES IN ENRICHED GREEN ONIONS USING HPLC-ICP-MS AND ESI-ITMS …………………………………………………………………..25
2.1. Abstract 2.2. Introduction 2.3. Experimental 2.3.1. Instrumentation 2.3.2. Reagents and Standards 2.3.3. Cultivation and preparation of selenium enriched green onions 2.3.4. Total Selenium determination 2.3.5. Chromatographic speciation analysis 2.3.6. ESI-ITMS analysis 2.4. Results and Discussions 2.4.1. Plant growth and total selenium accumulation 2.4.2. Molecular weight distribution of selenium in green onions 2.4.3. Speciation studies of selenium by HPLC-ICP-MS 2.4.4. Investigation of unknown selenium species by ESI-ITMS 2.5. Conclusions 2.6. References
CHAPTER 3: CHARACTERIZATION OF IODINE CONTAINING SPECIES IN COMMERCIALLY AVAILABLE SEAWEEDS………………………46
3.1. Abstract 3.2. Introduction 3.3. Experimental 3.3.1. Instrumentation 3.3.2. Reagents and Standards 3.3.3. Procedures 3.3.3.1. Sample collection and preparation 3.3.3.2. Total iodine determination
1
3.3.2.3. Extraction of iodine from seaweed in various media 3.3.2.4 Extraction of high M.W. iodine species from Wakame 3.3.3.5. Extraction of polyphenol-bound iodine from Wakame 3.3.4 Chromatographic Conditions 3.4. Results and Discussion 3.4.1. Total iodine concentration in samples and extracts 3.4.2. Optimization of SEC-ICP-MS 3.4.3. Fractionation studies of iodine in seaweed extracts 3.4.3.1. SEC-ICP-MS of Wakame extracts 3.4.3.2. Study of the association of iodine to biological molecules 3.4.3.3. SEC-ICP-MS of Kombu extracts 3.4.4. IC-ICP-MS for speciation of inorganic iodine 3.4.5. RP-HPLC-ICP-MS for studying iodine species 3.5. Conclusions 3.6. References
CHAPTER 4: DETERMINATION OF PHOSPHORIC ACID TRIESTERS IN HUMAN PLASMA USING SPME AND GC-ICP-MS…………………………………71
4.1. Abstract 4.2. Introduction 4.3. Experimental 4.3.1. Chemicals and Standards 4.3.2. Instrumentation 4.3.3. SPME fibers 4.3.4. Sample Preparation 4.4. Results and Discussion 4.4.1. Injection and gas chromatography conditions 4.4.2. ICP-MS conditions i. Forward Power and argon carrier gas flow rate ii. Optimization of ICP collision cell gas flow rate 4.4.3. Solid Phase micro-extraction i. Selection of SPME coating 4.4.4. Optimization of extraction procedure i. Influence of plasma matrix and deproteinization ii. Effect of Salt iii. Effect of pH iv. Effect of extraction temperature v. Effect of extraction and desorption time 4.5. Analytical performance Characteristics 4.6 Application to human plasma samples 4.7. Conclusions 4.8. References
2
CHAPTER 5: EXPLOING THE MINOR SELENIUM VOLATILES FROM ENRICHED GREEN ONION………………………………………………………………93
5.1. Abstract 5.2. Introduction 5.3. Experimental 5.3.1. Reagents and standards 5.3.2. Instrumentation 5.3.3. Preparation of selenium enriched green onions 5.3.4. Sample preparation 5.3.5. Synthesis of matching standards 5.4. Results and Discussion 5.4.1. Mass Defect Analysis (developed by Dr. Juris Meija) 5.4.2. Search for trace level Se volatiles i. GC/TOF-MS analysis
5.5. Conclusions 5.6. References
CHAPTER 6: LOCALIZATION AND CHARACTERIZATION OF SELENIUM CONTAINING HIGH MOLECULAR WEIGHT SPECIES IN MUNG BEAN (VIGNA RADIATA) SPROUTS……………………………………119
6.1. Abstract 6.2. Introduction 6.3. Experimental 6.3.1. Instrumentation 6.3.2. Reagents and Standards 6.3.3. Preparation of selenium-enriched mung bean seedlings 6.3.4. Total Selenium determination 6.3.5. Fractionation Procedure 6.3.6. Chromatographic speciation analysis 6.3.7. Gel electrophoresis on protein fraction 6.3.8. In-gel tryptic digestion 6.3.9. MALDI-MS analysis 6.4. Results and Discussions 6.4.1. Total selenium concentration 6.4.2. Distribution of selenium in sprouts 6.4.3. Distribution of selenium in various sub-cellular components 6.4.4. Gel electrophoresis on SEC fraction 6.4.5. MALDI-MS analysis 6.5. Conclusions 6.6. References
CHAPTER 7: CONCLUSIONS……………………………………………………………..139
3
CHAPTER 1 | INTRODUCTION
4
1.1 HETEROELEMENT CONTAINING MOLECULES
It is widely recognized that several important molecules of biological interest include trace elements playing vital roles in their functional properties. For example, actions of many proteins in living organisms depend critically upon their interaction with transition elements such as Fe,
Cu, Zn etc.1 Likewise, human nutrition, biochemical and physiological processes in plants and animals, clinical medicine and environmental chemistry are some of the important areas in which various physiochemical forms of elements (metals and non-metals) have been identified as major participants exhibiting toxicological or beneficial effects.2 These species either can arise as a metabolism product (endogenous or bio-induced molecules) in plants and animals1 or can be of anthropogenic origin (pesticides, herbicide, flame retardants etc) and be present in the samples of biological interest where their high concentrations might cause health concerns. The molecular nature of the species around the heteroelement (elements apart from C, H, O and N) promotes its behavior in terms of its mobility, bioavailability, retention and specific biological functions. So characterization of these various forms of element is very important in order to assess their health and environmental impacts. The increasing recognition of possible roles (benefit/toxic) of these species in living organism have stimulated the development of analytical techniques for their qualitative characterization and, if possible, quantitative determinations in various matrices.
Analyses that lead to characterization of heteroelement containing molecules in biological and environmental matrices is termed as “elemental speciation”.3
Elemental speciation has been rapidly growing over last few years. The studies vary from understanding the distribution of particular element in a sample to complete characterization of various elemental forms, including their quantification depending on the need of the analysis.
5
The determination may involve elucidation of oxidation states, isotopic composition and/or
complex or molecular structure of species of interest.
The scope of speciation analyses ranges from studying the fate of exogenous environmental contaminants and their degradation products to compartmentalization and characterization of endogenous heteroelement containing species, such as natural metabolites or metal complexes with bioligands. Both volatile and non-volatile components have received equal focus in speciation of heteroelement containing species in various matrices. In most of the cases, the species of interest constitute very small portions of the total sample. Therefore, their determination requires development of the analytical techniques with remarkable specificity and sensitivity. Sample preparation, analyte extraction, pre-concentration or enrichment also play important roles in trace element speciation and sometimes they are the key factors, since often speciation analyses are in sub-ppb range.4
The aim of this dissertation is characterization of selenium, iodine containing molecules in
various food samples and those of other complex matrices. On the basis of the results,
conclusions are drawn about their chemical association with biological molecules, both micro
(volatile and non-volatile forms) and macro molecules and their metabolites in the studied
system. Another aspect of this dissertation is method development for the fast and sensitive
determination of anthropogenic molecules such as phosphorus containing fire retardants so that
their presence can be confirmed in the biological matrices.
1.2. CHARACTERIZATION OF HETEROATOM CONTAINING MOLECULES
As stated above, trace levels of the target species, complex matrix composition and also their
lability (weak metal ligand interaction or instability of certain forms at particular conditions)
6
demands an analytical method with high specificity and selectivity for characterization of heteroatom containing molecules in complex matrices.5 The conventional species-specific detection method such as ion-selctive electrodes, nuclear magnetic resonance, flourimetry etc. lack these analytical characteristics.3 Instrumental development in mass spectrometric methods over last few decades has helped to overcome this disadvantage. Heteroatom determination via atomic spectrometric techniques is usually employed to incorporate the high element specificity in the procedure. For this reason, inductively coupled plasma mass spectrometry has become the detector of choice as it is widely known for its exceptional sensitivity, high throughput analysis, multielemental detection capability and detector response irrespective of the species’ chemical form. Multi-element detection capability and the high dynamic range of ICP-MS can be exploited together in a single run for simultaneous detection of various elements present in a sample with a wide concentration difference.2 However one limitation of ICP-MS is that it cannot be used as a stand alone technique in species specific characterization owing to the loss of structural information during ionization process. Hence, a separation technique is usually incorporated for resolution of various physiochemical forms of species into individual components prior to their element specific detection via ICP-MS – this requiring a retention time standard.
Hyphenation of separation techniques with ICP-MS has become the most common instrumental approach for characterization of heteroatom containing molecules. There could involve various possibilities for the on-line coupling of a separation technique with an element specific detector, depending primarily on the physiochemical properties of target species. High performance liquid chromatography, electrophoresis and gas chromatography have been successfully hyphenated to ICP-MS in numerous speciation studies. The easy coupling and
7
availability of commercial interface has attributed to the expansion of hyphenated techniques in
the analysis of heteroatom containing molecules at trace levels in complex matrices.6
Another dimension in characterization of heteroatom containing molecules is introduced
by incorporation of molecular mass spectrometry in speciation analysis. The shifting focus from
environmental chemistry where anthropogenic pollutants and their degradation products are
analyzed to the characterization of endogenous natural metabolites have necessitated the
intervention of techniques such as electrospray (ES-MS), Matrix assisted lased desorption
ionization (MALDI) and electron-impact time of flight mass spectrometry (EI-TOF-MS), which
are able to provide structural information of the unknown target species. Characterization of
these molecules remains incomplete by ICP-MS as majority of these species have not been
isolated to sufficiently pure state to be used as retention time standards.6 In such cases, molecular mass spectrometry is usually employed in parallel to determine the identity of eluted species.
The synergy of these two methods of detection has been successfully applied in many recent speciation studies where information obtained by one completes and also compliments the information by other.
1.2.1 ICP-MS FOR HETEROATOM DETERMINATION
Element specific determination by ICP-MS is based on the ionization of elements in the
argon plasma produced at the temperature in the range of 5000-9000K. The energy produced is
sufficient to ionize most of the elements in the periodic table except those restricted by Ar
ionization potential (He, Ne, F) or atmospheric gases (O, H, C). The argon plasma is formed in a
quartz torch by coupling radiofrequency energy at 27.1 or 40 MHz through a copper load coil to form an oscillating magnetic field. The front end of torch is connected to the sample introduction
8
device, which consists of the nebulizer and the spray chamber. The nebulizer breaks the liquid
sample into the aerosol at its tip by the action of the nebulizer gas. The spray chamber eliminates the larger droplets by condensation while fine aerosol drops are carried to the plasma
by the nebulizer gas flowing from the tip to the central tube (injector tube) of the quartz torch.
Aerosol containing sample is then desolvated, vaporized, atomized, and finally ionized. In case of solid and gaseous samples, the interface is appropriately modified for maximum transport of the analytes. Once ions are formed in the plasma, they enter the mass spectrometer portion of the
ICP-MS through the sampler and skimmer cones usually made of nickel and occasionally platinum. Each cone has a small aperture through which the ions pass. After passing through the cones, the ions are focused by a set of lenses called the extraction lens and omega lens (Agilent
7500) held at different potentials, the optimum value determined by maximizing the S/N at the selected mass to charge ratio. These lenses are located before the gate valve separating the atmospheric pressure region of the interface from the high vacuum region of the mass spectrometer. After the omega lenses, ions enter the octopole reaction cell (ORC), which is on- axis with the mass analyzer and detector, thereby enhancing the high ion transmission and hence, increasing the sensitivity enabling the trace and ultra-trace analysis.7 The focused ions enter the
quadrupole mass analyzer, which acts as a filter allowing only ions of the one mass to charge
ratio to be transmitted at a time depending on the applied voltages (ac and dc) to the quadrupole
rods, while other ions are electrostatically directed from their original path. Quadrupole mass
analyzers have unit mass resolution, which means that polyatomic interferences can prevent the
determination of some elements in some matrices, for e.g. 75As in sea water (interference from
40Ar35Cl+). However, the overall performance is remarkably good, especially following chromatography where the interfering ions are often separated from the analyte species. 6
9
Quadrupole based mass analyzers are the by far the most commonly used analyzers for
ICP-MS because of their low cost, relative simplicity, and rapid scanning capabilities.3 However, when the poor mass resolution and the need to determine simultaneous multielemental analysis become critical, other types of mass spectrometers that are preferred for use, in conjunction with the ICP ion source, are time of flight and sector field mass spectrometry depending upon the need of the analysis and availability of the instrument.
Octopole reaction cell: Elemental analysis by ICP-MS may be hindered by spectroscopic
interferences. These type of interferences arise when another species having the same mass to charge ratio is present for target species. Formation of polyatomic species from sample matrix
(O, N, C, H, Cl, S, F) and plasma (Ar) components, isobaric interferences, doubly charged ions are some of the cause of spectroscopic interferences. For example, determination of P and Se is made difficult by the formation of polyatomic ions such as 15N16O+, 14N17O+ and 14N16O1H+
40 + at31amu, Ar2 at mass 80amu for P and Se, respectively. The presence of these polyatomic
species is troublesome when interferences exist at the m/z for the element of interest, particularly
if it is monoisotopic or at the most abundant isotope to be monitored.8 These polyatomic species
are unavoidable and are reflected in higher background counts leading to higher limits of
detection. The most widely used quadrupole mass analyzer does not have sufficient resolution to
separate the target isotope from interfering polyatomic ions having similar nominal mass to
charge ratio. To resolve this problem, high resolution double focusing sector field mass analyzers
were adapted for ICP, which are capable of reducing interferences through the high resolution
capabilities. However, in this process sensitivity is compromised for high resolution.9 Another development toward the removal of polyatomic interferences has been introduction of the collision/reaction cell, which is placed before quadrupole mass analyzer.10 The advent of
10
collision cell technology has helped to increase signal to noise ratio for the isotopes impossible
to be monitored by the low resolution quadrupole mass analyzer without significant loss in
sensitivity. These devices have been used in many recent applications to obtain relatively
interference free responses and eventually lower detection limits for trace analysis of elements in
biological matrices. Detection limits are in the range of high ng l-1 for phosphorus containing
pesticides using ICP-MS equipped with octopole collision cell 11, 12.
The collision/reaction cell is operated by pressurizing the cell interior with gases such as H2 and He in reaction and collision mode, respectively. In the case of the hydrogen mode, reaction of hydrogen to the argon based polyatomic ions result in removal of these ions by either charge transfer, proton transfer or atom transfer as described below.
Charge transfer: + + Ar + H2 H2 + Ar (Interference on 40Ca+)
Proton transfer: + + Ar2 + H2 Ar2H + H (Interferences on 78Se+ and 80Se+)
Atom transfer: + + ArO + H2 H2O + Ar (Interference on 56Fe+)
While in case of the helium mode, removal of polyatomic interferences takes place by one of two
mechanisms: a) Collision induced dissociation b) Kinetic energy discrimination.7 Dissociation of polyatomic ions into individual components takes place as a result of their collision with He.
This phenomenon is limited by the fact that energy associated with collision is greater than the bond energy of the polyatomic species. A few common polyatomic interferences have low enough bond energies for this to occur.7 They include, ArC+, which can interfere with the
11
measurement of 52Cr. However in majority of the cases, the second mechanism is followed.
Kinetic energy discrimination is based on the fact that polyatomic ions undergo more energy
reducing collisions with the He gas than analyte ions owing to their larger cross-section.
Stopping potential is then applied between octopole and quadrupole to reject the lower energy
interferent ions, while allowing the higher energy analyte ions to enter the quadrupole mass
analyzer.11 Polyatomic interferences such as 15N16O+, 14N17O+ on m/z 31 are removed by this
method.
Thus, an inductively coupled plasma mass spectrometer equipped with the octopole
reaction cell represents a most powerful element specific detection method known for high
sensitivity, relatively interference free response and the capability to perform multielemental
analysis with compound unspecific response (Figure 1.1). It is capable for detecting not just
metals but also metalloid and certain non-metals such as phosphorus and sulfur with the detection limits in the ng ml-1 to pg ml-1 range.
Figure 1.1: A comparative figure showing three advantages of ICP-MS such as high sensitivity A) Overlay of HPLC-SF-ICP-MS (195Pt) and HPLC-UV signal of mixture containing 5’ GMP and cisplatin - 195 incubated for 20h. showing detection of mono (cis-[Pt(NH3)2Cl(GMP)] ) in Pt chromatogram, element selectivity B) Overlay of HPLC-ICP-MS (31P) and HPLC-UV traces of tryptic digest of β-caesin showing relativity simple ICP-MS chromatogram, and compound unspecific response C) Overlay of HPLC-ICP-
12
MS and HPLC-ESI-MS traces of phospholipid mixture PG, PE and PC respectively. (Figure adapted from Hann et al.13 and Axelsson et al.14 )
1.2.2 COUPLING OF CHROMATOGRAPHY AND ICP-MS
1.2.2.1 LC-ICP-MS
Liquid chromatography has long been traditional mainstay for separation in speciation analysis.
It is a powerful technique which allows separation of complex mixtures into individual
components. Major advantages of HPLC over other commonly used separation method such as
GC is that it requires minimum sample preparation, operates at ambient temperature and not
limited by thermal stability of the analytes. These important advantages have made it a preferred
technique of choice for the separation of macromolecules and ionic species, labile natural
products and also thermolabile environmental contaminants. Adequate resolution,
reproducibility, sensitivity and simplicity of the interface between HPLC with ICP-MS make it
the most important technique for speciation studies.
Coupling between the two can be achieved by connecting the outlet of the column to the
liquid sample inlet (nebulizer) using an inert or polymeric tubing.15 The efficient transfer of the
analyte from the solution phase to the gaseous phase is determined by the proper nebulization of
the chromatographic eluent. Applications involving regular bore columns, most commonly
employ concentric or cross-flow nebulizers (optimum flow rate 0.5-2ml/min) while for capillary or magabore HPLC systems micronebulizers (DIN, DIHEN, micromist, PFA etc.) are employed.9 Although coupling is simple and straightforward, separation conditions are limited
by the fact that excessive salt or organic solvent introduced into the plasma can affect the
sensitivity over the experimental period (salt or carbon deposit on the cones), plasma stability
and also eventually cause it to extinguish. Lower salt concentrations with volatile components
13
are chosen to increase the compatibility of chromatographic eluent with the ICP-MS. In order to
circumvent the cooling effect caused by high organic solvent content introduced into the plasma,
high forward power up to 1500W is used.16 Membrane desolvators, addition of oxygen to the nebulizer gas flow or to the sample aerosol leading to oxidation of organic solvents and sometimes lowering the temperature of spray chamber below 0oC is also implemented to
minimize the problems associated with high organic load in plasma.17 Platinum cones are used when oxygen is introduced into the plasma. Post column dilution of the chromatographic eluent is also practiced to reduce the organic solvent into the plasma.18,19
Based on the nature of the target molecules and demand of the speciation study one or
more of several high performance liquid chromatographic methods such as reverse phase (RP),
reversed phase-ion pairing (RP-IP), ion exchange (IC) and size exclusion (SEC) can be
employed for the isolation and separation of organometallic compounds. One of the major
considerations in development of any analytical method for speciation analysis is conservation of
native forms of the element during procedure. Natural distribution of the species should not be
disturbed either due to chemical interaction of the analyte with the stationary phase or due to
extreme pH/ salt concentration of the buffer. Since the packing material in a reversed phase
column is composed of silicates functionalized with C-2, C-8 or C-18 hydrocarbons, separation
between the analytes is solely based on hydrophobicity of molecules, which makes it a most
suitable type of chromatographic method for speciation analysis as mildest mobile phase
conditions are used to achieve effective separation. RP-IP-HPLC is the variation of RP-HPLC in
which ions in the solution are paired or neutralized using small concentration of appropriate ion-
pairing reagents for separation on the reversed phase column. The alkyl chain of ion-pairing
reagents imparts hydrophobicity to the ionic and highly polar species for their retention and
14
hence, separation on RP columns. Perfluorinated carboxylic acids20 and alkyl sulfonic acids21, 22 are most commonly used ion-pairing reagents in speciation studies. RP-IP-HPLC has been extensively employed for wide variety of compounds with molecular weight less than 3kDa.23
However, IC-HPLC and SEC-HPLC are also widely used.
Ion chromatography is based on exchange of analytes with the similarly charged mobile phase ions already present on oppositely charged functional group sites available on the stationary phase of the column. The difference in the interaction of analytes with the charged stationary phase brings about their separation. Selectivity in separation can be achieved by gradient elution optimizing the ionic strength, changing the pH or by adding the charged additive to the mobile phase. IC-HPLC has been extensively used for the speciation of arsenic species.
Not only separation but prior purification of extracts can also be performed using ion exchange based phases for removal of matrix components, which helps to increase the signal to noise ratio in MS spectra and eventually, identification of species of interest. Bendahl et al.24 purified the
fraction containing selenomethylselenoglactosamine using strong anion exchange cartridge
before its confirmation through CE-nESI-MS in basal human urine. Mcsheehy et al.25 employed
preparative anion-exchange followed by semi-preparative cation exchange or anion exchange
HPLC to isolate, purify and isolate organoarsenic species in marine certified reference material.
The role of size exclusion chromatography in speciation studies is to provide preliminary
information about the distribution of elements in a sample. It gives the size estimate of
metallospecies by separating them on the basis of their molecular weight. Macromolecules, such
as proteins and polysaccharides elute at the dead volume, while smaller molecules appear later
due to longer retention in the SEC column. However, due to its limited peak capacity complete
separation of compounds in complex biological matrices cannot be obtained by size exclusion
15
chromatography. In such cases application of other kinds of HPLC techniques such as RP-IP,
cation exchange or anion exchange is usually followed on SEC fractions when greater resolution
is desired for species identification. Because of high tolerance to the biological matrix SEC is
also applied as the sample purification step prior to other separation techniques.24
Multidimensional approaches are becoming increasingly popular in speciation studies, especially in cases when peak purity is critical for characterizing particular species by molecular mass spectrometry such as ESI-MS and ESI-MS-MS.
1.2.2.2 GC-ICP-MS
Gas chromatography is another very popular separation technique used in conjunction with ICP-
MS for characterization of heteroatom containing species that are either volatile or can be converted to the volatile form by derivatisation. High resolving power of the GC and the high sensitivity of the ICP-MS have made the combination a very suitable technique for the analysis of target species in environmental, industrial and biological samples.26 Since samples are
introduced to the plasma in gaseous form, atomization and ionization of sample is more complete
as energy from the plasma is not required for the desolvation and volatilization process. The
mobile phase in GC is usually helium gas, which not only enables the quantitative transport of
the sample (nearly 100%) to the detector without nebulisation, but also accounts for a low
background contribution in the detector itself.27 This overall process contributes to lowering the
detection limits of the method.
The interface between GC and ICP-MS is designed such that condensation of analytes is
prohibited during their transport from the GC column to the inlet of ICP-MS torch. For this purpose, usually the heated transfer line is employed which prevents the loss of analytes due to
16
condensation and eventually helps to maintain the peak shape for high sensitivity and low limits
of the detection. The entire transfer line is uniformly heated to avoid any cold regions.
Depending on the volatility of the analyte species, the
Figure 1.2: Design of heated GC-ICP-MS interface, Adapted from http://www.analitica.cl/pdf/F.7500a,- ce;.pdf
degree of heating is varied. In this case, the conventional spray chamber is removed and the
transfer line inserted into the central channel of the torch. A carrier or make-up gas flow is also
introduced to the GC effluent to attain a sufficient flow to get the analytes into the central
channel of the plasma. This type of interface is available commercially (Figure 1.2). In the second type of design, GC effluent is mixed with an aqueous aerosol in the spray chamber prior to introduction into the plasma. However the sensitivity of the system is reduced owing to the loss of plasma energy required to desolvate the wet aerosol. The design also fails with complex matrices, which may condense in the spray chamber. 28
17
There are several heteroatom containing organic molecules which are volatile enough to
be separated by GC in their native state. They include species (MenEt4-nPb) (n=1 to 4),
methylselenium compounds (e.g. Me2Se, Me2Se2 etc), some organomercury compounds
+ (MeHg , Me2Hg), some non-metal containing pesticides, herbicides, flame retardants as well as
naturally occurring metalloporpyrins. However, there are also a multitude of organometallic
compounds which exist in quasi-ionic polar forms that have relatively high boiling points and
often, poor thermal stability. For such compounds various kinds of derivatization procedures
have been applied prior to their separation with the GC system. Mainly three derivatization
techniques have been employed for the elemental speciation a) hydride generation, (b) extraction
into an organic solvent and derivatization with an alkylating agent, and (c) in situ ethylation with sodium tetraethylborate (NaBEt4) followed by head-space analysis. Regarding the development of derivatization procedures, the implementation of sodium tetra(n-propyl)borate (NaBPr4) reagent represents the most recent advance in this area.28
For the separation of volatile organometallic compounds, capillary GC column is usually
preferred over the packed GC column because of its higher resolving powers and better
efficiency. However, the sample introduction in this case becomes challenging because of the
reduced sample volume and high dilution factor caused by the make-up necessary to match the
spectrometer’s optimum flow rate, resulting in loss of sensitivity.27 Therefore, the extraction and
pre-concentration of the analytes of interest from the complex matrix components becomes very
important prior to injection on the GC column. Different sampling techniques such as
microwave-assisted extraction, headspace solid-phase microextraction (SPME), stir bar sorptive
extraction (SBSE), and purge and capillary trapping have been applied in various studies for
analyte extraction and preconcentration.
18
SPME: Solid phase micro-extraction is a very simple solvent free pre-concentration technique
that has been successfully applied in extraction of many organometallic species in recent years. It
has a numerous benefits such as simplicity, reduced sample volume for analysis, low cost and the
compatibility with on-line analytical procedures. Also, the enrichment factors obtained are very
high, enabling the trace and ultra-trace analysis of target species present in biological and environmental samples. The technique is based on establishment of the equilibrium between the analyte and fused fiber coated with the stationary phase, which can be a liquid polymer, solid sorbent or combination of both.29 The fiber may be immersed into the sample or may be exposed
in the headspace above the sample for the extraction process depending upon the volatility of the
target species. Several factors affect extraction efficiency of SPME and are evaluated during
method development. Optimization of parameters affecting SPME extraction is performed in
order to increase the volatilization process and hence increase the analyte recovery and decrease
the sampling time. Once the equilibrium is achieved, the fiber is transferred to a GC injector by
means of a microsyringe, and the analytes are thermally desorbed inside the heated injector onto
a capillary chromatographic column. For certain heavier analytes the time required for the
analytes to reach the equilibrium condition is long, in such cases extraction at non-equilibrium
condition is performed in order to reduce the sampling and analysis time provided the
repeatability for the analyte is demonstrated. SPME is very sensitive to the experimental
conditions so any factor that affect the absorption coefficient and absorption rate affect the
amount of analyte absorbed on the fiber and in turn reproducibility.30 The parameters that affect the SPME process are fiber type, ionic strength, pH, temperature, agitation and extraction time.
SPME optimization can be achieved in traditional univariate trial where each factor is studied separately or by the multivariate approach where several factors are studied simultaneously.31
19
There are only a few studies in the literature where the latter approach is applied for the SPME process. SPME is a non-exhaustive technique and the recovery of the analytes from the system is comparatively small so its use in conjunction with a sensitive detection system like GC-ICP-
MS is quite beneficial.27
1.2.3 MOLECULAR MASS SPECTROMETRY IN SPECIATION
ANALYSIS
In recent years, most of the speciation studies are aimed at understanding the role of various elements (both essential and toxic) in living organisms. Identification of various metals containing biomolecules present either as intrinsic metalloenzymes or formed under the stress conditions as a result of metabolism/detoxification process demands the involvement of techniques which can provide molecular and structural information for successful characterization. Role of ICP-MS in these analyses becomes more of a n elemental screening technique since structural information is lost during the plasma ionization process. Introduction of a separation step prior to detection and using retention time matching only facilitate the successful identification when standards are available. However, confidence in the detection process diminishes when dealing with the complex matrices where probability of co-elution of several species in one peak cannot be ignored. In such cases, molecular mass spectrometry comes into play.
Molecular mass spectrometry, such ESI-MS and MALDI-MS, has been recognized as an
essential tool in the biochemical speciation, usually applied for identification and confirmation of
unknown peaks seen in LC-ICP-MS analysis. Molecular mass and structural information
obtained by molecular mass spectrometry not only completes, but also compliments the
20
information obtained by ICP-MS. Analysis of metal containing species using mass spectrometry
is facilitated by the fact that most of the elements have their characteristic isotopic pattern, which
gives them specific identity and distinguishes them from multitude of peaks arising from the
matrix components. In such cases, molecular ion information obtained by ESI or other molecular
mass spectrometric technique may prove to be sufficient enough for the unequivocal
identification of the metal containing species. However, some widely studied elements for
biochemical speciation such as arsenic and phosphorus are monoisotopic and no characteristic
isotopic abundance pattern can be obtained. In such cases, application of ESI-MS/MS and
structural fragmentation patterns from this (also known as CID, collision induced dissociation)
are considered necessary to avoid misidentifying metal containing compounds using only
chromatographic retention time and molecular mass information.32
Unlike ICP-MS, the ionization process in ESI-MS is highly dependent on the matrix
components and a significant amount of peak suppression can be possible due to the presence of high concentration of salts in mobile phase or due to matrix components. Higher detection limit
as compared to ICP-MS also demands for cleaner sample extracts for the analysis of trace
amounts of metal containing species that are easily detected by ICP-MS. Fraction collection and
pre-concentration of several fractions from LC-ICP-MS prior to the fraction entering the plasma
is usually employed to compensate for the lower sensitivity of ESI-MS as compared to ICP-
MS.33 Desalting is often practiced before the analysis as increasing concentration of salt can
suppress the ES ionization efficiency.34 In some cases ESI is replaced by MALDI as it is less
vulnerable to the matrix effects than ESI.35 The most common mass analyzers used in conjunction with these ionization techniques are time of flight and quadrupole time of flight for
21
tandem MS because of their high sensitivities and high mass accuracy, which helps to increase confidence in the structural characterization.
To conclude, application of ICP-MS and its combined use with molecular mass spectrometric techniques in conjunction with various separation techniques has become a promising analytical tool in characterization of important molecules of interest in biological, biochemical and environmental speciation. Easy on-line coupling with variety of separation methods such as HPLC, GC have increased the scope of ICP-MS in various applications. The unparalled element selectivity and high sensitivity (irrespective of chemical form of the element) obtained by intervention of ICP-MS in these fields, not only simplifies but also speeds up the overall analysis. On the other hand, its combined use with molecular mass spectrometry overcomes the drawbacks of existing methods and leads to results unachieved by conventional detection methods.
22
1.4 REFERENCE:
(1) Szpunar, J. Analyst 2005, 130, 442-465. (2) Shah, M.; Caruso, J. A. Journal Of Separation Science 2005, 28, 1969-1984. (3) Caruso, J. A.; Sutton, K.; Ackley, K. L. Elemental Speciation.New Approaches for Trace Element Analysis, Comprehensive Analytical Chemistry; Elsevier: Amsterdarm, 2000. (4) Meija, J. PhD Dissertation, University of Cincinnati, Cincinnati, 2005. (5) Kannamkumarath, S. S. PhD Dissertation, University of Cincinnati, Cincinnati, 2004. (6) Szupnar, J.; Lobinski, R. Hyphentated techniques in speciation analysis; Royal Soceity of Chemistry: Cambridge, 2003. (7) Wilbur, S.; Soffey, E.; Agilent Technologies, 2004. (8) McSheehy, S.; Mester, Z. Trac-Trends in Analytical Chemistry 2003, 22, 210-224. (9) Szpunar, J.; Lobinski, R.; Prange, A. Applied Spectroscopy 2003, 57, 102A-112A. (10) Tanner, S. D.; Baranov, V. I.; Bandura, D. R. Spectrochimica Acta Part B-Atomic Spectroscopy 2002, 57, 1361-1452. (11) Vonderheide, A. P.; Meija, J.; Montes-Bayon, M.; Caruso, J. A. Journal of Analytical Atomic Spectrometry 2003, 18, 1097-1102. (12) Profrock, D.; Leonhard, P.; Wilbur, S.; Prange, A. Journal of Analytical Atomic Spectrometry 2004, 19, 623-631. (13) Hann, S.; Zenker, A.; Galanski, M.; Bereuter, T. L.; Stingeder, G.; Keppler, B. K. Fresenius Journal of Analytical Chemistry 2001, 370, 581-586. (14) Axelsson, B. O.; Jornten-Karlsson, M.; Michelsen, P.; Abou-Shakra, F. Rapid Communications in Mass Spectrometry 2001, 15, 375-385. (15) Montes-Bayon, M.; DeNicola, K.; Caruso, J. A. Journal of Chromatography A 2003, 1000, 457-476. (16) Browner, R. F.; Canals, A.; Hernandis, V. Spectrochimica Acta Part B-Atomic Spectroscopy 1992, 47, 659-673. (17) Caruso, J. A.; Montes-Bayon, M. Ecotoxicology and Environmental Safety 2003, 56, 148- 163. (18) Shah, M.; Wuilloud, R. G.; Kannamkumaratha, S. S.; Caruso, J. A. Journal of Analytical Atomic Spectrometry 2005, 20, 176-182. (19) Kannamkumarath, S. S.; Wuilloud, R. G.; Stalcup, A.; Caruso, J. A.; Patel, H.; Sakr, A. Journal of Analytical Atomic Spectrometry 2004, 19, 107-113. (20) Kotrebai, M.; Tyson, J. F.; Block, E.; Uden, P. C. Journal of Chromatography A 2000, 866, 51-63. (21) Wrobel, K.; Parker, B.; Kannamkumarath, S. S.; Caruso, J. A. Talanta 2002, 58, 899-907. (22) Wrobel, K.; Caruso, J. A. Journal of Analytical Atomic Spectrometry 2002, 17, 1048- 1054. (23) Szpunar, J. Analyst 2000, 125, 963-988. (24) Bendahl, L.; Gammelgaard, B. Journal of Analytical Atomic Spectrometry 2004, 19, 950- 957. (25) McSheehy, S.; Mester, Z. Journal of Analytical Atomic Spectrometry 2004, 19, 373-380. (26) Wuilloud, J. C. A.; Wuilloud, R. G.; Vonderheide, A. P.; Caruso, J. A. Spectrochimica Acta Part B-Atomic Spectroscopy 2004, 59, 755-792. (27) Szpunar, J.; Lobinski, R. Hyphentated techniques in speciation analysis; Royal Soceity of Chemistry: Cambridge, 2003.
23
(28) Wuilloud, J. C. A.; Wuilloud, R. G.; Vonderheide, A. P.; Caruso, J. A. Spectrochimica Acta Part B-Atomic Spectroscopy 2004, 59, 755-792. (29) Dietze, C.; Sanz, J.; Camara, C. Journal of Chromatography A 2006, 1103, 183-192. (30) Sides, A.; Robards, K.; Helliwell, S. Trac-Trends in Analytical Chemistry 2000, 19, 322- 330. (31) Polo, M.; Llompart, M.; Garcia-Jares, C.; Cela, R. Journal of Chromatography A 2005, 1072, 63-72. (32) McSheehy, S.; Pohl, P.; Velez, D.; Szpunar, J. Analytical and Bioanalytical Chemistry 2002, 372, 457-466. (33) Shah, M.; Kannamkumarath, S. S.; Wuilloud, J. C. A.; Wuilloud, R. G.; Caruso, J. A. Journal of Analytical Atomic Spectrometry 2004, 19, 381-386. (34) Polec, K.; Garcia-Arribas, O.; Perez-Calvo, M.; Szpunar, J.; Ribas-Ozonas, B.; Lobinski, R. Journal of Analytical Atomic Spectrometry 2000, 15, 1363-1368. (35) Encinar, J. R.; Ruzik, R.; Buchmann, W.; Tortajada, J.; Lobinski, R.; Szpunar, J. Analyst 2003, 128, 220-224.
24
CHAPTER 2 | IDENTIFICATION AND CHARACTERIZATION OF SELENIUM SPECIES IN ENRICHED GREEN ONIONS BY HPLC- ICP-MS AND ESI-ITMS
25
2.1 ABSTRACT
In this work speciation of selenium in selenium-enriched green onions (Allium fistulosum) was done with high performance liquid chromatography (HPLC) and size-exclusion chromatography (SEC) coupled on-line to ICP-MS for selenium specific detection. Green onions were grown to maturity in a greenhouse using commercially available media. The extract obtained using sodium hydroxide (0.1 mol l-1) analyzed by SEC (CAPS 10 mmol l-1, pH 10.0) with ICP-MS detection showed the incorporation of selenium in both high as well as low molecular weight fractions. Protein bound selenoamino acids were released using enzymes (Proteinase K & Protease
XIV) and selenoamino acids found in cytosol in free form were extracted using 0.1M HCl. The extracts were analyzed for speciation studies by reversed phase ion pairing HPLC [0.1% (v/v) heptaflourobutyric acid, 5% (v/v) methanol, pH 2.5]. The identification of the selenium compounds was performed by matching the chromatographic retention times with commercially available standards. Further characterization of unknown selenium species was done by electrospray ionization-ion trap mass spectrometry (ESI-ITMS).
26 2.2 INTRODUCTION
Selenium is a key trace element required in small amounts in humans and animals. The key
nutritional role of selenium is its association with enzymes like glutathione peroxidase and
thioredoxin reductase involved in antioxidant activity and iodothyronine deiodinases involved in
thyroid metabolism.1-3 Anti-carcinogenic properties of selenium have also been suggested by
various studies.4-7 The critical selenium containing metabolite responsible for anticancer activity
is methylselenol (CH3SeH), which is generated in body tissues by chemical transformation of
various selenoamino acids.8 Selenium is not yet known as an essential element for plants but they
can incorporate selenium in their tissues following the sulfur assimilatory pathway if grown in
soil enriched with different forms of inorganic and organic selenium. Different species of
saccharomycaes, brassica and allium families have been known to be accumulators of
selenium.9-11 Selenium enriched yeast and different species of allium have been proposed as
dietary supplements due to their several medicinal and especially anti-carcinogenic properties. It
was shown by Clark et al. that human dietary supplementation with selenium-enriched yeast
decreased cancer incidence and mortality rate by nearly 50% for certain tumorigenic cancers.12
The major species in enzymatic extracts of yeast and garlic have been found to be selenomethionine and γ-glutamyl-Se-methylselenocysteine, respectively.13, 14 The difference in
anticancer activity of selenomethionine and γ-glutamyl-Se-methylselenocysteine lies in the fact
that the later species serves as a carrier of Se-methylselenocysteine. Se-methylselenocysteine
(MSC) is metabolized in vivo by a β-lyase to methylselenol (key species in chemoprevention of
cancer), whereas selenomethionine must undergo a multi-step chemical transformation process
before converting into methylselenol.5 During this chemical transformation process
selenomethionine may be incorporated into proteins in place of methionine, which makes it less
available than MSC for cancer chemoprevention.
Speciation analytical methodologies have been developed to determine the distribution of
selenium among various forms of selenoamino acids. Size-exclusion chromatography (SEC) and
high performance liquid chromatography (HPLC) are used to study the distribution and
separation of different selenium compounds in conjunction with ICP-MS for element specific
detection.11,15 Additionally, reversed phase ion-pairing chromatography has become widely
accepted for separating selenium compounds obtained from plant extracts.9, 16 Tandem mass
spectrometry, a species-specific detection technique, is well established for better identification
and characterization of the species of interest. Electrospray ionization-mass spectrometry (ESI-
MS) is a popular technique applied in speciation studies of selenium compounds.17 It
complements and completes the results obtained by ICP-MS. Both on-line14 and off-line18
HPLC-ESI-MS have widely been used in the speciation studies of selenium compounds in various samples.
In the particular case of vegetables from the Allium family, research has been selectively focused on bulbs and the information available on the speciation of selenium in leaves of allium vegetables is scarce. In fact, the presence of selenium species in leaves of regular onions (Allium cepa) has been recently demonstrated.19 However, the consumption of regular onions leaves is
not common among people. On the other hand, the major part of green onions (Allium
fistulosum) corresponds to leaves and they are consumed in many parts of the world as a major
ingredient of many recipes. Therefore, the identification and determination of selenium species
in green onion is of great interest. However, speciation studies of selenium in green onions have
not been reported in the literature.
28
The aim of the present work has been the characterization of selenium species in green
onion (Allium fistulosum). The SEC-ICP-MS coupling was exploited to understand the
distribution of selenium in covalently bound high molecular weight species (mainly proteins) and low molecular weight (small peptides or free amino acids) species. Reversed phase ion-pair chromatography coupled to ICP-MS was used to separate the selenium compounds obtained after different extraction procedures. Identification of selenium compounds was performed by matching the retention time with chromatographic peaks of a standard mixture containing different selenium species. Electrospray ionization ion trap mass spectrometry (ESI-ITMS) was used for identification of unknown species. Molecular mass was identified in the MS mode and structural information was obtained from MS2 mass spectra.
2.3 EXPERIMENTAL
2.3.1 Instrumentation
High performance liquid chromatography (HPLC) used a Shimadzu (Shimadzu Scientific
Instrument Inc, Columbia, MD, USA) LC-6A pump with a 100 μl loop (Rheodyne 7725
injection valve, Rheodyne, Cotati, USA). Reversed-phase chromatography was performed using
a C8 Alltima (Alltech, Deerfield, IL, USA) column (150 mm x 4.6 mm, 5 μm particle size);
while for SEC separations a Superdex Peptide HR 10/30 (Pharmacia Biotech) was used.
An Elan 6000 inductively coupled plasma mass spectrometer (Perkin Elmer Sciex,
Canada) equipped with a cross-flow nebulizer and a Scott-type double-pass spray
chamber made from Ryton was used for the determination of total selenium as well as for
selenium-specific detection after chromatographic separation. The column effluent was
introduced on-line to ICP-MS. The instrumental operation conditions are given in Table 2.1.
29
Table 2.1. ICP-MS, ESI-ITMS and chromatographic instrumental parameters ICP-MS parameters
Forward power 1250 W
Plasma gas flow rate 15.0 l min-1
Auxiliary gas flow rate 0.87 l min-1
Carrier gas flow rate 0.975 l min-1
Dwell time 0.1 s per isotope
Isotopes monitored 77Se, 78Se, 82Se
ESI-ITMS parameters
Drying temperature 350°C
Nebulizer gas and pressure He; 40 psi
-1 Drying gas and flow rate N2; 9 l min
Capillary voltage 3500 V
SEC chromatography parameters
Column Superdex Peptide HR 10/30
Mobile phase 10 mmol⋅l-1 CAPS buffer, pH 10.0
Flow rate 0.6 ml⋅min-1
Injection volume 100 μl
RP-IP-HPLC parameters
Column Alltima C8 (150 mm x 4.6 mm, 5 μm)
Mobile phase 0.1% (v/v) heptafluorobutyric acid, 5% (v/v)
methanol, pH 2.5)
Flow rate 0.9 ml min-1
Injection volume 20 μl
The MS experiments were performed using an Agilent 1100 Series LC/MSD Ion Trap
Mass Spectrometer (Agilent Technologies, Tokyo, Japan). The mass spectrometer was equipped
30
with an electrospray interface as the ionization source, which was operated under the conditions
indicated in Table 2.1. Full-scan mass spectra (100-500 amu) were recorded every 39.227 ms in
the positive ion mode. For MS/MS (MS2) experiments, experimental conditions were: isolation
width: 4 m/z; fragmentation energy: 30-200%; fragmentation time: 40 ms. The protonated
molecular ion [M+H]+ (m/z=313) was chosen as the parent ion for isolation and fragmentation of
an unknown selenium species.
2.3.2 Reagents
Analytical reagent grade chemicals and HPLC-grade methanol (Fisher Scientific, Pittsburgh, PA,
USA) were used. Doubly deionised water (18.2 MΩ.cm, NanoPure treatment system (Barnstead,
Boston, MA, USA) was used throughout.
L-selenomethionine, selenocystine, sodium selenite and sodium selenate were purchased from Aldrich (Milwaukee, WI, USA), Se-methylselenocysteine was from Sigma (St Louis, MO,
USA). Stock solutions containing 1 mg ml-1 selenium compound were prepared in 10 mmol l-1
hydrochloric acid (Sigma) and were stored frozen. Working solutions were prepared daily by
appropriate dilution.
For the determination of total selenium working solutions were prepared daily by
appropriate dilution of 1 mg ml-1 Se(IV) standard solution. Nitric Acid (Suprapure) 68% from
Pharmaco and Hydrogen Peroxide 30% from Fisher Scientific were used for sample digestion.
The chromatographic mobile phases (Table 2.1) were prepared from Sigma reagents.
Solutions of the following Sigma reagents were used: Heptafluorobutyric acid (HFBA) and 3-
cyclohexylamino-1-propane –sulfonic acid (CAPS).
31
2.3.3 Cultivation and preparation of selenium-enriched green onion
The allium species selected for the speciation study in the present work was Allium fistulosum
(evergreen long white bunching onion) as it is one of the commonly consumed varieties of green
onion in the United States and in Asian countries. It has a single, long branched pseudo stem up
to 50 cm in length and toughened leaves. It does not develop a large bulb since it lacks any long
dormancy. The seeds were purchased from Burpee (Warminster, PA). A number of 3 to 5 seeds
were seeded in cell packs having four cell each of dimensions (5”x 5” x 4”) each. Each cell
contained Promix BX (Natorp landscape, Cincinnati, OH), which is mixture of sphagnum, peat moss, perlite, vermiculite, dolomite and calicitic lime Stone. Cells were watered daily and fertilized when needed. Light watering was done to avoid any run out from the cells. A broad- spectrum fungicide was also applied when needed. Plants were grown in a greenhouse with temperatures controlled between 16-21 oC. Before any selenium treatment, the cells were thinned
to 3 plants per cell. One of the cell packs was not enriched with Se and used as control. Two
months after seeding, the cells were treated with Na2SeO3 at the following concentrations: 1, 2, 3,
5 and 15 μg g-1 Se. To compensate for loss from run out, two weeks after the first treatment the cells were treated with the same concentration of selenium as before. Subsequently, plants were
grown to maturity. As the container size was restrictive, the plants and bulb were smaller than
field-grown plants. Harvesting was done when the growth stopped four months after the sowing
process began.
After harvesting, green onions were washed with water. The bulbs and leaves were
separated from the root. Since the bulb in evergreen long bunching onion is tiny, the leaves were
not separated from the bulb and were combined together for speciation studies. The cleaned
32
green onions were frozen at -20oC using liquid nitrogen. Then, plants were ground and freeze dried.
2.3.4 Total Selenium Determination
For determinations of total selenium by ICP-MS, microwave digestion was employed. The
mineralization program used for microwave digestion involved three steps with first and second
step of 5 minutes each and power applied was 100 and 600 W, respectively. Power was increased
to 1000W in the third step and was held at that value for another 10 minutes. Approximately 0.05
g of dried onion (leaves or bulbs) was digested by using 3 ml HNO3 (65%) and 1 ml H2O2
(30%). The end solutions were diluted with distilled water up to 10 ml. The reagent blank was digested in the same way. The total selenium concentration was determined by ICP-MS. The isotope 78Se, 77Se, 82Se was used for the measurements. The total selenium content of each
sample is summarized in Table 2.1.
2.3.5 Chromatographic speciation analysis
Before SEC separation, 0.05 g of freeze dried onion sample was weighed in a plastic tube and
then extraction of the selenium species was performed using 3 ml of sodium hydroxide solution
(0.1 mol l-1). The mixture was shaken on a Vortex system for 15 min and centrifuged at 5000
rpm for 10 min. The supernatant was then filtered using a 0.45 μm hydrophilic polyvinyldene
fluoride filter (PVDF) and 100 μl of the filtrate was introduced into the SEC-ICP-MS system.
The specific use of PVDF filters was adequate to avoid protein losses.
Additionally, two different extraction procedures were applied to analyze selenium
species found in free form and bound to the proteins using HPLC. For extraction of free
selenium compounds, 2 ml of 0.1 mol l-1 HCl were added to 0.075 g of sample weighed in a
glass vial. The mixture was then stirred on a magnetic stirrer for 1 hour and then filtered using
33
0.45 μm PVDF filter. The resulting solution (20 μl) was injected into the reversed phase ion-
pairing HPLC-ICP-MS system for separation and determination of the selenium species.
To extract selenium compounds bound to proteins present in the sample, 5 ml of a Tris
buffer solution (pH 8.0) containing 1mM CaCl2 was added to 0.1 g of sample and 0.03 g of
Proteinase K enzyme. The solution was kept at a constant temperature of 50 oC and constantly
stirred for 18 hours. After that time, 0.03 g of Protease XIV was added to the above mixture and
kept at 50 o C with constant stirring for 12 hours. The final mixture was filtered with 0.45 μm
PVDF filter. The resulting solution (20 μl) was injected into the ion-pairing HPLC-ICP-MS
system.
2.3.6 ESI-ITMS analysis
In order to characterize unknown selenium compounds for which we had no retention time
standards, the appropriate chromatographic peaks were collected 20 times and lyophilized. The
residue was dissolved in 50 μl of methanol. The mixture was analyzed by ESI-ITMS. Molecular
ion and MS2 spectral information was obtained utilizing the conditions showed in Table 2.1.
2.4 RESULTS AND DISCUSSION
2.4.1 Plant growth and total selenium accumulation
The Green Onions were grown in promix medium enriched with different concentrations of
Se(IV). The total selenium content of the plants grown in presence of different Se(IV) amounts is summarized in Table 2.2. Plants grown in a medium enriched with 1 μg g-1 of Se did not
accumulate detectable amounts of selenium. It can be observed that the selenium amount
accumulated by the green onion plants was proportional to the Se(IV) added to the growth
media. No differences in terms of growth between the sample plants and the control plants were
observed due to the presence of selenium in potting medium during the green onion growth.
34
Therefore, selenium at the concentrations used did not seem to inhibit appreciably the growth
and maturity of the green onions in the studied range of selenium. The selenium content obtained
with 15 μg g-1 added to the growth medium was 30.3 μg Se g-1 of plant tissue. The speciation
studies were performed on the plants with the highest Se concentration (30.3 μg Se g-1 of plant tissue).
Table 2.2. Uptake of Selenium by green onions from growing media (95% confidence
interval, n=6).
Se amount added to growing media Se content in green onion plant
(μg g-1) (μg g-1)
None ND
1 ND
2 1.0 ± 0.1
3 1.4 ± 0.2
5 5.4 ± 0.4
15 30 ± 2.1
2.4.2 Molecular weight distribution of selenium in green onion
Calibration of the SEC column was accomplished with a standard mixture of myoglobin (17
kDa), lysozyme (14.4 kDa), substance P (1.35 kDa) and (Gly)6 (0.36 kDa), which gave a range
2 showing a good linear response for the log10 of molecular weight vs. retention time (r = 0.9928).
The information obtained by different extraction procedures: alkaline, acidic and
enzymatic applied on plant samples was compared. In the first approach to selenium speciation, the dried sample was treated with 0.1 mol l-1 sodium hydroxide for alkaline extraction of both
35
high and low molecular weight fractions. The selenium species extracted by that procedure were
separated and determined by SEC-ICP-MS. The size-exclusion chromatogram obtained for plant
sample is shown in Figure 2.1(a).
Figure 2.1(a): SEC-ICP-MS (78Se ) chromatogram of NaOH extract of enriched green onions
The fractionation profile shows a predominant high molecular weight (HMW) selenium-
containing peak that corresponds to an apparent molecular weight of about 12 kDa. Additionally,
two other selenium-containing fractions were observed at 1.6 kDa and 0.8 Da. These findings
confirm association of selenium in green onion not only to HMW compounds, perhaps proteins,
but also to LMW fractions like inorganic selenium, selenoamino acids and selenopeptides. In
fact, previous studies have demonstrated the possible association of selenium to proteins in
allium vegetables.13 It may be that the HMW fraction observed in the SEC chromatogram of the
NaOH extraction is due to the presence of one or more selenium containing proteins in green
onion. The peak area measurement showed that around 44% of the total selenium present in the
injection volume was attached to the HMW fraction, while about 56% was distributed among the
36
two LMW fractions. Selenium containing proteins potentially represent an important proportion
of the total selenium extracted by alkaline media.
Since the aim of this work was the identification of LMW selenium-containing species in
green onion, two additional extraction procedures were tested in this study. Extraction of non-
bound selenium species was performed by treating the freeze dried plant sample with 0.1 mol l-1
HCl solution. The extracted solution was injected into the SEC-ICP-MS for speciation of selenium. The chromatographic profile obtained is presented in Figure 2.1(b). As can be seen, the elution profile reveals that primarily LMW fractions were extracted with 0.1 mol l-1 HCl.
Figure 2.1(b): SEC-ICP-MS (78Se ) chromatogram of HCl extract of enriched green onions
These LMW fractions could correspond mainly to the presence of inorganic selenium,
free amino acids and small peptides which can be solubilized in the acidic media. On the other
hand, HMW compounds, such as proteins, were not significantly extracted due to the low solubility of these compounds at the low pH value. This result matches with results obtained in our laboratory on Brassica leaves20 and may confirm the association of selenium to the HMW
fraction in the form of selenium containing proteins.
37
In order to release selenium species bound to proteins, enzymatic hydrolysis was
employed on the green onion plant sample. It has been demonstrated that extent of protein
hydrolysis can be increased by the action of two enzymes more than by one acting alone21.
Consequently, enzymatic hydrolysis was performed with the combination of two enzymes,
Proteinase K and Protease XIV. The extract was analyzed by using the SEC-ICP-MS. The chromatographic fractionation profiles revealed a complete conversion of the HMW peak in the
Figure 2.1(a) to a single LMW selenium-containing fraction (Figure 2.1(c)). This can be explained by hydrolysis of selenium containing proteins obtaining LMW compounds such as selenoamino acids and selenopeptides generated by incomplete hydrolysis of proteins. The results support presumptions about the association of selenium to proteins obtained from the extractions of green onion samples in alkaline and acidic media.
Figure 2.1(c): SEC-ICP-MS (78Se) chromatogram of enzymatic extract of enriched green onions
2.4.3 Speciation studies of selenium by HPLC-ICP-MS
Reversed phase ion paring chromatography has been previously used for the speciation of
selenium compounds in plant samples.13 Ion paring reagents like perfluorinated carboxylic acids
and hexanesulfonic acids have been used. Generally, heptafluorobutyric acid (HFBA) is used for
38
compounds eluting earlier and for later eluting compounds trifluoroacetic acid (TFA) is
employed in the mobile phase. In the present work, complete separation of the selenium species
present in the different extraction solutions was obtained with a mobile phase containing 0.1%
(v/v) HFBA and 5% (v/v) MeOH. The use of a methanol concentration lower than 10% (v/v) in the mobile phase, avoided instability effects on the plasma permitting the sensitive detection of selenium by ICP-MS. Under these conditions (Table 2.1), separation of the selenium species was reached with a minimal retention time of 20 min.
Chromatographic information on the speciation analysis of selenium in green onion can
be observed in Figure 2.2. Since the identification of the selenium species was performed by
retention time matching of the peaks with those corresponding to different selenium standards, of
the retention time for the selenium standard was needed. Figure 2.2(a) shows a typical
chromatogram of mixed selenium standards with the following order of elution: tret = 2.0 min
inorganic selenium; tret=4.9 min Se-cystine; tret=7.2 min Semethylselenocysteine; and tret=19.9
min Se-methionine.
Figure 2.2 (a): RP-IP-HPLC-ICP-MS (78Se ) chromatograms of mixed selenium standards
39
Figure2.2(b): RP-IP-HPLC-ICP-MS (78Se) chromatogram of HCl extract of enriched green onion
The chromatographic profile shown in Figure 2.2(b) indicates that only two predominant
selenium species are present in the HCl extraction media. The first peak, at tret=7.2 min, was
identified as Se-methylselenocysteine, while the last small peak, at tret=19.8 min was assigned to
Se-methionine since the retention time of these peaks matched with the retention time of
selenium standards. The presence of these selenium species was also confirmed by standard
addition. Another predominant selenium-containing species is also present in the chromatogram
but its retention time did not match with retention time of any of the selenium standards
available. Therefore, identification of the unknown peak had to be performed utilizing molecular
mass spectrometry.
The enzymatic extract of the green onion plant sample was introduced into HPLC-ICP-
MS system for speciation analysis of selenium. From the chromatographic information shown in
Figure 2.2(c), it is possible to observe that inorganic selenium, Se-cystine, and Se-methionine
are the primary species found in the enzymatic hydrolysis media. Additional peaks
40
corresponding to unknown selenium-containing species were also observed. One of them, at
tret=11.4 min, was similarly eluted to that obtained in the HCl extract solution (Figure. 2.2(b)).
Figure 2.2(c): RP-IP-HPLC-ICP-MS (78Se) chromatograms of enzymatic extract of enriched green onion 2.4.4. Investigation of unknown selenium species by ESI-ITMS
Several fractions of the unknown peak at tret =11.4 min in the chromatogram of the HCl extract
were collected and after preconcentration, the resultant solution was introduced into ESI-ITMS.
The mass spectrum of the injected solution and the MS/MS spectra of the molecular ion [M+H]+
(m/z = 313) are shown in Figure 2.3(a) and 2.3(b). As observed from the MS/MS spectra
(Figure. 2.3(b)) the fragments obtained for the isolated molecular ion m/z=313 match what might be expected for γ-glutamyl-Se-methylselenocysteine (Figure. 2.3(c)). The fragmentation pattern obtained (Figure. 2.3 (b)) is consistent with those obtained by others14, 15.
41
Figure 2.3(a): Mass spectrum showing the resultant molecular ion with m/z=312 (m/z=313 for [M+H]+) of the unknown peak in HCl and enzymatic extract.
Thus, the primary species identified in the HCl extract were Se-methylselenocysteine and
γ-glutamyl-Se-methylselenocysteine plus Se-methionine found at considerably lower levels.
These studies indicate that selenium incorporation in the green onion bulb and leaves leads to the
formation of the same predominant species as in bulbs of other allium plants.8, 14, 15 Since γ-
Glutamyl-Se-methylselenocysteine is carrier of selenomethylselenocysteine, which is the major
selenoamino acid producing methylselenol (the major metabolite of anticancer activity), green
onion may be useful as a source of dietary selenium when Se enrichment conditions are utilized
in growing the plant.
42
b
c
Figure 2.3 (b) Mass spectrum of the fragmentation of m/z=313; (c) Fragmentation of the γ-glutamyl-Se- methylselenocysteine molecule.
Two remaining unidentified species were found in the chromatogram after enzymatic hydrolysis (Figure. 2.2(c)). Efforts were made to identify these species using electrospray mass spectrometry but the poorer sensitivity of ESI-MS in comparison to ICP-MS and considering that the species concentration in the sample was very low, the initial identification attempts were not
successful, although peptides are possible9,15, 16. This is also supported by the absence of these
selenium compounds in the HCl extract. Future studies will include enrichment of green onions
with higher concentrations of selenium for more detailed identification of selenium species and
43
effect of concentration of selenium in growth medium on the distribution of different selenium
species.
2.5 CONCLUSIONS
In this study, the chromatographic separation was coupled with ICP-MS detection to study the speciation of selenium in green onion plants grown in the presence of Se(IV). The green onions accumulated a significant amount of selenium in the plant tissue when this element was added to the growing medium. SEC experiments showed the incorporation of Se into HMW compounds
(≥12 kDa) as well as LMW compounds (1.8 kDa and 0.6 kDa fractions). The application of different extraction procedures, including NaOH, HCl, and enzymatic hydrolysis, suggests the association of 43% of the total Se present in the injection volume to proteins. The NaOH solution was effective for the extraction of both HMW and LMW selenium species, while HCl only permitted the extraction of LMW non-bound selenium species from the green onion tissues.
Reversed phase ion-pairing HPLC-ICP-MS and ESI-ITMS were useful analytical tools
for identification of γ-glutamyl-Se-methylselenocysteine, Se-cystine,
Selenomethylselenocysteine, Se-methionine, and inorganic selenium in green onions. These
studies indicate that selenium incorporation in the green onion leads to the formation of the same
predominant species as in bulbs of other allium plants. Since γ-Glutamyl-Se-
methylselenocysteine is carrier of selenomethylselenocysteine, which is the major selenoamino
acid producing methylselenol (the major metabolite of anticancer activity), green onion may be
useful as a source of dietary selenium when Se enrichment conditions are utilized. Future studies
will include enrichment of green onions with higher concentrations of selenium for more detailed
identification of selenium species and effect of concentration of selenium in growth medium on
the distribution of different selenium species.
44
2.6 REFERENCES
(1) Rayman, M. P. Lancet 2000, 356, 233-241. (2) Arteel, G. E.; Sies, H. Environmental Toxicology and Pharmacology 2001, 10, 153-158. (3) Holben, D. H.; Smith, A. M. Journal of the American Dietetic Association 1999, 99, 836- 843. (4) Clark, L. C.; Dalkin, B.; Krongrad, A.; Combs, G. F.; Turnbull, B. W.; Slate, E. H.; Witherington, R.; Herlong, J. H.; Janosko, E.; Carpenter, D.; Borosso, C.; Falk, S.; Rounder, J. British Journal of Urology 1998, 81, 730-734. (5) Ip, C.; Lisk, D. J.; Stoewsand, G. S. Nutrition and Cancer-an International Journal 1992, 17, 279-286. (6) Ip, C.; Lisk, D. J. Nutrition and Cancer-an International Journal 1993, 20, 129-137. (7) Ip, C.; Lisk, D. J. Carcinogenesis 1994, 15, 1881-1885. (8) Ip, C.; Dong, Y.; Ganther, H. E. Cancer and Metastasis Reviews 2002, 21, 281-289. (9) Montes-Bayon, M.; Yanes, E. G.; de Leon, C. P.; Jayasimhulu, K.; Stalcup, A.; Shann, J.; Caruso, J. A. Analytical Chemistry 2002, 74, 107-113. (10) B'Hymer, C.; Caruso, J. A. Journal of Analytical Atomic Spectrometry 2000, 15, 1531- 1539. (11) Bronzetti, G.; Cini, M.; Andreoli, E.; Caltavuturo, L.; Panunzio, M.; Della Croce, C. Mutation Research-Genetic Toxicology and Environmental Mutagenesis 2001, 496, 105- 115. (12) Ip, C.; Birringer, M.; Block, E.; Kotrebai, M.; Tyson, J. F.; Uden, P. C.; Lisk, D. J. Journal of Agricultural and Food Chemistry 2000, 48, 2062-2070. (13) Kotrebai, M.; Birringer, M.; Tyson, J. F.; Block, E.; Uden, P. C. Analyst 1999, 125, 71- 78. (14) Kotrebai, M.; Birringer, M.; Tyson, J. F.; Block, E.; Uden, P. C. Analytical Communications 1999, 36, 249-252. (15) Chassaigne, H.; Chery, C. C.; Bordin, G.; Rodriguez, A. R. Journal of Chromatography A 2002, 976, 409-422. (16) Montes-Bayon, M.; LeDuc, D. L.; Terry, N.; Caruso, J. A. Journal of Analytical Atomic Spectrometry 2002, 17, 872-879. (17) McSheehy, S.; Pohl, P.; Szpunar, J.; Potin-Gautier, M.; Lobinski, R. Journal of Analytical Atomic Spectrometry 2001, 16, 68-73. (18) Casiot, C.; Vacchina, V.; Chassaigne, H.; Szpunar, J.; Potin-Gautier, P.; Lobinski, R. Analytical Communications 1999, 36, 77-80. (19) Wróbel, K.; Wróbel, K.; Kannamkumarath, S. S.; Caruso, J. A.; Wysocka, I. A.; Bulska, E.; Swiatek, J.; Wierzbicka, M. Food Chemistry 2003, in press. (20) Montes-Bayon, M.; Grant, T. D.; Meija, J.; Caruso, J. A. Journal of Analytical Atomic Spectrometry 2002, 17, 1015-1023. (21) Wrobel, K.; Kannamkumarath, S. S.; Caruso, J. A. Analytical and Bioanalytical Chemistry 2003, 375, 133-138. (22) McSheehy, S.; Yang, W. J.; Pannier, F.; Szpunar, J.; Lobinski, R.; Auger, J.; Potin- Gautier, M. Analytica Chimica Acta 2000, 421, 147-153. (23) Uden, P. C.; Bird, S. M.; Kotrebai, M.; Nolibos, P.; Tyson, J. F.; Block, E.; Denoyer, E. Fresenius Journal of Analytical Chemistry 1998, 362, 447-456.
45
CHAPTER 3 | CHARACTERIZATION OF IODINE CONTAINING SPECIES IN COMMERCIALLY AVAILABLE SEAWEEDS
3.1 ABSTRACT
Speciation of iodine in commercially available commonly consumed seaweed samples was
performed using a multidimensional chromatographic approach coupled with inductively coupled plasma mass spectrometry (ICP-MS) for element specific detection. Analysis of alkaline extract (0.1 mol l-1 NaOH) by size-exclusion chromatography coupled to ICP-MS (0.03 mol l-1
Tris-HCl pH 8.0) indicated the association of iodine in both high as well as low molecular weight fractions in Wakame while in case of Kombu, only low molecular weight iodine species were found. Likely association of iodine with protein as well as polyphenolic species was indicated in case of Wakame. Anion-exchange chromatography coupled to ICP-MS (0.005 mol l-
1 NaOH) confirmed that the most predominant inorganic iodine species present in both type of seaweeds is iodide. Protein bound iodinated species were hydrolyzed by enzymatic digestion
using Proteinase K. Analysis of the hydrolysate using reversed-phase HPLC-ICP-MS (0.01 mol
l-1 Tris-HCl pH 7.3: 0.01 mol l-1 Tris-HCl pH 7.3 and 50%MeOH) revealed the presence of
monoiodotyrosine and di-iodotyrosine in Wakame, which was later identified by matching the
chromatographic retention time with the retention time of commercially available standards.
47
3.2 INTRODUCTION
Iodine is an essential micromineral for human nutrition. Its role in the synthesis of thyroid
hormones necessary for human growth and development such as thyroxin, (tetraiodothyronine
1-3 (T4)) and triiodothyronine (T3), is widely known. Iodine deficiency leads to goiter and various
disorders associated with growth and development like dwarfism, mental retardation and
neuromuscular defects, commonly referred as Iodine deficiency disorders (IDD).4, 5 The
recommended dietary allowance (RDA) of iodine is 150µg/d in the United States. In European
and other countries, ranges between 150-200 µg/d have been established.5 In order to prevent
iodine deficiency disorders, supplementation of foodstuff with iodine is commonly practiced.6
Milk, a source of iodine of animal origin obtains most of it from iodine supplemented cattle feed.
Usage of iodized salt has also become popular worldwide.6 Iodine is plentiful in oceans and
marine animals and sea plants such as some marine algea (seaweed) are naturally occurring
sources of dietary iodine. Some seaweed can accumulate exceptionally high quantities of iodine available from the sea. Hizikia (Hiziki), Undaria (Wakame), Laminaria (Kombu) and Porphyra
(Nori) are some of the commercially available seaweed commonly consumed with very high
iodine content.7 Total iodine content of seaweed depends on the species as well as the region in
which they are found. The enrichment factor of Laminaria japonica for iodine reaches 106.8 The consumer preference of natural products over artificial ones provides impetus for studying seaweed as source of iodine.9 Consumption of seaweed as a dietary source of iodine is not only
restricted to Asian countries, but also increasingly in Europe and some African countries like
Ghana.5, 10 Recently, possibility of using marine algae Wakame (Undaria pinnatifida) and
Kombu (Laminaria digita japonica) as a food supplement has also been evaluated.11
48
The key to good thyroid function is adequate, but not excessive, iodine intake, which can
not only cause high-iodine goiter, but also damage the nervous system leading to retarded brain
development and functional impediment.12-14 Another important fact associated with consumption of iodine is that, like other elements, bioavailability and toxicity is species dependent. Inorganic forms of iodine such as iodide and iodate are less toxic than molecular
iodine and some organically bound iodine.8 Likewise, bioavailability of organically bound iodine
such as monoiodotyrosine (MIT), di-iodotyrosine (DIT) is also less than mineral iodide.9
Considering above, and the increasing use of seaweed as a dietary source of iodine, total analysis and characterization of iodine species in seaweed is an important pursuit.
Many different analytical techniques have been developed for iodine speciation in food, environmental and biological matrices. Interfacing various separation techniques such as reversed-phase high performance liquid chromatography (RP-HPLC)19, 20, ion chromatography
(IC)21, 22, size-exclusion chromatography (SEC)23 and capillary electrophoresis (CE)24 with element selective ICP-MS detection has been employed in speciation studies of iodine to separate and characterize various iodine species. Although, numerous papers have been published reporting total analysis of iodine in food samples4, 15-17, speciation studies of iodine in
seaweed has only been recently demonstrated, in spite of the extensive consumption.18
Moreover, the well known capabilities of multidimensional chromatographic techniques coupled to elemental specific detectors such as ICP-MS, for fast and reliable speciation analysis of iodine in seaweed, has never been reported.
The aim of the present study is an initial characterization and identification of iodine species in commercially available seaweed samples. An approach involving multidimensional chromatographic techniques coupled to ICP-MS is utilized for the separation and determination
49
of iodine species in seaweed. Size exclusion chromatography was used to investigate the association of iodine with various molecular weight fractions extracted from different media and to separate inorganic iodine from organically bound iodine. Anion-exchange chromatography was used for separating inorganic forms of iodine. Reversed-phase high performance liquid chromatography was used for the separation and identification of low molecular weight iodine species in seaweed samples. Identification of iodine species was performed by matching the peak retention times with those of standards.
3.3 EXPERIMENTAL
3.3.1 Instrumentation
Chromatographic separations were performed using an Agilent 1100 liquid chromatographic
system (Agilent Technologies, Palo Alto CA, USA) equipped with an HPLC binary pump, an
auto sampler, vacuum degasser, a thermostated column compartment and a diode array detector.
The chromatographic columns used were a Superdex 75 HR (10mm x 300mm x 13um particle
size) column (Amersham Pharmacia Biotech AB, Uppsala, Sweden) for size-exclusion
chromatography, an Ion pac AS-11 (Dionex, Sunnyvale, CA, USA) column (2.0mm i.d. x
250mm length x 13 um particle size) for ion exchange chromatography and a C18 Alltima
(Alltech, Deerfield, IL, USA) column (4.6mm i.d. x 150mm length x 5um particle size) for RP-
HPLC. Chromatographic conditions are summarized in Table1.
An Agilent 7500ce ICP-MS (Agilent Technologies, Tokyo, Japan) equipped with a
micromist nebulizer and a Peltier cooled spray chamber (2oC) and a shielded torch system was
used for iodine specific detection. The outlet of the UV detector was connected online to the
liquid sample inlet of the ICP-MS nebulizer using a 300 mm long by 0.25 mm PEEK tubing. For
RP-HPLC, online dilution of the chromatographic eluent containing organic solvent was
50
performed as follows before its nebulization into the plasma: the outlet carrying the chromatographic eluent at a flow arte of 0.5 ml min-1 was connected to the one arm of PTFE tee piece while through the other arm 2% nitric acid was introduced at a flow rate of 0.5 ml min-1 in order to provide a dilution factor of 1:2. Dilution was performed to reduce the organic solvent
(methanol) load introduced into the plasma. The instrumental operating conditions are summarized in Table 3.1.
51
Table 3.1. ICP-MS and chromatographic instrumental parameters
ICP-MS parameters Forward power 1500 W Plasma gas flow rate 15.0 l min-1 Auxiliary gas flow rate 0.87 l min-1 Carrier gas flow rate 1.20 l min-1 Dwell time 0.1 s per isotope Isotopes monitored 127I SEC chromatography parameters Column Superdex 75 HR 10/30 Mobile phase 0.03 mol⋅l-1 Tris-HCl buffer, pH 8.0 Flow rate 0.6 ml⋅min-1 Injection volume 100 μl Ion chromatography parameters Column Ion Pac AS-11 anion exchange column( 250 mm x 2.0mm i.d. x 13µm) Mobile phase 0.005 mol l-1 Sodium hydroxide Flow rate 0.3 ml min-1 Injection volume 20 μl RP-HPLC parameters
Column Alltima C18 (150 mm x 4.6 mm, 5 μm) Mobile phase (A) 0.01 mol l-1 Tris-HCl(pH 7.3): (B) 0.01 mol l-1 Tris-HCl (pH 7.3) and 50% MeOH Flow rate 0.5 ml min-1 Injection volume 50 μl Make up solution 2% (v/v) HNO3; 0.5 ml min-1 Gradient 0-5 min-100% A to 45% B; 5-8 min-45% B to 85% B; 8-10 min-85% B to 100% B and 10-40 min 100% B
3.3.2 Reagents and Standards
All reagents used were analytical grade reagents and presence of iodine was not detected in the working range. All solutions were prepared in 18 MΩ cm doubly deionized water generated by a
52
NanoPure treatment system (Barnstead, Boston, MA, USA). The following reagents were
purchased from Sigma (Sigma-Aldrich Co, St-Louis, Mo, USA): tris (hydroxymethyl)
aminomethane (TRIS), ammonium bicarbonate and sodium dodecyl sulfate (SDS). Hydrochloric
acid, sodium hydroxide, and potassium iodate were obtained from Fisher (Fairlawn, NJ, USA).
HPLC grade methanol, ethanol, and acetone (Fisher Scientific, Pittsburgh, PA, USA) were used
throughout. Individual standard solutions of 3-iodotyrosine (MP Biomedicals, Irvine, CA, USA)
and 3,5-diiodotyrosine (Cambridge Corporation, San Diego, CA, USA) were prepared by
dissolution of chemicals in methanol. For the determination of total iodine, working solutions
were prepared daily by appropriate dilution of the 10 µg ml-1 iodide standard solution obtained from High-Purity Standards (Charleston, SC, USA). Nitric acid (Suprapure) 68% from Pharmaco and hydrogen peroxide 30% from Fisher Scientific were used for sample digestion.
3.3.3 Procedures
3.3.3.1 Sample collection and preparation
Commercially available dried seaweed samples [marine algae Kombu (Laminaria japonica) and
Wakame (Undaria pinnatifida)] were obtained from local Asian stores in USA for total iodine
analysis and speciation studies. The above mentioned marine algae were selected for speciation
studies because they are commonly consumed, especially in various Asian recipes and they
represent sources of high iodine concentration. The dried algae samples were ground in a
household coffee grinder.
3.3.3.2 Total iodine determination
Both types of seaweeds were analyzed for total iodine content by ICP-MS after complete
digestion using an MES 1000 closed vessel microwave digestion system (CEM corp., Matthews,
NC, USA). Sample amounts of 0.1g were weighed and 5 ml of 65% concentrated nitric acid and
53
1ml of 30% hydrogen peroxide were added in microwave vessels. The following mineralization
program was applied for complete digestion: the microwave power was increased over three
steps with 5 min interval from 250 W to 500 W and finally to 1000 W and held at 1000 W for
another 25 min. Temperature limits of 120,150, 160 and 200 oC were set for each of four steps.
The mixture was cooled for 20min at room temperature. The digested samples were adjusted to a
pH between 9 and 10, by addition of 3% NH3 at ratio of 1:1. The addition of NH3 was performed
to give a stable iodide signal4. And finally the samples were diluted to 100 ml with doubly
deionized water. Reagent blank was digested the same way. Tellurium was added as an internal
standard. Appropriate dilution was performed when needed. The method of standard additions
(at µg l-1 of iodine) was applied for quantification of the iodine present in the samples.
3.3.2.3 Extraction of iodine from seaweed in various media
For SEC studies, iodine was extracted from samples with each of the following solutions: 0.1mol
l-1 NaOH for extracting high as well as low molecular weight fractions25, 0.1 mol l-1 HCl for extraction of low molecular weight fraction26 and aqueous buffer 0.03 mol l-1 Tris-HCl (pH-8.0).
Approximately 0.1g of both type of seaweed samples were weighed in different glass vials and
treated separately with 10 ml of above mentioned solutions. Extraction was carried out for 30
min at room temperature with constant stirring using a magnetic stirrer. The extracts were
centrifuged for 10 min at 5000 rpm and the supernatant was separated from the residue. Dilutions
of the above extracts were performed 1:10 and 1:50 times for Wakame and Kombu samples, respectively. After previous filtration with 0.45 µm PVDF filters, 100 µl of the resulting solutions was introduced into SEC-UV-ICP-MS system.
54 3.3.2.4 Extraction of high M.W. iodine species from Wakame
Investigation of association of iodine with high molecular weight fractions was performed by
extracting possible proteins from the Wakame sample according to Hou et al18 and analyzing the
extract by SEC-UV-ICP-MS. About 1.0g of sample was weighed in a glass vial and leached
three times with acetone to remove pigments from the plant matrix. Centrifugation was
performed each time in order to separate residue from the supernatant (acetone). The residue
obtained from the above step was treated with a solution containing 1% CaCl2 and 0.5% caffeine
to remove carbohydrates and polyphenols, respectively from the plant matrix. Centrifugation was
performed at 5000 rpm for 10 min and supernatant was discarded and the residue was collected.
Finally, the residue was sonicated for 48h in 10ml of a 0.03 mol l-1Tris-HCl (pH-8.0) solution
containing 1% SDS, 0.05% NaN3 and 100 mg of PVPP (polyvinyl-poly-pyrrolidone). PVPP is
also used for absorbing polyphenolic compounds from the plant matrix. After centrifugation (10
min, 5000 rpm), proteins were precipitated by adding acetone up to a final concentration of 80% and the solution was kept at -20oC for 12 h. Finally, the precipitate was collected and solubilized in 3 ml of the above Tris-HCl solution containing 1% SDS. Before introduction to SEC-UV-ICP-
MS system, the above solution was diluted 1:10 with Tris-HCl buffer used in the mobile phase.
Iodine-containing high M.W. species (most likely proteins) were digested by employing
enzymatic hydrolysis with Proteinase K.27 About 10 ml of a Tris-HCl (pH 8.0) buffer solution
containing 1 mM CaCl2 was added to 1 ml of protein extract with 0.03g of Proteinase K enzyme.
The solution was kept at constant temperature of 50oC with continuous stirring for 12 h. Addition
of equal amount of Proteinase K and stirring were repeated. The final mixture was filtered with
0.45µm filters. A volume of 100µl of the resulting solution was injected into SEC-ICP-MS
system and 20 µl was injected to RP-HPLC-ICP-MS system. Prior to injection into RP-HPLC-
ICP-MS system, the hydrolysate was filtered through 10 kDa molecular weight cut-off
membrane filters (Centricon) to remove partially digested proteins and enzyme.
3.3.3.5 Extraction of polyphenol-bound iodine from Wakame
Iodine bound to polyphenols was extracted by treating about 0.5g of sample with 10 ml of 75% ethanol three times at room temperature. Supernatants from above step were combined and concentrated to a small volume under reduced pressure at 50oC. Final solution was filtered and
100µl was introduced into SEC-ICP-MS system.
3.3.4 Chromatographic Conditions
Calibration of Size-exclusion column was performed by following set of protein standards:
albumin (66 kDa); myoglobin (17.6 kDa); aprotinin (6.5 kDa) and Vitamin B12 (1.35 kDa)
In case of RP-HPLC-ICP-MS, optimization of chromatographic conditions was
performed so that final flow introduced into the ICP-MS system does not interfere with the
normal nebulization process and the stability of plasma was maintained throughout the
chromatographic run. Final flow of 1.0 ml/min was introduced into to the sample introduction
system of ICP-MS obtained by combining the flow from the chromatographic column with an
external flow of 2% nitric acid solution (both 0.5ml/min) using a T junction after the column.
3.4 RESULTS AND DISCUSSION
3.4.1 Total iodine concentration in samples and extracts
Initially, total iodine determination was performed on the ground sample. Total iodine was also determined in the three extraction media for both seaweed samples.
Results of the total iodine determination are summarized in Table 3.2. For validation purposes,
standard reference material (Citrus leaves-NIST 1572) was also analyzed for total iodine as
described in experimental section. The obtained result (2.0 ± 0.05 µg g-1) was in reasonable
56
agreement with the certified value. (1.84 ± 0.03 µg g-1). For HCl and Tris-HCl extract of both the
samples, the pH of was adjusted in the range of 9-11 with diluted ammonia solution for stabilizing the iodine signal.
Table 3.2: Comparison of different media for the extraction of iodine from seaweed
samples and inorganic iodine species present (95% Confidence Interval, n = 6)
Total Content (µg Extractability (% ) Iodide (µg g-1) Iodate (µg g-1)
g- 1) (%RSD) NaOH HCl Tris-HCL
Kombu 4170 (5.6) 93±2.5 42±0.5 79±1.2 3940 Not detectable
Wakame 226 (4.8) 98±2.8 20 ±0.8 30±0.3 140 4.16
3.4.2 Optimization of SEC-ICP-MS
Size exclusion chromatography was utilized as a preliminary method to understand the
distribution of iodine in different molecular weight fractions in the seaweed extracts.
Chromatographic conditions were optimized to obtain separation between various forms of
iodine in the shortest time possible. Buffers including 0.03 mol l-1 ammonium bicarbonate (pH
8.0), 0.03 mol l -1 ammonium carbonate (pH 9.0) and 0.03 mol l-1 Tris-HCl (pH 8.0) were
assayed as mobile phases for SEC. Tris-HCl was selected as it provided least amount of non-
specific interaction between the inorganic iodine and the stationary phase of the column,
resulting in faster elution of inorganic iodine species from the column as compared to above two
buffers.
57 3.4.3 Fractionation studies of iodine in seaweed extracts
3.4.3.1 SEC-ICP-MS of Wakame extracts
As a first approach to iodine speciation, seaweed was treated with the 0.1 mol l-1 sodium hydroxide solution. Alkaline solutions are known to solubilize both high as well as low molecular weight fractions. It is clear from Figure 3.1(a) that in Wakame iodine is associated to both low as well as high molecular weight fractions. The fractionation profile shows that three peaks of significant intensities are found in the size exclusion chromatogram. The first peak eluting at the dead volume of the column (≥ 70kDa) represents the association of iodine with the high molecular weight compounds presumably proteins and/or carbohydrates as these species are soluble in alkaline media.
Figure 3.1 (a): 127I SEC-ICP-MS chromatogram of NaOH extract of Wakame
Another peak at 20.8 min indicates association of iodine with medium molecular weight fraction
of about 6.5 kDa. This peak may be due to the presence of peptides containing iodo amino acids
and iodinated polyphenolic species. The last peak eluting after the elution volume of size
exclusion column is identified as the inorganic iodine species, which represents a significant
portion of iodine in the Wakame algae. Delayed elution of iodide is attributed to its non-specific
interaction between polysaccharide based stationary phases of the size-exclusion column. Since the size exclusion column does not have the capability to separate inorganic forms of iodine, further speciation of inorganic iodine fraction was performed by anion exchange chromatography, which is discussed later in the text.
Extraction of iodine species in 0.1 mol l-1 HCl and in Tris-HCl solution was also evaluated.
In acidic medium, fractions containing only low molecular weight iodinated species are extracted, which is depicted in corresponding size-exclusion chromatogram Figure 3.1(b). Only one significant peak is depicted at 21.7 min in the molecular weight of about 3.1 kDa. This peak is might be due to iodine containing peptides. Since the oxidation of iodide to elemental iodine is enhanced in acidic solutions, peak corresponding to the inorganic form of iodide could be diminished. It also is noted from the chromatogram that high molecular weight species eluting at the dead volume of the column are not extracted in acidic media. This is probably due to low solubility of biological macromolecules, such as proteins, at acidic pH.
Figure 3.1 (b): 127I SEC-ICP-MS chromatogram of HCl extract of Wakame
A comparison of the chromatograms shown in Figures 3.1(a) and (b) indicates an association of iodine to high molecular weight fractions, which can be presumed to be proteins
59
based on their different solubility in acidic or basic media. An aqueous buffer Tris-HCl (pH 8.0) was investigated for extracting proteins from algae as it is known to solubilize water soluble proteins.28 It can be noted from the size-exclusion chromatogram Figure 3.1(c) that high
molecular weight fraction is extracted in this medium eluting at dead volume of the column. This
suggests that in marine algae Wakame iodine is incorporated into water soluble high molecular
weight proteins.
Figure 3.1 (C): 127I SEC-ICP-MS chromatogram of Tris-HCl extract of Wakame 3.4.3.2 Study of the association of iodine to biological molecules
In order to confirm that iodine containing peak corresponding to the high molecular weight
fraction in size-exclusion chromatogram Figure 3.1(c) is only due to presence of iodine
containing proteins, a systematic approach leading to specific precipitation of proteins was
followed as described above. Interferences from other high molecular weight molecules such as
pigments, carbohydrates and polyphenols were avoided by removing them from plant matrix
using acetone, calcium chloride and PVPP respectively according to Hou et al.18 SDS, an anion
surfactant was added in the Tris buffer to extract the high molecular weight proteins.
Precipitation of proteins extracted in above solution was achieved by addition of 80% acetone.
60
High organic solvent concentrations have been used in the past for precipitation of both high as
well as low molecular weight proteins .28
The proteins thus extracted were injected in the SEC-UV-ICP-MS system and as shown in
Figure 3.2(a), only one peak containing high molecular weight iodinated species is eluted at the
dead volume of the column. The elution profile of this peak is similar to the elution profile of the
high molecular weight peak eluting at the column dead volume in the size-exclusion
chromatogram Figure 3.1(a). Therefore, it can be concluded that high molecular weight species
extracted in the alkaline solution is due to iodine-containing proteins. Presence of peak at the
column dead volume in the UV elution profile (295nm), shown in the same chromatogram, also
leads to the presence of proteins in the extract.
Figure 3.2 (a): 127I SEC-ICP-MS chromatogram of protein extract of Wakame (----) spectrophotometric detection (295
nm).
A further experiment was performed to understand the nature of iodine bonding to
proteins in Wakame. Proteinase K was employed for hydrolysis of proteins. The extract was
analyzed by SEC-ICP-MS. The chromatographic fractionation profile revealed the conversion of high molecular weight proteins Figure 3.2(b) into a low molecular weight fraction eluting at
61
about 26.7min. These species having molecular weight below 1.3 kDa may correspond to the
iodinated peptides and iodinated amino acids released after protein hydrolysis. This also suggests
that iodine is associated with proteins through covalent bonding.
Figure 3.2 (b): 127I SEC-ICP-MS chromatogram of enzymatic extract of Wakame
The presence of iodinated polyphenolic species was also investigated in Wakame
seaweed. Extraction of polyphenolic species was performed with 75% (v/v) ethanol according to
Hou et al.18 A volume of 100µl lyophilized solution was injected into the SEC-ICP-MS system.
It is suggested from Figure 3.2(c) that in Wakame seaweed iodine is also associated to polyphenolic species. Intensity of peak eluting at the column dead volume is very low, which can be explained by low solubility of proteins in organic solvent. Major polyphenolic species associated with iodine elute at 18.7 minutes, which corresponds to a molecular weight of approximately 7.9kDa. Iodine bound polyphenolic species have also been reported by Hou et al in Sargassum Kjellanianum.18 These polyphenolic species may correspond to the polymer of
iodine substituted p-chloroglucinol, since they are commonly found in brown seaweed.18
62
Figure 3.2 (c): 127I SEC-ICP-MS chromatogram of polyphenolic extract of Wakame 3.4.3.3 SEC-ICP-MS of Kombu extracts
A similar approach to that developed for Wakame seaweed was also followed for studying the distribution of iodine in Kombu seaweed. Three different extractions: alkaline, acidic and Tris-
HCl as in case of Wakame were performed. 100µl of extracted solution was injected into the
SEC-ICP-MS system for speciation of iodine The fractionation profile of iodine species extracted in alkaline medium shows only one predominant low molecular weight species corresponding to retention time of inorganic iodine species (Figure 3.3) Evaluation of iodine species extraction in acidic and Tris-HCl medium reveals the same results. This leads to the conclusion that metabolism of iodine in Kombu does not lead to formation of any significant organoiodine species and it is significantly different to the results obtained for Wakame. The major iodine species found in Kombu corresponds to inorganic iodide species and as indicated below, has been further identified by IC-ICP-MS.
63
Figure 3.3: 127I SEC-ICP-MS chromatogram of alkaline extract of Kombu seawwed 3.4.4 IC-ICP-MS for speciation of inorganic iodine
Ion chromatography has been previously applied for the separation of anionic iodine species.21, 22
In the present study, separation between iodide and iodate was performed with an anion- exchange column. Composition of the mobile phase and sample solution in which standards were prepared was optimized. Complete separation of iodide (1-tret = 1.9min) and iodate (2-tret=
6.7min) was performed using 5 mM sodium hydroxide as mobile phase (Figure 4(a)). Since the stability of the iodine species is affected by the pH of the solution, stability of iodine and iodate in alkaline medium (0.1 mol l-1 NaOH) was performed. This study was necessary in alkaline medium because the maximum extraction efficiency of iodine from the seaweed samples was obtained in this medium (Table 2). Composition of the mixture of the two species remains stable in this medium. The peak area precision (%RSD) of iodide and iodate standards injected at 10ng ml-1, based on three replicate measurements were well below 1.0% and 1.2% respectively.
Absolute limits of detection for both the species, calculated based on 3σ of the blank signal were
found to be 0.12 µg l-1 and 0.2 µg l-1.
64
Peak areas of a particular concentration of iodide and iodate in alkaline medium were found to be
similar to the peak areas of corresponding species when injected individually at the same
concentration. This suggests that in 0.1 mol l-1 NaOH solution, no significant loss or
interconversion of one form of iodine to another takes place, which makes it a suitable medium
for speciation of inorganic iodine species.
Figure 3.4 (a): 127I IC-ICP-MS chromatogram of mixed iodine standards at concentration of 10 ng ml-1 each prepared
in 0.1 mol l-1 NaOH solution. Alkaline extract of both Wakame and Kombu seaweed was analyzed for possible presence of
iodide and iodate. A volume of 20µl diluted solution was injected into the IC-ICP-MS system.
The chromatographic profile shown in Figure 3.4(b) reveals that predominant inorganic iodine species present in Wakame in alkaline extraction medium is iodide with a low amount of iodate is also present. While in the case of Kombu, the only iodine species present is iodide, which is depicted in IC-ICP-MS chromatogram (Figure 3.4(c)).
65
Figure 3.4 (b): 127I IC-ICP-MS chromatogram of alkaline extract of Wakame, and (c) Kombu seaweed 3.4.5 RP-HPLC-ICP-MS for studying iodine species
Reversed phase high performance liquid chromatography was used for speciation of low
molecular weight iodine species. Since association of iodine with proteins was proven in
Wakame seaweed, the extract obtained after enzymatic hydrolysis of the protein fraction was analyzed for the presence of iodoamino acids. Separation of iodoamino acids has been previously
achieved by RP-HPLC.30, 31 For the present study, a similar separation procedure as published by
Michalke20 was followed with some variation in the flow rate of mobile phase and gradient
program. In Figure 3.5(a) chromatograms of three laboratory available iodine standards is
presented. It is shown that baseline resolution was obtained between inorganic iodine (1-tret = 2.9
66
min), monoiodotyrosine (2-tret = 9.6min) and diiodotyrosine (3-tret = 12.1min). Also, it can be
noted from Table 1 that about 50% methanol was required for elution of iodoamino acids from the column. In order to reduce the organic load introduced into the plasma, online dilution of the chromatographic eluent, using 2% HNO3, was performed with a dilution factor of 1:2. Dilution using HNO3 not only increases the MeOH tolerance of the plasma, since it partially oxidizes the
methanol,20 but also reduces the deposition of carbon residue on the sampling and skimmer
cones surfaces. Rf plasma power was also increased to 1500W to circumvent the cooling effect
due to the introduction of organic solvent.
Figure 3.5 (a): 127I RP-HPLC-ICP-MS chromatogram of mixed iodine standards at 100 ng ml-1 each, and (b)
enzymatic extract of the Wakame seaweed.
67
Before injection, the hydrolysate was filtered through Centricon-10 (molecular weight
cut-off filter) to remove unhydrolysed higher M.W. material and Proteinase K to avoid column
loading. A volume of 50µl-filtered solution was introduced into the RP-HPLC-ICP-MS system.
From the chromatographic information shown in Figure 3.5(b), it is depicted that iodide,
monoiodotyrosine and di-iodotyrosine are the primary species present in the enzymatic extract.
Standard addition also gave enough evidence for the presence of these species. It can be noted
that no other iodo amino acids are eluted from the column, which suggests that iodine
metabolism in Wakame leads to formation of only mono and di-iodinated tyrosine, unlike human
metabolism where iodination can lead to formation of tetraiodinated species.
3.5 CONCLUSIONS
In this study, the applicability of several chromatographic techniques including SEC, IC-HPLC,
and RP-HPLC coupled to ICP-MS for iodine speciation in seaweed has been demonstrated.
Moreover, the use of hyphenated techniques for iodine speciation in seaweed extracts allowed obtaining important information on the association of iodine to the various matrix components of seaweed. Whereas iodide is about the most predominant species present in Kombu, a more complicated distribution of iodine is present in Wakame seaweed. This study shows that incorporation of iodine in different seaweeds follows different metabolic pathways, notwithstanding that both of them belong to same class, Phaeophyceae. Presence of iodide was proven in Kombu, while in case of Wakame, iodine monoiodotyrosine and di-iodotyrosine are also present and probably bound to the proteins. Since bioavailability of iodide is better than any other form of iodine, Kombu seaweed would be preferred as a natural dietary supplement.
68
3.6 REFERENCES
(1) Underwood, E. J. Trace elements in human and animal nutrition; Academic Press: New York, 1977. (2) Buchberger, W. Journal of Chromatography A 1988, 439, 127-135. (3) Glowa, G. A.; Mezyk, S. P. Radiation Physics and Chemistry 1998, 53, 127-135. (4) Eckhoff, K. M.; Maage, A. Journal of Food Composition and Analysis 1997, 10, 270- 282. (5) Serfor-Armah, Y.; Nyarko, B. J. B.; Carboo, D.; Osae, E. K.; Anim-Sampong, S.; Akaho, E. H. K. Journal of Radioanalytical and Nuclear Chemistry 2000, 245, 443-446. (6) Anke, M.; Groppel, B.; Muller, M.; Scholz, E.; Kramer, K. Fresenius Journal of Analytical Chemistry 1995, 352, 97-101. (7) Cann, S. A.; van Netten, J. P.; van Netten, C. Cancer Causes & Control 2000, 11, 121- 127. (8) Hou, X. L.; Chai, C. F.; Qian, Q. F.; Yan, X. J.; Fan, X. Science of the Total Environment 1997, 204, 215-221. (9) Aquaron, R.; Delange, F.; Marchal, P.; Lognone, V.; Ninane, L. Cellular and Molecular Biology 2002, 48, 563-569. (10) Ruperez, P.; Saura-Calixto, F. European Food Research and Technology 2001, 212, 349- 354. (11) Kolb, N.; Vallorani, L.; Milanovic, N.; Stocchi, V. Food Technology and Biotechnology 2004, 42, 57-61. (12) Fisher, D. A. Clinical Chemistry 1996, 42, 135-139. (13) Kim, J. Y.; Kim, K. R. Yonsei Medical Journal 2000, 41, 22-28. (14) Key, T. J. A.; Thorogood, M.; Keenan, J.; Long, A. Journal of Human Nutrition and Dietetics 1992, 5, 323-326. (15) Gelinas, Y.; Krushevska, A.; Barnes, R. M. Analytical Chemistry 1998, 70, 1021-1025. (16) van Netten, C.; Cann, S. A. H.; Morley, D. R.; van Netten, J. P. Science of the Total Environment 2000, 255, 169-175. (17) Fecher, P. A.; Goldmann, I.; Nagengast, A. Journal of Analytical Atomic Spectrometry 1998, 13, 977-982. (18) Hou, X. L.; Yan, X. J.; Chai, C. F. Journal of Radioanalytical and Nuclear Chemistry 2000, 245, 461-467. (19) Michalke, B.; Schramel, P.; Witte, H. Biological Trace Element Research 2000, 78, 67- 79. (20) Michalke, B.; Schramel, P.; Witte, H. Biological Trace Element Research 2000, 78, 81- 91. (21) Stark, H. J.; Mattusch, J.; Wennrich, R.; Mroczek, A. Fresenius Journal of Analytical Chemistry 1997, 359, 371-374. (22) Leiterer, M.; Truckenbrodt, D.; Franke, K. European Food Research and Technology 2001, 213, 150-153. (23) Sanchez, L. F.; Szpunar, J. Journal of Analytical Atomic Spectrometry 1999, 14, 1697- 1702. (24) Michalke, B.; Schramel, P. Electrophoresis 1999, 20, 2547-2553.
69
(25) Shah, M.; Kannamkumarath, S. S.; Wuilloud, J. C. A.; Wuilloud, R. G.; Caruso, J. A. Journal of Analytical Atomic Spectrometry 2004, 19, 381-386. (26) Montes-Bayon, M.; Grant, T. D.; Meija, J.; Caruso, J. A. Journal of Analytical Atomic Spectrometry 2002, 17, 1015-1023. (27) Wrobel, K.; Kannamkumarath, S. S.; Caruso, J. A. Analytical and Bioanalytical Chemistry 2003, 375, 133-138. (28) Mounicou, S.; Meija, J.; Caruso, J. Analyst 2004, 129, 116-123. (29) Evenson, J. K.; Sunde, R. A. Proceedings of the Society for Experimental Biology and Medicine 1988, 187, 169-180. (30) delaVieja, A.; Calero, M.; Santisteban, P.; Lamas, L. Journal of Chromatography B 1997, 688, 143-149. (31) Takatera, K.; Watanabe, T. Analytical Chemistry 1993, 65, 759-762.
70
CHAPTER 4 | DETERMINATION OF PHOSPHORIC ACID TRIESTERS IN HUMAN PLASMA USING SPME AND GC-ICP-MS
71
4.1 ABSTRACT
A simple and sensitive method for determination of phosphoric acid triesters at trace levels in human plasma sample is described. In this work, solid-phase microextraction (SPME) is employed as a sample preparation procedure for extraction and pre-concentration of alkyl and aryl phosphates followed by gas chromatography coupled to inductively coupled plasma mass spectrometry (GC-ICP-MS) for phosphorus-specific and very sensitive determination of these compounds in human plasma. The detection limits from blood plasma were 50 ng L-1 (tripropyl phosphate), 17 ng L-1 (tributyl phosphate), 240 ng L-1 (tris(2-chloroethyl) phosphate) and 24 ng
L-1 (triphenyl phosphate). Sample preparation involves plasma deproteinization followed by direct immersion SPME with 65 µm poly(dimethylsiloxane/divinylbenzene) fiber. Extraction was performed at 40 oC for 30 min and at pH 7.0 in 10 mM sodium carbonate buffer. The reported method, to our knowledge, describes the first application of SPME with element- specific detection for analysis of phosphoric acid esters. Application of the method to the plasma samples, previously stored to poly(vinyl chloride) plasma bags revealed the presence of triphenyl phosphate, which was further confirmed by SPME GC time-of-flight high-resolution mass spectrometry.
72
4.2 INTRODUCTION
Phosphoric acid triesters are used as flame retardants and plasticizers in variety of products
ranging from electronic equipment, building-materials, lubricants, glues, poly(vinyl chloride)
plastics and polyurethane foams.1, 2 However, some of the alkyl phosphates like tris(2- chlororethyl) phosphate show neurotoxic and carcinogenic properties.3 Similarly, aryl
phosphates such as triphenyl phosphate and 2-ethyl-hexyl diphenyl phosphate show allergenic4 and haemolytic effects5 in addition to other undesired side-effects. As a consequence of these
biological effects, phosphoric acid triesters have stimulated the development of many analytical
methods in environmental and biological samples.5-8 Analysis of these species in human blood plasma is gaining increasing attention due to their possible leaching from the plastic plasma collection bags. 9, 10 Presence of triphenyl phosphate and 2-ethylhexyl-diphenyl phosphate in
blood donor plasma collected from whole bag collection systems has been reported with the
levels of 2-ethylhexyl-diphenyl phosphate (up to 1.3 µg g-1) being of the same magnitude as its
-1 9 haemolytic EC20 (effective concentration at which 20% of population is affected) (1.1 µg g ).
Similarly, the reported levels of triphenyl phosphate (up to 0.1 µg g-1) are of enormous
significance for the medical treatment of patients who have a large portion of their plasma
exchanged for this type of donor plasma.9 The authors employed a relatively time-consuming
microporous membrane liquid-liquid extraction combined with GC-MS for their analysis. The
method required large amount of blood plasma sample (5.0 g) for the extraction. In order to
minimize the sample volume, miniaturized dynamic liquid-liquid extraction involving a hollow
fiber based XT-tube extractor was also employed by the same authors.10 Some studies have also
explored the possibility of on-line restricted access material/LC/tandem mass spectrometry to
eliminate the cumbersome and time consuming liquid-liquid extraction for faster analysis.11
73
However, some matrix effects were observed in case of APCI ionization with MS detection.
In this study solid-phase microextraction (SPME) is utilized as a sample preparation step for extraction and pre-concentration of phosphate esters from the human plasma samples previously stored in conventional poly(vinyl chloride) plasma bags. Advantage of SPME over previously used methods is that it is a simple solvent-free technique involving extraction, pre-concentration and sample introduction into a single step.12 To date there are only a few applications of SPME for the analysis of biological samples.13 Various parameters such as fiber coating, ionic strength, matrix pH, extraction time and temperature are optimized for reproducible recovery of analytes from the plasma samples. GC coupled to collision cell equipped inductively coupled plasma mass spectrometry (ICP-MS) for phosphorus specific detection of the compounds is utilized for separation and quantification of the compounds. ICP-MS offers lower detection limits as compared to electron impact ionization MS with phosphorous-selective detection, which is an attractive alternative to the existing methods. Phosphorus detection through standard ICP-MS suffers from low ionization efficiency and isobaric interferences such as 15N16O+ and
14N16O1H+14. However, with proper optimization of ICP parameters and the use of collision cell, which eliminates polyatomic interferences, sensitive and interference-free response can be obtained, which eventually leads to lower detection limits for trace analysis of phosphorus containing species in complex matrices. Recent applications have utilized this approach for detection of pesticides and herbicides in environmental samples.14-16 The optimized method was then applied to the analysis of human blood plasma stored in plasma bags. The presence of phosphoric acid triesters in human plasma is further validated by SPME GC time-of-flight high resolution mass spectrometry (GC/TOF-MS).
74
4.3 EXPERIMENTAL
4.3.1 Chemical and Standards
Doubly deionized water (18 MΩ cm) was used throughout, prepared using a NanoPure
treatment system (Barnstead; Boston, MA, USA). All reagents used were analytical grade
reagents and used without any further purification. Tripropyl phosphate, tributyl phosphate,
tris(2-chloroethyl) phosphate and triphenyl phosphate were purchased from Aldrich (Milwaukee,
WI, USA). Stock solutions containing 1 mg P mL-1 of each compound were prepared in
acetonitrile and used for not more than 3 weeks. Working solutions were prepared from the
appropriate dilutions of the stock solution with acetonitrile when required. Acetone, sodium acetate, sodium bicarbonate, sodium chloride and sodium borate were obtained from Fisher
scientific (Fair Lawn, New Jersey, USA).
4.3.2 Instrumentation:
An Agilent Technologies (Agilent Technologies; Palo Alto, CA, USA) model HP 6890 series
GC was used for the separation of the species. The GC was interfaced to an Agilent 7500cs ICP-
MS (Agilent Technologies; Tokyo, Japan) through a heated GC/ICP-MS interface (Agilent
Technologies; Tokyo, Japan). Separation of phosphoric acid esters was performed on a DB-5 capillary column, 30 m × 0.320 mm i.d. × 0.25 µm (J&W Scientific; Folsom, CA, USA) in the case of ICP-MS detection. A Micromass GCT orthogonal time-of-flight mass spectrometer
(Micromass; Manchester, U.K.) coupled to the Agilent 6890N GC was used for mass spectral characterization of the phosphoric acid esters found in the plasma samples stored in poly(vinyl chloride) bags. Heptacosa(fluorotributylamine) was used as a reference compound and as the lock mass compound (using the 218.9856 Da ion) for mass calibration. Average mass accuracy usually was no greater than 1 mDa. The instrument was recalibrated at any time the mass
75
accuracy was determined to exceed this limit. All instrumental conditions are indicated in Table
4.1.
76
Table 4.1 Instrumental conditions
ICP-MS parameters Forward power 1150 W Plasma gas flow rate 15.0 L Ar min-1 Auxiliary gas flow rate 0.87 L Ar min-1 Carrier gas flow rate 1.0 L Ar min-1 Sampling depth 6 mm
Optional gas Ar:N2 (50:50), ~ 4% relative to carrier gas Isotopes monitored 31P (100 ms dwell time) Collision cell gas 1.2 mL He min-1 GC parameters Capillary column DB-5 (5% phenyl-95% methyl-polysiloxane) GC carrier gas 4 mL He min-1 (constant flow mode) Oven temperature 70 oC initial ramped at 30 oC/min to 310 oC and held for 2 min Injection mode Pulsed splitless at 414 kPa (0.5 min) Injection port temperature 270 oC ICP transfer line temperature 310 oC GC-TOF-MS parameters Capillary column DB-5 GC carrier gas 2 mL He min-1 (constant flow mode) Oven temperature 80 oC initial ramped at 30 oC/min to 200 oC then ramped at 35 oC/min to 310 oC Injection port temperature 270 oC MS transfer line temperature 310 oC Ionization Electron impact: 70 eV at 180 oC + Calibrant and lock mass Perfluorotributylamine. 218.9856 Da (C4F9 ) ion was used as a lock mass.
77 4.3.3 SPME fibers:
Following fibers with different stationary phases and film thickness were obtained from
Supelco: poly(dimethylsiloxane), 100 µm (PDMS); poly(dimethylsiloxane), 6 µm; poly(dimethylsiloxane/ divnylbenzene), 65 µm (PDMS-DVB); polyacrylate, 85 µm; carboxen/poly(dimethylsiloxane), 75 µm (cPDMS). Fibers were initially conditioned according to manufacturer specifications before use to remove contaminants and to avoid passive extraction of analytes from ambient. SPME holder for manual sampling was purchased from Supelco
(Bellefonte, PA, USA). The 65 µm PDMS-DVB fiber was selected and used throughout the study. The fiber was cleaned every day by putting it into the injection port during a whole run of the GC system.
4.3.4 Sample Preparation:
Human plasma samples employed for the study were obtained from Hoxworth Blood Center
(Cincinnati, OH) with appropriate permission. Samples were stored at 4 oC prior to analysis. 50
µL of methanol was added to 500 µL aliquot of plasma sample. The obtained mixture was agitated thoroughly and the precipitation of plasma protein was then performed by adding 600
µL of acetone. Subsequently, samples were centrifuged at 5000 rpm for 10 min, the supernatant was placed into 4.0 mL vials and 0.70 g NaCl was added to each sample. The final mixture was diluted to the 3.5 mL with 0.01 M sodium carbonate buffer (pH = 7.0). The vials were sealed with PTFE faced septum cap and SPME was performed by placing the PDMS/DVB fiber into the solution, which was stirred with magnetic stirrer at constant speed. After the extraction time of 30 min the fiber was inserted directly into the GC injector for the analysis. After each analysis, the fiber was cleaned with immersing it in distilled water for 1 min and reconditioning in the GC injector for 2 min.
In order to establish optimal conditions for SPME extractions, plasma samples spiked with
10 ng P mL-1 standard were extracted after protein precipitation by PDMS/DVB fiber at different pH (3.0, 5.0, 7.0, 9.0), by varying the ionic strength (10, 15, 20 and 30% NaCl), temperature (30,
40, 50, 60 oC), extraction (15, 30, 45 min) and desorption times (0, 1, 2 min).
4.4 RESULTS AND DISCUSSION
4.4.1 Injection and gas Chromatography conditions
At first, splitless injection was tried for the injection of the analytes into the gas chromatographic column. In splitless injection mode, the entire injected amount is forced rapidly onto the capillary column. This improves the detection limits. Large peak tailing was observed for triphenyl phosphate; however, switching from splitless to pulsed splitless injection dramatically improved the peak shape of triphenyl phosphate, hence pulsed splitless injection was employed throughout this study.
In the present study, various chromatographic variables were studied in order to improve the retention time and peak shape of the phosphoric acid triesters using the standard DB-5 capillary column. Analyte elution was tested at several He carrier gas flow rates. Better peak shapes with the least retention time was observed at 4.0 mL He min-1 (Figure 4.1). In addition to flow rate another variable that was studied for achieving faster separation and better peak shape was a higher oven temperature program. High ramping temperatures are also known to decrease the retention times. The combination of high flow rate and oven program resulted in elution of phosphoric acid esters in less than 8 min.
79
Figure 4.1: Effect of helium flow rate through the GC column on the retention time and peak shapes of (1) tripropyl phosphate, (2) tributyl phosphate, (3) tris(2-chloro ethyl) phosphate, (4) triphenyl phosphate, each at concentration of 1ng P µL-1 (2 µL injection). Other conditions as stated in Table 4.1.
4.4.2 ICP-MS conditions
Historically, phosphorus has been recognized as one of the elements difficult to analyze in argon plasma because of the large polyatomic interferences and its low ionization potential.14
With the collision cell, optimization of ICP-MS parameters can easily minimize or eliminate the polyatomic interferences leading to detection limits coincident with trace analysis of P- containing molecules. So in this study, various ICP-MS conditions, such as forward power, carrier gas flow rate and collision cell gas flow rate were optimized. It has been demonstrated previously that addition of N2 as an alternate gas to the central argon plasma channel leads to
14 enhancement of phosphorus signal. The fraction of N2 that proved to be most advantageous in
80
mixed-gas plasma was found to be approximately 4% relative to Ar carrier gas. For that purpose,
50:50 mixture of Ar: N2 was used as optional gas throughout the experiments.
Forward Power and Argon Carrier gas flow rate. Forward Rf power and argon carrier gas flow rate are two important parameters that have considerable effect on the ionization of phosphorus in the GC/ICP-MS system. Optimization of these conditions was performed by injecting the same concentration of tripropyl phosphate at different settings of these parameters. Result of application of different Rf power with ~ 4% Ar: N2 as optional gas reveal that relatively low power plasma (1100-1150 W) results in maximum sensitivity and hence, maximum signal to noise ratio. While for argon carrier gas flow rate, the maximum peak area for triphenyl phosphate was obtained at 1.0-1.1 L min-1 (Figure 4.2). A flow rate of 1.0 L min-1 was selected throughout the study since the background signal increases with the carrier gas flow rate.
Figure 4.2: Effect of argon carrier gas flow rate on ionization of tripropyl phosphate (10 ng P mL -1 (2 µL injection). Other conditions as stated in Table 1.
Optimization of ICP collision cell gas flow rate. As stated earlier, determination of monoisotopic phosphorus suffers from polyatomic interferences such as 15N16O+ and 14N16O1H+ at m/z 31.14
This poses difficulties in phosphorus determination with a standard ICP-MS equipped with low-
81
resolution quadrupole mass analyzer. In the present study, nitrogen is added as a plasma component to enhance the ionization of phosphorus in argon based plasma. However, this is also accompanied by increase in background signal at m/z 31 due to formation of nitrogen based polyatomic interferences. Removal of these interferences becomes necessary for trace analysis of phosphorus in plasma samples. For that purpose, octopole collision cell is employed, which is placed before the quadrupole mass analyzer. Collision cells helps to remove the polyatomic interferences through combination of collisional interactions and energy discrimination.17,18
Pressurizing the collision cell with He gas leads to its collision with polyatomic ions and the element ions. Since collisions with the larger cross-section polyatomic ions occur more frequently than those with the element ions, this results in the selective decrease of polyatomic ion energy. Finally the application of energy difference potential between the octopole (-18 V) and quadrupole (-17 V) stops the lower energy polyatomic ions from entering the quadrupole while allowing the passage of the element ions under investigation.16 Helium was used as collision cell gas at 1.2 mL min-1 throughout the experiments in order to minimize interferences at m/z = 31.
4.4.3 Solid-phase micro-extraction (SPME)
Selection of SPME coating. The first step towards the SPME optimization was choice of appropriate fiber. Along with PDMS (both 7 and 100 µm), which is recommended by Supelco for analysis of non-polar species, PDMS/DVB, polyacrylate and cPDMS SPME fibers were tested for the extraction efficiency of phosphoric acid esters from the spiked plasma samples. For this purpose, 500 µL aliquots of plasma was placed in a glass vial, diluted to 3.5 mL with deionized water and the resulting solution was then spiked with phosphate esters standard mixture (each compound at 500 ng P mL-1). After the extraction time of 30 min, the
82
fiber was exposed directly to the hot GC injector for the analysis. Previously 100 µm PDMS has also been used for air sampling of gaseous organophosphate triesters.19, 20 As depicted in Table
4.2, extraction efficiency of PDMS/DVB fiber is found to be considerably higher than other fiber coatings for the analytes of interest in this study. Higher extraction of tris(2-chloroethyl) phosphate can be explained due to bipolar nature of PDMS/DVB coating whereas π-π interactions between the aromatic cycle of triphenyl phosphate and divinylbenzene coating may lead to additional analyte/fiber interaction 20.
Table 4.2. Performance of various fiber coatings relative to PDMS/DVB coating for extraction of phosphoric acid ester species by SPME a
Extraction efficiency (%) relative to PDMS/DVB Compound PDMS (100 µm) cPDMS Polyacrylate PDMS (7µm) Tripropyl phosphate 31.5 65.0 2.05 1.62 Tributyl phosphate 61.5 27.5 2.37 0.50 Tris (2-chloro ethyl) 11.6 12.8 13.0 3.90 phosphate Triphenyl phosphate 16.5 14.2 16.6 4.71 aResults obtained by diluting 500 µl of plasma to 3.5 ml and spiking the resulting solution with phosphoric acid esters standard mixture at conc. 500 ng P ml-1 each. Extraction time was 30 min.
4.4.4 Optimization of the extraction procedure
Influence of plasma matrix and deprotenization. In order to study the effect of plasma matrix on the low extraction yield of phosphoric acid esters, 3 kDa molecular weight cut-off filter (Millipore) was used. Spiking the plasma prior to filtration showed no recovery of triphenyl phosphate from the filtrate. This shows that considerable binding of the analytes with the plasma matrix takes place that eventually prevents the analytes from passing through the membrane as a small molecular weight compounds.
In order to reduce the effect of plasma matrix on the extraction yield, their removal before the SPME extraction was necessary. Since most of the plasma matrix constitutes proteins, deprotenization of plasma was performed. Prior to this, methanol was added which is known to
83
assist in releasing the analytes bound to the plasma proteins.21 Precipitation of proteins was performed using acetone rather than conventionally used protein precipitating agents (such as trichloroacetic acid) due to the hydrophobic nature of higher organophosphate esters. This procedure increased the extraction yield of tributyl and triphenyl phosphate from 11% and 5% to
43 and 66%, respectively, at the same plasma dilution so remaining conditions for SPME extraction were optimized after the deproteinization and dilution of the resulting supernatant. On the other hand, extraction yield of tripropyl and tris(2-chloroethyl) phosphate did not increase significantly due to their low solubility in hydrophobic solvents.
Effect of salt. In order to investigate the effect of salt on the extraction efficiency, various amounts of NaCl (10, 15, 20, and 30%) were added to the spiked plasma samples after protein precipitation and dilution as described above in the sample preparation section. As shown in
Figure 4.3a, the addition of NaCl has significant effect on the extraction efficiency of phosphoric acid esters. By increasing the ionic strength of solution with 30% NaCl, the amount of extracted tripropyl, tributyl, and tris(2-chloroethyl) phosphates also increased; however, for the triphenyl phosphate opposite effect was observed. Thus, 20% NaCl was selected as the
appropriate condition throughout the study.
84
Figure 4.3 (a): Effect of ionic strength of the sample on the extraction of the phosphoric acid esters at 10 ng P ml-1 .
Effect of pH. The pH of plasma matrix can also be altered to effect the distribution of analyte between the sample and SPME fiber. In case of acid and basic analytes, proper adjustment of pH can increase the sensitivity of the method by converting the ionic form of the analyte into neutral
(thus more volatile) form for efficient extraction on neutral SPME coating.22 In case of phosphoric acid esters, the variation of pH over a range of 3.0 to 9.0 did not significantly affect the extraction yield by the fiber (Figure 4.3b). Overall, these findings are consistent with the fact that phosphate esters are neutral molecules with no acid/base characteristics. pH = 7.0 was selected for the sample dilution after protein precipitation throughout the study, since extreme pH values can damage the fiber coating and reduce its lifetime.
85
Figure 4.3 (b): Effect of pH of the sample on the extraction of the phosphoric acid esters at 10 ng P ml-1 .
Effect of extraction temperature. The effect of sample temperature in the extraction yield was investigated varying the temperature between 25 and 60 oC with a constant extraction time of 45 min. The extraction and temperature profiles obtained for the phosphoric acid triesters are depicted in Figure 4.3c. Increase in temperature from 25 to 40 oC results in slight recovery enhancement which then remains constant up to 50 oC. However, a further rise in temperature to
60 oC leads to decrease in the extracted amount of the analytes. This can be explained by the fact that extraction is the exothermic process and therefore, high temperature favors the partition of analytes between the solution and fiber in the opposite direction (e.g. into the solution).23 An extraction temperature of 40 oC was selected for the study.
86
Figure 4.3 (c): Effect of temperature of the sample on the extraction of the phosphoric acid esters at 10 ng P ml-1 . Effect of extraction and desorption time. Analyte extraction time is perhaps the most important parameter in the SPME method. It is known from the previous studies (performed with air samples) that equilibrium sampling of organophosphates with SPME can take up to several hours.19, 20 In order to keep a short analysis time of the entire method, extraction times were investigated only up to 45 min. Time dependent adsorption behavior of phosphoric acid esters on the PDMS/DVB fiber is shown in Figure 4.3d. It is clear that the equilibrium conditions are not reached during this period of time which is in agreement with previous studies.19, 20 For practical reasons, an extraction time of 30 min was selected. For all comparisons and quantification, the
extraction conditions and times were held constant.
87
Figure 4.3 (d): Effect of sampling time on the extraction of the phosphoric acid esters at 10 ng P ml-1 . GC desorption of the analytes from the PDMS/DVB fiber was found to be instantaneous
(data not shown). However, the fiber was generally held for complete chromatographic run time in the injection port in order to avoid memory effects.
4.5 ANALYTICAL PERFORMANCE CHARACTERISTICS
To check the performance of the developed method, SPME analysis of spiked plasma samples containing known amounts of phosphoric acid ester standards was performed. The assay was linear (with r2 > 0.993) over concentration range of 0.1 to 50 ng P mL-1for each phosphoric ester studied. The limits of detection were calculated using the chromatographic sideband noise 14, 24,
25:
(1)
Here ci is the concentration of the analyte, sb is the average standard deviation of background near the peak of the analyte (~ 200 data points), Ai is the peak area of the analyte corrected for
88
the background signal, and n is the number of data points in the analyte peak (usually n = 20–
100). The results are shown in Table 4.3. The detection limits were 50 ng L-1 for tripropyl phosphate, 17 ng L-1 for tributyl phosphate, 240 ng L-1 for tris(2-chloroethyl) phosphate and 24 ng L-1 for triphenyl phosphate. Recovery of triphenyl phosphate increased from 5% to 66% after deproteinization of plasma samples while that for tripropyl, tributyl and tris (2-chloroethyl) phosphates was in the range of 35, 43 and 49%, respectively, after sample deproteinization at 10 ng mL-1 of spiked concentration. Note that such a low analyte recovery is commonly encountered in drug determination from plasma due to considerable binding with the plasma proteins. For example, optimum recovery of lidocaine from human plasma after protein precipitation in a recent study was only 12%.26 When direct immersion SPME is performed after 50-fold sample dilution of the plasma samples (without protein precipitation) recoveries of tripropyl-, tributyl- and tris(2-chloroethyl) phosphate were 92, 60 and 91% respectively. Although such an approach results in high analyte recovery, it is of little use, since the increase in recovery is compromised with large sample dilution. On the other hand higher dilution of the plasma could be performed for the shorter extraction time in order to improve recoveries if the analyte concentration is high enough to be detected by the method of analysis.
The precision of the method was obtained by analysis of ten replicate spiked plasma samples consecutively at 1 ng P mL-1. The relative repeatability was below 15% for all the analytes. Validation of the method could not be performed due to lack of commercially available certified reference material for determination of analytes in the plasma matrix.
89
Table 4.3: Analytical Performance Characteristics of the phosphoric acid triester detection in human plasma.
Phosphoric acid triesters Limit of Retention r2 Method Recovery % detection, time, min precision, ng/l %RSD Tripropyl phosphate 50 3.36±0.03 0.998 8 35 Tributyl phosphate 17 4.48±0.01 0.999 11 43 Tris(2-chloro ethyl) phosphate 240 4.98±0.01 0.993 7 49 Triphenyl phosphate 24 7.21±0.04 0.995 14 66
4.6 APPLICATION TO HUMAN PLASMA SAMPLES
In this study, human plasma collected from a plasma bag was analyzed for the organophosphate esters using the developed method. Presence of tributyl phosphate and triphenyl phosphate was detected in the plasma that was exposed to the poly(vinyl chloride) plasma collection bag for a two week period (Figure 4.4b) while these compounds were absent in the same plasma that had not been stored in the conventional plastic storage bags (Figure
4.4a). Levels of triphenyl phosphate were in the range of 0.2 ng P mL-1 while that of tributyl phosphate was close to the detection limit of the method (0.02 ng P mL-1). Both 31P GC/ICP-MS and GC/TOF-MS chromatograms for the non-spiked human plasma are presented in Figure 4.4.
Presence of triphenyl phosphate in natural plasma stored in poly(vinyl chloride) bags was confirmed with high resolution TOF-MS measurements. The identity of the triphenyl phosphate was verified by retention time matching, correct isotope pattern (M.+ and [M-H]+) and accurate mass measurements (within 1 mDa accuracy) as seen in Figure 4.4. However, presence of tributyl phosphate could not be confirmed through GC/TOF-MS due to its very trace levels in the plasma samples.
90
Figure 4.4: Analysis of phosphoric acid triesters in human plasma with GC/ICP-MS (left) and GC/TOF-MS showing extracted ion chromatogram for triphenyl phosphate and M.+ and [M-H]+ions for sample and standard (right). (a) Native human plasma, (b) human plasma that has been stored in a poly(vinyl chloride) bag, and (c) human plasma spiked with 1 ng P mL-1 (1 ppb) of phosphoric acid triesters. SPME extraction was performed after sample deproteinization, and addition of 0.70 g NaCl at pH = 7.0. Extraction was carried out with a 65 µm PDMS-DVB fiber for 30 min at 40 oC.
4.7 CONCLUSIONS
Application of SPME in conjunction with GC/ICP-MS proved to be a very promising analytical method for determination of trace amounts of phosphoric acid esters in complex biological samples such as human plasma. The developed method is relatively simple, sensitive, reasonably fast and solvent free as compared to previously described methods of detection. To our knowledge this is the first application of SPME with element-specific detection for analysis of phosphoric acid esters.
Low detection limits obtained in the parts-per-trillion range also assisted in determination of triphenyl phosphate in human plasma previously stored in conventional plasma storage bags.
Levels of this were found to be 3 orders of magnitude lower than its haemolytic EC20 value.
91
Combination of GC/ICP-MS with GC/TOF-MS helped to confirm the presence of this species in plasma collected from the bag. The presence of this is attributed to the fact that triphenyl phosphate is applied as non-flammable plasticizer in poly(vinyl chloride) bags. Despite previous reports of large levels of ethylhexyl diphenyl phosphate in plasma stored in poly(vinyl chloride) bags, no evidence of this compound was found in our study.
4.8 REFERENCES
(1) Carlsson, H.; Nilsson, U.; Becker, G.; Ostman, C. Environmental Science & Technology 1997, 31, 2931-2936. (2) Nagase, M.; Toba, M.; Kondo, H.; Yasuhara, A.; Hasebe, K. Analytical Sciences 2003, 19, 1617-1620. (3) Matthews, H. B.; Eustis, S. L.; Haseman, J. Fundamental of Applied toxicology 1993, 20, 477-785. (4) Camarasa, J. G.; Serra-Baldrich, E. Contact dermatitis 1992, 26, 264-265. (5) Jonsson, O. B.; Dyremark, E.; Nilsson, U. L. Journal of Chromatography B 2001, 755, 157-164. (6) Fries, E.; Puttmann, W. Journal of Environmental Monitoring 2001, 3, 621-626. (7) Toda, H.; Sako, K.; Yagome, Y.; Nakamura, T. Analytica Chimica Acta 2004, 519, 213- 218. (8) Rodil, R.; Quintana, J. B.; Reemtsma, T. Analytical Chemistry 2005, 77, 3083-3089. (9) Jonsson, O. B.; Nilsson, U. L. Journal of Separation Science 2003, 26, 886-892. (10) Jonsson, O. B.; Nilsson, U. L. Analytical and Bioanalytical Chemistry 2003, 377, 182- 188. (11) Amini, N.; Crescenzi, C. Journal of Chromatography B 2003, 795, 245-256. (12) Lambropoulou, D. A.; A, T. A. Journal of Chromatography A 2001, 922, 243-255. (13) Queiroz, C. M. E.; Lancas, M. F. LCGC North America 2004, 22, 970-980. (14) Vonderheide, A. P.; Meija, J.; Montes-Bayon, M.; Caruso, J. A. Journal of Analytical Atomic Spectrometry 2003, 18, 1097-1102. (15) Sadi, B. B. M.; Vonderheide, A. P.; Caruso, J. A. Journal of Chromatography A 2004, 1050, 95-101. (16) Profrock, D.; Leonhard, P.; Wilbur, S.; Prange, A. Journal of Analytical Atomic Spectrometry 2004, 19, 623-631. (17) Tanner, S. D.; Baranov, V. I.; Bandura, D. R. Spectrochimica Acta Part B-Atomic Spectroscopy 2002, 57, 1361-1452. (18) Bandura, D. R.; Baranov, V. I.; Tanner, S. D. Fresenius Journal of Analytical Chemistry 2001, 370, 454-470. (19) Isetun, S.; Nilsson, U.; Colmsjo, A.; Johansson, R. Analytical and Bioanalytical Chemistry 2004, 378, 1847-1853. (20) Isetun, S.; Nilsson, U.; Colmsjo, A. Analytical and Bioanalytical Chemistry 2004, 380, 319-324.
92
(21) Krogh, M.; Grefslie, H.; Rasmussen, K. Journal of Chromatography B: Biomedical Applications 1997, 689, 357-364. (22) Wuilloud, R. G.; de Wuilloud, J. C. A.; Vonderheide, A. P.; Caruso, J. A. Journal of Analytical Atomic Spectrometry 2003, 18, 1119-1124. (23) James, K.; Stack, M. Journal of High Resolution Chromatography 1996, 19, 515. (24) Meija, J.; Montes-Bayón, M.; Duc, D.; Terry, N.; Caruso, J. A. Analytical Chemistry 2002, 74, 5837-5844. (25) Brushwyler, K. R.; Furuta, N.; Hieftje, G. M. Talanta 1990, 37, 23-32. (26) Koster, E. H. M.; Wemes, C.; Morsink, J. B.; Jong, G. J. D. Journal of Chromatography B 2000, 739, 175-182.
93
CHAPTER 5 | STUDYING MINOR SELENIUM VOLATILES FROM SELENIUM ENRICHED GREEN ONIONS
5.1 ABSTRACT
In this study the presence of minor Se-containing volatiles were investigated in Se-enriched green onions (Allium fistulosum). The Combination of high-resolution mass spectrometry and inductively-coupled plasma mass spectrometry confirmed the structures of volatiles reported previously, along with several unreported small molecular weight Se-containing volatiles such as
MeSeSeSMe. A simple mass defect-based algorithm is presented to convert electron impact mass spectra into the ICP-like (element-specific) mass spectra. This data analysis technique was useful to link the results obtained from molecular and elemental mass spectrometry thus aiding in the search for new trace level Se-containing volatiles.
95 5.2 INTRODUCTION
Ever since the Clark trial1, there has been increasing interest in understanding the selenium biochemistry within living organisms due to the suggested Se anti-carcinogenic and antioxidant properties. Both volatile and non-volatile forms of selenium have been explored within many biologically important systems in order to obtain as complete information as possible on the occurrence of selenium in its various forms. Plants from genus Allium represent one of the most widely studied groups as they are known to accumulate appreciable amounts of sulfur and selenium. Selenium occurs predominantly in Allium. plants in form of Se-methylselenocysteine
(Allium tricoccum)2 and γ-glutamyl-Se-methylselenocysteine (A. sativum and A. fistulosum)3, 4 and at intermediate levels as Se-methionine (A. cepa)3. These compounds are widely recognized for their selenium-related medicinal properties from epidemiological studies1. On the other hand, the information on Se-containing volatile constituents of Allium plants is restricted to only a few studies.5, 6 The volatile components of the Allium genus are released from their non-volatile precursors, such as Se-methylselenocysteine-Se-oxide, by an enzyme mediated degradation which takes place when the plants are crushed. While dimethylselenide is the major volatile in
Se-accumulating plants and dimethyldiselenide is the major Se-volatile in Se-hyperaccululating plants, due to their high sulfur content Allium plants contain considerable amounts of selenosulfenates as their major Se-containing volatiles. Alk(en)yl groups are mainly a combination of propyl-, 1-propenyl-, allyl-, and methyl- groups, depending on the plant species.
5, 6 (RSnSeR’: n = 0–2; R and R’ = methyl-, allyl-, 1-propenyl-).
Previously headspace gas chromatography with atomic emission detection (HS/GC-AES) has been employed to detect and identify various selenium volatiles released from cut Allium spp.6
However, the unavailability of selenium-containing volatile standards has left many heavier
selenium volatile compounds either unidentified or unconfirmed. The aim of present study is to explore the use of modern mass spectrometric techniques to improve the knowledge of minor selenium metabolites present in the Allium sp. Selenium enriched green onions (Allium fistulosum) were selected in this study.
Recent developments in combined use of inductively coupled plasma mass spectrometry
(ICP-MS) (for element specific detection within a molecule) and molecular mass spectrometry for corresponding structural characterization (a metallomics approach) have helped to confirm trace amounts of previously unidentified species in biological matrices.7 ICP-MS offers very high sensitivity, multielemental detection capability and low detection limits among other element selective detectors when coupled to a GC system. Previously the coupling of GC to ICP-
MS has been performed for detection of selenium and sulfur-containing plant volatiles.8, 9 Solid phase microextraction (SPME) was used for selective extraction and pre-concentration of the volatiles and for analysis, direct injection into the GC/ICP-MS system and GC/TOF-MS.
Extraction using SPME ensures preconcentration up to several orders of magnitude as compared to other time consuming and tedious sample preparation methods. ICP-MS, on the other hand, is also known for its high sensitivity and superior (lower) detection capabilities. So combined use of two highly sensitive techniques at the sample preparation and detection stages leads to detection of ultratrace species in complex biological samples. This approach is applied in this study for the determination of trace selenium containing volatiles present in the headspace.
Complimentary to this, gas chromatography-time of flight mass spectrometry (GC/TOF-MS) is used as an additional technique for characterization and confirmation of novel selenium- containing volatiles, since the ability to do exact mass measurements is highly beneficial for studying the mass spectral fragmentation pathways of the selenium compounds.10 A simple mass
97
defect-based algorithm is also presented in order to simplify electron impact high resolution mass spectra, thereby appearing similar to ICP-like (element-specific) mass spectra. ‘This was developed by Dr. Juris Meija and is further described elsewhere. The scheme is included here for completeness of the data.” This data analysis technique was highly useful to relate the conclusions obtained from molecular and elemental mass spectrometry.
5.3 EXPERIMENTAL
5.3.1 Instrumentation
Gas chromatography: An Agilent Technologies (Agilent Technologies; Palo Alto CA, USA) model HP 6890 series GC was used for the separation of the species. Separation of selenium volatile compounds was performed on a DB-5 capillary column, 30 m × 0.320 mm i.d. × 0.25
µm (J&W Scientific; Folsom CA, USA) in the case of ICP-MS detection. Separation conditions are depicted in Table 5.1.
ICP-MS: An Agilent 7500ce ICP-MS (Agilent Technologies; Tokyo, Japan) was used for the element specific detection. The instrument is equipped with octopole collision/reaction cell and can be operated in both normal and collision cell mode. Hydrogen was used as a collision cell gas for interference free detection of selenium at m/z 80 and 78.
98
Table 5.1: Instrumental conditions
ICP-MS parameters Forward power 1150 W Plasma gas flow rate 15.0 L Ar min-1 Auxiliary gas flow rate 0.87 L Ar min-1 Carrier gas flow rate 1.0 L Ar min-1 Sampling depth 6 mm
Optional gas Ar:N2 (50:50), ~ 4% relative to carrier gas Isotopes monitored 33S, 34S, 77Se, 78Se, 80Se, and 82Se (100 ms dwell time) -1 Collision cell gas 3.2 mL H2 min GC parameters Capillary column DB-5 (5% phenyl-95% methyl-polysiloxane) GC carrier gas 2.4 mL He min-1 (constant flow mode) Oven temperature 40 °C for 4 min 40-125 °C @ 15 °C min-1 Hold 5 min 125-300 °C @ 35 °C min-1 Hold 1 min Injection mode Pulsed splitless at 10 psi (0.5 min) Injection port temperature 250 oC ICP transfer line temperature 250 oC GC-TOF-MS parameters Capillary column DB-5 GC carrier gas 2 mL He min-1 (constant flow mode) Oven temperature 50-220 °C @ 10 °C min-1 Hold 7.5 min Injection port temperature 250 oC MS transfer line temperature 250 oC Ionization Electron impact: 70 eV at 180 oC + Calibrant and lock mass Perfluorotributylamine. 218.9856 Da (C4F9 ) ion was used as a lock mass. EI + Temp 180 °C Trap current 400 µA m/z monitored 40-620 Scan duration 0.9 s Interscan delay 0.1 s GC re-entrant 250 °C Reference reservoir 80 °C
99
Introduction of collision cell gas not only leads to removal of isobaric interferences, but also enables the detection of trace amount of selenium volatile species released due to low background noise at these masses. Sulfur isotope at m/z 34 was also monitored along with selenium for multielemental analysis. The GC was interfaced to ICP-MS through a heated GC/ICP-MS interface (Agilent Technologies; Tokyo, Japan).
GC/TOF-MS: A Micromass GCT orthogonal time-of-flight mass spectrometer (Micromass;
Manchester, U.K.) coupled to the Agilent 6890N GC was used for mass spectral characterization of the selenium volatiles released from the enriched green onions. Heptacosa
(fluorotributylamine) was used as a reference compound and as the lock mass compound (using the m/z 218.9856 ion) for mass calibration. Average mass accuracy usually was less than 0.001
Da. The instrument was recalibrated at any time the mass accuracy was determined to exceed this limit.
5.3.2 Reagents and standards
All reagents were analytical grade reagents and were used without purification. All solutions were prepared in 18 MΩ cm doubly deionized water generated by a NanoPure treatment system
(Barnstead; Boston MA, USA). The following reagents were purchased from Fluka/Sigma-
Aldrich (Milwaukee WI, USA): Dimethyl selenide, dimethyl sulfide, and dimethyl disulfide, dimethyl diselenide, diethyl disulfide, dimethyl trisulfide and heptacosafluorotribuytlamine.
Diethyl diselenide was purchased from Strem Chemicals (Newburyport MA, USA) Individual stock solutions of the standards were prepared by dissolution of chemicals in GC/MS grade hexane (Fisher Scientific; Fair Lawn NJ, USA). Standards were diluted in hexane as required.
1000 μg mL-1 stock solutions of selenite was prepared by dissolving appropriate amount of
100
Na2SeO3 (ICN Biomedicals; Aurora OH, USA) in doubly deionized distilled water. The stock solution was appropriately diluted for supplementation purposes.
5.3.3 Preparation of selenium-enriched green onions
The plant compartment selected for the characterization of volatile selenium in the present work was the leaves of the evergreen long white bunching onion, Allium fistulosum. The seeds were purchased from Burpee (Warminster PA, USA). Cell packs containing three to five seed and with four cells each of dimensions (0.13 m × 0.13 m × 0.10 m) were employed. Each cell contained Promix BX (Natorp landscape; Cincinnati OH, USA), which is mixture of sphagnum, peat moss, perlite, vermiculite, dolomite and calicitic lime stone. Cells were watered daily and fertilized with commercially available 15N:30P:15K nutrient solution when needed. Plants were grown in a greenhouse with temperatures between 16-21 oC. Before any selenium treatment, the cells were thinned to 3 plants per cell. One of the cell packs was not enriched with Se and was used as a control. Two months after seeding, the cells were treated with Na2SeO3 at the concentration 20 μg g-1. To compensate for loss from run out, one week after the first treatment the cells were treated with the same concentration of selenium as before. Plants were harvested ten days after supplementation began. The green onions were washed with water and then were separated from the roots. Total selenium determination (ICP-MS) after complete digestion of sample revealed that plants accumulated Se at the level of 200 μg Se g-1 of fresh sample. It should be noted that no attempt was made to establish sterile conditions.
5.3.4 Sample Preparation
Enriched green onion plant was thoroughly washed and dried in air. The plant sample was crushed using pestle and mortar and quickly placed into the sample vial. The pestle and mortar
101
and the sample vial was purged with N2 prior to use to prevent the oxidation of selenium volatiles. The vial was closed with the Teflon septum and the SPME fiber was exposed to the vial headspace for extraction of the volatiles.
Commercially available SPME fiber was used for extraction and pre-concentration of volatile species. The fiber used was Carboxen/PDMS (film thickness 75 µm) obtained from the
Supelco (Bellefonte PA, USA). The SPME fiber was exposed to the headspace of the sample.
After the extraction period of 30 min, the analytes were desorbed at 250 oC in the GC injection port using the 0.75 mm i.d. inlet liner (Supelco). The fiber was held in the liner throughout the run to reduce any memory effects.
5.3.5 Synthesis of Matching Standards
Dimethylselenosulfenate (MeSeSMe) and other selenium- and sulfur-containing volatiles were prepared in solution by mixing equal volumes of methanol or pentane solutions of dimethyl trisulfide and dimethyl diselenide (1000 ppm each) in a closed vial. The resulting solution was allowed to equilibrate at room temperature for a few hours and, after dilution with pentane, was subjected to chromatographic separation.9
5.4 RESULTS AND DISCUSSION
Since the volatile selenium constituents originate from their non-volatile precursors, their identification provides indirect information on the selenium biochemistry and metabolism in the plants of interest. Many recent studies have been dedicated to identify the various selenium volatiles in biological systems. The analysis of selenium compounds in biological samples in not a trivial task since there are many techniques for the element-specific detection of Se-species, such as ICP-MS, but only few that offer molecular structure characterization of the detected compounds at the trace levels.11 The valuable information obtained from the element-specific
102
detection remains incomplete because of loss of structural information during ionization process.
The species specific characterization of selenium compounds by ICP-MS is limited to only few selenides and diselenides such as methyl and ethyl derivatives since these are commercially available whereas, in complex biological matrices a multitude of selenium containing volatiles are possible and require identification.
Unlike selenium accumulating plants, such as Brassica juncea, Allium plants volatiles are mainly allyl- group containing selenides and diselenides. Mixed selenium and sulfur containing compounds have also been detected in the allium plants.6 The lack of commercially available standards require in-house synthesized standards and their subsequent characterization with the help of molecular mass spectrometry.9 One of the main problems in trace element speciation is the unmatched detection capabilities of inductively coupled plasma mass spectrometry and molecular mass spectrometry techniques. Even though modern molecular spectrometry techniques, such as high resolution time-of-flight mass spectrometry offers comparably low detection levels just as ICP-MS, the complexity of the mass spectra sometimes prohibits direct identification of unknown species. The presence of the isotope pattern is helpful, although this is not a universal solution to the problem, since many elements (such as iodine or arsenic) are monoisotopic. Isotope pattern distortions due to the presence of isobaric interferences are also common.10
Analytical problems, such as lack of commercially available standards for direct analysis by element specific detection via retention time matching and standard additions, require often complex molecular mass spectra for trace identification of the species of interest. For simplification purposes, a mathematical approach, such as mass defect analysis as a means of digital sample-clean-up for the molecular mass spectrometry, helps in interpretation. So,
103
combined use of molecular mass spectrometry and mass defect analysis with element specific detection have been employed to provide and complete the preliminary information on selenium volatiles obtained by ICP-MS.
5.4.1 Mass defect mask analysis (developed by Dr. Juris Meija) Due to the differences in the formation energy (stability) of various elements and the mass- energy relationship (∆E = ∆mc2), exact masses of elements differ from the total mass of the
16 32 constituent protons, neutrons and electrons. For example, isobars such as O2 and S have different fractional masses (mass defect or the difference between the exact mass and the nearest integer mass). Since the mass defect of heavy atoms is larger than for lighter major elements, the exact mass of molecular interferences usually will be larger than that of single elements – a feature commonly utilized in high resolution sector-field ICP-MS. In this study we reduce the complexity of mass spectra by tracking the selenium-containing species using the mass defect of the elements.12 In this approach, a mass defect mask is applied to observed signals and as a result, signals matching certain mass defects are visualized while others are suppressed. This is a post-acquisition type of chemical noise suppression method (suppressing unwanted chemical signals) and such a visualization approach is well suited for elemental speciation of trace elements in complex biological systems because of the large differences in mass defects between trace elements of common interest (such as Se, As, Sn and I) and the major elements H, C, N, O,
S and P.13, 14 Mathematical mass defect analysis is also used in peptide and protein identification15 and in petroleum classification and characterization (Kendrick mass defect plots).12
The mass defect of molecules is a function of the molecular weight due to the contributions from other elements, such as hydrogen or oxygen. However, the mass difference between the various aliphatic selenides and hydrocarbons remains constant as shown in Figure 5.1.
104
Figure 5.1 | Differences in exact mass of aliphatic Se/S chalcogenides, hydrocarbons and poly(dimethylsiloxanes) as a basis for the mass defect mask analysis.
The mass defect of hydrocarbon CnH2n+2 (with molecular weight M) can be calculated from the formula
Δm(CnH2n+2) = a·(M + 12) (1) where a = (MH – 1)/(6 + MH). Then, for any given mass spectral signal at mass M (for charge z =
1) one calculates the mass defect difference between the alkanes, Δm(CnH2n+2), alkyl chalcogenides, Δm(X), and the signal of interest, Δmi:
|Δm(CnH2n+2) – Δm(X) – Δmi| (2)
The value for the Δm(X) is the relative mass defect (relative to the hydrocarbons) read directly from Figure 5.1, e.g. Δm(Se) = 0.17 u.
Gas chromatographic mass spectra usually contain large amounts of unwanted chemical information, which comes either from hydrocarbon or poly(siloxane) fragment ions. These ions have odd mass whereas electron impact mass spectra of Se-containing volatiles produces abundant molecular ion clusters with even mass Se isotopes covering 92% and 86% of Se1 and
Se2 isotope patterns, respectively. Mass parity thus can be effectively incorporated to remove any
105
odd-mass ions. For example, ~80% of the ion current of a typical mass spectrum of a poly(dimethylsiloxane) capillary column bleed is of odd mass (data not shown). Overall the element specific noise filtering algorithm of high-resolution mass spectra is summarized in
Scheme 5.1.
Scheme 5.1 | Algorithm for the mass defect specific filtering of high-resolution mass spectra using contextual information. In this example, volatile selenosulfenates (containing Se-S)are the targeted species.
According to this scheme every m/z channel at every scan is subjected to this logical test. The first step rejects all the ions with mass below the target substructure while the second step filters out all the odd-m/z values. The third step demands that the mass defect of the selected ion is matching the theoretically predicted mass defect of alkyl selenosulfenate at the m/z value of the selected ion as outlined above. The matching tolerance in this example is set to 0.025 u.
5.4.2 Search for trace level Se volatiles
The first approach toward the identification of minor selenium volatiles in enriched green onions was application of SPME-GC/ICP-MS. Figure 5.2 depicts the SPME-GC/ICP-MS
106
chromatogram for the headspace analysis of green onion plants (supplemented with Se(IV)) after crushing to initiate the action of lyases and allinases in the plant tissue (a rapid acceleration of sensecense). As shown, several selenium containing peaks are identified in the sample. Retention time matching with commercially available standards was performed, which revealed the presence of only a few earlier eluting commonly found selenium compounds such as dimethyl selenide (MeSeMe), dimethyl diselenide (MeSeSeMe). In addition to these principle volatiles, several other mixed selenium sulfur containing volatile species have also been detected in the headspace of green onions enriched with Se(IV) (Figure 5.2). Identification of these mixed selenium sulfur species was performed by all-in-one laboratory synthesized standard mixture obtained from the S/Se exchange reaction between the dimethyl diselenide, MeSeSeMe, and dimethyl trisulfide, MeSSSMe. The chalcogen exchange reaction between the two species leads to formation of various selenosulfates and polychalcogenides such as dimethyl selenosulfenate,
16 MeSeSMe, bis(methylthio) selenide, (MeS)2Se, and bis(methylseleno) sulfide, (MeSe)2S.
Retention time matching with the standards, also confirms the presence of MeSeSMe, and the trace levels of (MeS)2Se and (MeSe)2S in the sample. However, further characterization using molecular mass spectrometry was required to confirm the presence of these minor volatiles. The presence of bis(methylseleno) sulfide, (MeSe)2S, from the plant samples has never been reported. The two isomers of (MeS)2Se, MeSSeSMe and MeSSSeMe exhibit slight difference in retention behavior on the GC column used, hence their individual identification in the sample could also be performed using the current chromatographic method. From these two compounds,
MeSeSSMe has been very recently reported in the headspace gases above genetically modified
Escherichia coli.17 Comparison of the mass spectra of synthetic MeSeSSMe and MeSSeSMe was also undertaken to confirm the identity of these species as described below.
107
350000 1
300000 U
250000 2 U Se cps
78 200000
150000
Abundance, Abundance, 100000 S 3 2 Se U MeSeSMe 50000 2 MeSeMe MeSeSeMe (MeSe) (MeS) 0 024681012 Time, Sec
Figure 5.2 : SPME/GC-ICP-MS chromatogram of enriched green onions headspace
The peaks labeled as U1, U2 and U3 from GC/ICP-MS could not be identified retention time matching because we lacked laboratory available standards. However, multielemental detection capability of ICP-MS helped to identify the presence of both S and Se isotopes in these three peaks. This was found to be useful in their identification through GC/TOF-MS. The search for m/z = 111.888 (SeS.+ ion) in the TIC helps to extract the peaks containing the SeS.+ fragment ion (Figure 5.3).
108
Figure 3: Extracted ion chromatogram from GC/TOF-MS showing peaks containing SeS.+ as a fragment ion
GC/TOF-MS analysis: GC/TOF-MS is a very common molecular mass spectrometric technique and is highly useful when used in conjunction with ICP-MS because of their comparable detection limits. The high sensitivity of TOF-MS along with high resolution has made identification and confirmation of many species found in complex biological samples possible, especially for those which cannot be identified using element specific detection due to lack of commercially available standards. In such cases, ICP-MS helps to selectively screen for the presence of unknown selenium or sulfur compounds, while TOF-MS helps in their identification by its ability to do exact mass measurements. The benefits of exact mass measurements become more important for accurate assignment of polyselenide molecular ions because of skewed isotopic distribution patterns. In this study, the complimentary use of
GC/TOF-MS has helped to characterize the unknown peaks and also confirm the novel selenium species from senescing plants, such as MeSSeSMe, MeSSSeMe and (MeSe)2S in the green onions samples supplemented with Se(IV). A close look at the time of flight mass spectra also helped to elucidate other later eluting selenium compounds, the presence of which could only be speculated in the previous studies.6
109
As similar column and chromatographic conditions were used for both GC/ICP-MS and
GC/TOF-MS experiments similar elution order of the selnium compounds was expected. Based on elution profile, the peaks at 7.03 min and 7.28 min in Figure 5.3 were assigned as the peaks labelled as U1 and U2. These unknown species were identified as methylpropyl selenosulfenate
(PrSSeMe) and methyl-1-propene selenosulfenate (MeSeSAll), respectively. The corresponding molecular ions at m/z = 169.971 and 167.954 were identified in the EI+ spectra. A close look at the fragment ions reveal the presence of ion cluster at m/z =92.933 and m/z = 94.958
80 .+ 80 .+ (corresponding to ions CH Se and CH3 Se ) in both cases and at m/z = 73.021 (corresponding
.+ to C3H6S ion) in allylmethyl selenosulfenate spectra leading to the above structures. The relative abundance of the fragment ions in the case of allylmethyl selenosulfenate is similar to described by Cai et al for this molecule.5
Figure 5.4: Extracted ion chromatogram from GC/TOF-MS of 169.967 of green onions supplemented with Se(IV). Corresponding mass spectra (A) of peak at 7.03 min for PrSSeMe and (B) of peak at 7.28 min for MeSeSAll.
110
Identity of these two species in the sample was also confirmed through mass defect mask analysis as described earlier in the text. Using the above signal filtering strategy we were able to easily locate Se-containing compounds of interest (in this example, alkyl selenosulfenates – containing -Se-S- group) from the complex GC/TOF-MS mass spectra. Figure 5.5 shows the strategy of such search. All the mass spectra were merged together from the chromatographic window of interest yielding a raw mass spectrum which then was subjected to the SeS-specific search as shown in Scheme 5.1. This resulted in a very simple spectrum where the presence of
Se-containing isotope pattern at m/z = 168 was evident. Using this information one can now obtain the conventional individual ion chromatogram of the identified mass (167.95 u) which resulted in two compounds: methylpropyl- and allylmethyl selenosulfenate (Figure 5.5). It is interesting to note that such strategy basically allows us to generate ICP-like element-specific chromatograms from the TOF-MS data as shown in Figure 5.5.
111
(a) (c)
(d)
(b)
Figure 5: The search for selenosulfenates from the SPME-GC/TOF-MS data of Se(IV) supplemented green onions. (a) Total ion chromatogram, (b) Extracted ion chromatogram of 167.95 ion (c) raw mass spectrum merged within 7.0 min and 8.0 min (d) filtered mass spectrum obtained after mass defect analysis
Accurate mass measurements (within 5 mDa accuracy) and the correct isotope pattern were also used to confirm other species such as MeSeMe, MeSeSeMe, MeSeSMe , (MeS)2Se and
(MeSe)2S in the sample (Table 5.2).
112
Table 5.2: Summary of the Se-containing volatiles identified from Se-enriched green onions
Structurea Abundanceb Method of identificationc
CH3SeCH3 (110 Da) 0.3 RT, MS, EC
CH3SeSCH3 (142 Da) 1.7 RT, MS, EC
CH3SeSeCH3 (190 Da) 1 RT, MS, EC
CH3CH2CH2SSeCH3 (170 Da) 14.2 RT, EC
CH3CH=CHSSeCH3 (168 Da) 3.8 RT, EC
CH3SSeSCH3 (174 Da) <0.3 RT, MS, EC
CH3SSSeCH3 (174 Da) <0.3 RT, MS, EC
CH3SSeSeCH3 (222 Da) <0.3 RT, MS, EC
CH3CH2CH2SSeSCH3 (202 Da) - EC
CH3CH2CH2SSeSCH2CH2CH3 (230 Da) - EC
aMonoisotopic molecular weight based on 80Se.
b Relative abundance from all the Se-containing volatiles as compared to CH3SeSeCH3 determined with
-1 GC/ICP-MS. CH3SeSeCH3 headspace concentration was found to be at about 0.45 ng µL
cRT: retention time matching with standards as performed on a 30 m and 60 m DB-5 (0.25 µm and 1 µm
respectively) capillary columns. MS: electron impact mass spectra compared to mass spectra of the standards.
EC: elemental composition and isotope patterns verified with high resolution GC/TOF-MS.
Also, the presence of two structural isomers of (MeS)2Se, MeSSeSMe and MeSeSSMe, were also identified first time in natural samples. The distinction between the symmetric and the asymmetric form was made from their differences in retention times and EI+ fragmentation patterns. The symmetric form (MeSSeSMe) eluted earlier than the asymmetric form (MeSSSeMe). As shown in
Figure 5.6, the selenium containing ion cluster at m/z = 141.938 (formed as chalcogen exchange product 16) is more pronounced in the species MeSeSSMe in the EI+ mass spectra. This is expected due to presence of MeSe- and MeS- containing group in the compound. Mass spectra of these two isomers also differ by the presence of fragment ion at m/z = 126.9093 (MeSSe+) in case
113
of MeSSeSMe. This ion is of at relatively small abundance the in case of MeSSSeMe. Both the isomers were identified in the sample (Figure 5.6b).
Figure 5.6: Extracted ion chromatogram (EIC) from GC/TOF-MS of 173.907 of a) standard mixture including
CH3SSSeCH3 and CH3SSeSCH3 b) and enriched green onion. Mass spectra of: peak (1) CH3SSeSCH3 from
standard, peak (2) CH3SSSeCH3 from standard, peak (3) CH3SSeSCH3 from sample, and peak (4) CH3SSSeCH3 from sample
114
In the case of C2H6Se2S, only one isomer could be detected from Se-enriched onions. In order to confirm the presence of this species, comparison of the mass spectra of synthetic
MeSeSSeMe and MeSeSeSMe species with the mass spectra from the sample was undertaken.The two isomers display different fragmentation pattern. As shown in Figure 5.7, the mass spectrum of MeSeSSeMe exhibits very pronounced selenium containing cluster at m/z =
189.882. On the other hand, MeSeSeSMe is characterized by the presence of two selenium
+ containing ion clusters at m/z= 159.834 and at m/z=141.936 corresponding to Se2 and
MeSeSMe+, respectively. As described earlier, ion at m/z= 141.936 is due to presence of MeS- and Me-Se group in the molecule. These two species also exhibit different retention time on the column. The mass spectrum of peak at 9.43 min from the sample is in agreement with the synthetic MeSeSeSMe with the characteristic presence of fragment ion m/z = 159.834 cluster
(with its characteristic Se2 isotope pattern) and absence of ion cluster at m/z = 189.882.
115
Figure 5.7 | Mass spectra of CH3SeSeSCH3 from the Se-enriched onions compared to the standard spectra of both
C2H6Se2S isomers, CH3SeSeSCH3 and CH3SeSSeCH3.
116
Hence only the asymmetric form of (MeSe)2S could be identified in the Se-enriched green onions.
A close look at the TOF mass spectra also helped to elucidate the presence of other later eluting selenium compounds which could only be suggested in the previous studies. One of such compounds eluted at 11.05 min with the exact mass of 201.9413 Da and elemental composition
C4H10S2Se (Figure 5.3). Accurate molecular weight measurement and the correct isotope pattern corresponds to the species PrSSeSMe. This species was suggested to be present in the onion oil by the Cai et al.6 However, lack of an available standard and its presence at very trace levels left its identification unconfirmed. PrSSeSPr (with the corresponding molecular ion at 229.9678 Da) was also found to be present in the sample (at the retention time of 13.35 min) (Figure 5.3).
Presence of this species in plant samples has never been reported.
It is interesting to note that wide variety of methyl and propyl containing selenosulfenates have been identified to be released from the allium plant investigated in this study. The result obtained is in contrast to the selenium accumulating plants where major selenium containing volatile released is dimethyl selenide. Of all the selenium volatiles present, methylpropyl selenosulfenate is found to be most abundant in the headspace of Se-supplemented green onions
(Table 5.2). Also, there are wide varieties of propyl group containing sulfur volatiles present in the sample such as methyl propyl disulfide, dipropyl disulfide, propyl propenyl disulfide, propyl trisulphide etc (data not shown). Dipropyl disulfide is the most abundant peak in GC/TOF-MS chromatogram. Since propyl group containing disulfides are the most abundant sulfur volatiles in onions, this might suggest that methylpropyl selenosulfenate might be an S/Se exchange product between the major propyl group containing disulfides and dimethyl diselenide.
117 5.5 CONCLUSIONS A study of the minor Se-containing volatiles in senescing Se-enriched green onions leads to the identification of previously unreported selenium species such as MeSSSeMe and MeSeSeSMe.
Both isomers of MeS2SeMe (MeSSeSMe and MeSSSeMe) have been detected, while only one
Me2Se2S isomer, MeSeSeSMe, was found to be present in the samples. Variety of propyl group containing selenosulfenates was also identified in the sample. The beneficial complementary use of both atomic and molecular spectrometry to approach the trace level Se-volatiles has been demonstrated.
5.6 REFERENCES
(1) Clark, L. C.; Combs, G. F.; Turnbull, B. W.; Slate, E. H.; Chalker, D. K.; Chow, J.; Davis, L. S.; Glover, R. A.; Graham, G. F.; Gross, E. G.; Krongrad, A.; Lesher, J. L.; Park, H. K.; Sanders, B. B.; Smith, C. L.; Taylor, J. R. Jama-Journal Of The American Medical Association 1996, 276, 1957-1963. (2) Whanger, P. D.; Ip, C.; Polan, C. E.; Uden, P. C.; Welbaum, G. Journal Of Agricultural And Food Chemistry 2000, 48, 5723-5730. (3) Kotrebai, M.; Birringer, M.; Tyson, J. F.; Block, E.; Uden, P. C. Analyst 1999, 125, 71- 78. (4) Shah, M.; Kannamkumarath, S. S.; Wuilloud, J. C. A.; Wuilloud, R. G.; Caruso, J. A. Journal Of Analytical Atomic Spectrometry 2004, 19, 381-386. (5) Block, E.; Cai, X. J.; Uden, P. C.; Zhang, X.; Quimby, B. D.; Sullivan, J. J. Pure And Applied Chemistry 1996, 68, 937-944. (6) Cai, X. J.; Uden, P. C.; Block, E.; Zhang, X.; Quimby, B. D.; Sullivan, J. J. Journal Of Agricultural And Food Chemistry 1994, 42, 2081-2084. (7) Meija, J.; Montes-Bayon, M.; Caruso, J. A.; Sanz-Medel, A. Trac-Trends In Analytical Chemistry 2006, 25, 44-51. (8) Meija, J.; Montes-Bayon, M.; Le Duc, D. L.; Terry, N.; Caruso, J. A. Analytical Chemistry 2002, 74, 5837-5844. (9) Meija, J.; Bryson, J. M.; Vonderheide, A. P.; Montes-Bayon, M.; Caruso, J. A. Journal Of Agricultural And Food Chemistry 2003, 51, 5116-5122. (10) Meija, J.; Beck, T. L.; Caruso, J. A. Journal Of The American Society For Mass Spectrometry 2004, 15, 1325-1332. (11) Arnault, I.; Auger, J. Journa of Chromatography A 2006, In press. (12) Hughey, C. A.; Hendrickson, C. L.; Rodgers, R. P.; Marshall, A. G.; Qian, K. N. Analytical Chemistry 2001, 73, 4676-4681. (13) Hall, M. P.; Ashrafi, S.; Obegi, I.; Petesch, R.; Peterson, J. N.; Schneider, L. V. Journal Of Mass Spectrometry 2003, 38, 809-816.
(14) Meija, J. Analytical And Bioanalytical Chemistry 2006, In press. (15) Zhang, X.; Hines, W.; Adamec, J.; Asara, J. M.; Naylor, S.; Regnier, F. E. Journal Of The American Society For Mass Spectrometry 2005, 16, 1181-1191. (16) Meija, J.; Caruso, J. A. Inorganic Chemistry 2004, 43, 7486-7492. (17) Swearingen, J. W.; Frankel, D. P.; Fuentes, D. E.; Saavedra, C. P.; Vasquez, C. C. V.; Chasteen, T. G. Analytical Biochemistry 2006, 348, 115-122.
119
CHAPTER 6 | LOCALIZATION AND CHARACTERIZATION OF SELENIUM CONTAINING HIGH MOLECULAR WEIGHT SPECIES IN MUNG BEAN (VIGNA RADIATA) SPROUTS
120 6.1 ABSTRACT
In this study, an approach involving combination of size-exclusion chromatography-ICP-MS with the gel electrophoresis was employed for the identification of selenium containing protein in the plant tissue. Sub-cellular fraction is performed for localization of selenium containing high molecular weight species in mung bean sprouts (Vigna radiata) tissue and also for matrix-clean for the determination of protein by mass spectrometric methods. The results revealed that selenium is incorporated in the proteins associated with the cell wall. Tryptic digest of the gel electrophoresis protein band followed by the MALDI-MS analysis showed the presence of the selenopeptide with the molecular mass of 2120 kDa, confirming the presence of selenium containing proteins in the mung bean sprouts. 6.2 INTODUCTION
Selenium is an essential trace element required for the proper function of the several enzymes such as selenoprotein P, Selenoprotein W, glutathione peroxidase, and iodothyronine deiodinase in animals.1 These proteins contain the selenoamino acid selenocysteine. SeCys, at the active site and undergo oxidation-reduction reactions to maintain the biological activity of the proteins.
Incorporation of SeCys into selenoprotein is achieved through cotranslation process coded by a
UGA codon. On the other hand, there is no evidence for the specific incorporation of SeCys into proteins in land plants.2 So essentiality of Se for land plants remains uncertain.
Selenium incorporation into amino acids and proteins can also take place non-specifically when the sulfur containing amino acids are replaced by selenoamino acids such as selenomethionine and selenocysteine. Although, there is no evidence of selenoproteins in higher plants, there are data showing the presence of selenoamino acids and small peptides, such as selenomethionine, selenocysteine, seleno-methylselenocysteine, selenocystathionine, γ-glutamyl-
Se-methylselenocysteine, etc. in plant tissue.3-8 Based on their physiological response to the selenium, higher plants can be divided into three groups, accumulators, hyperaccumulators and non-accumulators. Most of the plants in the above categories are not ready sources of proteins, hence determination of trace levels of selenium containing proteins in the presence of complex plant matrix is a difficult task.5 A sensitive analytical methodology is required, which allows the identification of selenoproteins after their purification from the plant matrix.
In an attempt to identify the selenium containing proteins in higher plants, selenium enriched mung bean (vigna radiata) seedlings have been used as a model system. It is a non- accumulating selenium deficient legume. Mung beans like other beans have high protein content and are widely consumed as sprouts in United States and are an important lentil for the Indian
diet. It has been shown in the past that in mung bean seedlings grown with Se, enhanced the respiratory control ratio and succinate dehydrogenase activity in the mitochondria of tissues, indicative of a role for Se in mitochondrial membrane functions. 9, 10 Also, it has been shown that
80% of the selenium is incorporated in the proteins in some non-accumulators plants.1 Hence both specific and non-specific incorporation of selenium into proteins is expected to be present in this non-accumulator legume.
This aim of this preliminary study is focused on localization of the selenium containing macromolecules in the plant tissues after subcellular fractionation. Subcellular fractionation into various organelles helps to divide the tissue into discrete functional units helps trace the presence of high molecule selenium containing species within cellular component.11 It also represents an attractive approach to determine the Se proteome, as the complexity of the plant matrix is reduced by fractionation into individual components. The matrix clean-up by sub-cellular fractionation is also advantageous for the determination of protein by mass spectrometric methods where signal suppression due to large amount of matrix components is unavoidable.
Before characterization using molecular mass spectrometric techniques, inductively coupled plasma mass spectrometry coupled to size exclusion chromatography (SEC-ICP-MS) is used for the purification and selective extraction of selenium containing species from the other matrix components. In order to separate the selenium containing fraction into individual protein components, gel electrophoresis is introduced after SEC-ICP-MS analysis of the extract. In-gel tryptic digest was performed of the protein bands and the extract was subjected to matrix-assisted laser desorption ionization (MALDI-MS) and electrospray mass spectrometry for possible molecular mass determination and peptide mapping of selenoproteins.
123
6.3 EXPERIMENTAL
6.3.1 Instrumentation
Chromatographic separations were performed using an Agilent 1100 liquid chromatographic system (Agilent Technologies, Palo Alto CA, USA) equipped with an HPLC binary pump, an auto sampler, vacuum degasser, a thermostated column compartment and a diode array detector.
The chromatographic column used was a Superdex 75 HR (10mm x 300mm x 13um particle size) column (Amersham Pharmacia Biotech AB, Uppsala, Sweden) for size-exclusion chromatography. Chromatographic conditions are summarized in Table1.
An Agilent 7500ce ICP-MS (Agilent Technologies, Tokyo, Japan) equipped with a micromist nebulizer and a Peltier cooled spray chamber (2oC) and a shield torch system was used for selenium specific detection. The outlet of the UV detector was connected online to the liquid sample inlet of the ICP-MS nebulizer using a 300 mm long by 0.25 mm PEEK tubing. The instrumental operating conditions are summarized in Table 6.1.
Table 6.1. ICP-MS and chromatographic instrumental parameters
ICP-MS parameters Forward power 1500 W Plasma gas flow rate 15.0 l min-1 Auxiliary gas flow rate 0.87 l min-1 Carrier gas flow rate 1.20 l min-1 Dwell time 0.1 s per isotope Isotopes monitored 77Se, 78Se, 80Se, 82Se SEC chromatography parameters Column Superdex 75 HR 10/30 Mobile phase 0.03 mol⋅l-1 Tris-HCl buffer, pH 8.0 Flow rate 0.7 ml⋅min-1 Injection volume 100 μl
124
6.3.2 Reagents
Analytical reagent grade chemicals and doubly deionised water (18.2 MΩ.cm, NanoPure treatment system (Barnstead, Boston, MA, USA) was used throughout.
For the determination of total selenium working solutions were prepared daily by appropriate dilution of 1 mg ml-1 Se(IV) standard solution. Nitric Acid (Suprapure) 68% from
Pharmaco and Hydrogen Peroxide 30% from Fisher Scientific were used for sample digestion.
The chromatographic mobile phase (Table 6.1) was prepared from Sigma reagent tris
(hydroxymethyl) aminomethane (TRIS). Following analytical reagents (purity > 90% ) were also purchased from sigma for in-gel tryptic digestion: Ammonium bicarbonate, trypsin, formic acid, acetonitrile.
6.3.3 Preparation of selenium-enriched mung bean seedlings
The mung bean seeds were purchased from local Indian grocery stores. The seeds were washed well with deionized water, surface sterilized with 0.1% mercuric chloride for 2 min, rinsed thoroughly with deionized water and soaked overnight in deionized water (without Se) as control and in batches individually supplemented with Na2SeO3 solution at 1, 2, 5 and 10 µg/ml Se levels. The seeds were germinated over moist filter paper in Petri dishes, maintained in the dark and replenished with the respective solutions.
6.3.4 Total Selenium Determination
For determinations of total selenium by ICP-MS, microwave digestion was employed. Prior to this, the fresh biomass was freeze-dried. The mineralization program used for microwave digestion involved three steps with first and second step of 5 minutes each and power applied was 100 and 600 W, respectively. Power was increased to 1000W in the third step and was held at that value for another 10 minutes. Approximately 0.05 g of dried sprouts sample was digested
125
by using 3 ml HNO3 (65%) and 1 ml H2O2 (30%). The end solutions were diluted with distilled water up to 10 ml. The reagent blank was digested in the same way. The total selenium concentration was determined by ICP-MS using 89Y as an internal standard. The isotope 78Se,
77Se, 82Se was used for the measurements. The total selenium content of each sample is summarized in Table 6.2.
Table 6.2. Se uptake by the mung bean seedlings
Se amount added to the growth Total Se content in the sprouts medium (µg g -1) (µg g -1) ± SD 1.0 3.10 ± 1.5 2.0 4.10 ± 2.5 5.0 32.9 ± 6.3 10.0 52.9 ± 8.5
6.3.5 Fractionation Procedure
Fresh biomass of the sprouts grown at the concentration of 10.0 µg g-1 was homogenized in liquid nitrogen. For cellular fractionation, ~10 g of the homogenized material was suspended in
30 ml of Tris-HCl buffer (50 mmol L-1, pH 7.0) containing 1 mM protease inhibitor PMSF and 5 mM EDTA. Fresh biomass (0.5 kg) of selenium enriched sprouts was homogenized in liquid nitrogen and the obtained powder (about 10 g) was suspended in 50 mL of Tris-HCl buffer (50 mmol/L, pH 7.0) containing protease inhibitors (1 μg/mL) [extraction buffer]. The extract was fractionated by centrifugation as described elsewhere (Scheme 6.1).12 Briefly, two subsequent centrifugations were then carried out (centrifugation I: 7300g, 4°C, 10 min; centrifugation II:
147,000g, 4°C, 60 min) yielding pellet I (cell walls), supernatant I, pellet II (mixed membrane fraction), and supernatant II (cytosol). Each cellular fraction was freeze-dried prior to further analysis.
126
Sprouts- fresh biomass
Homogenized in liquid nitrogen
Suspended in extraction buffer
Centrifugation I 7300 g, 4oC, 10 min
Pellet I Supernatant I Cell Walls Cytosol + membranes
Freeze-dried Centrifugation II 147000 g, 4oC, 10 min
Pellet II Supernatant II Membrane fraction Cytosol
Freeze-dried Freeze-dried
Scheme 6.1: Procedure for sub-cellular fractionation of sprouts tissue
6.3.6 Chromatographic speciation analysis
Depending upon the availability of the sample, about 0.02-0.1 g of each freeze dried cellular fraction was weighed in an eppendorf tube and extraction of the selenium species was performed using 0.5-1 ml of Tris buffer solution (0.03 mol l-1, pH 7.5). The mixture was shaken on a Vortex system overnight and centrifuged at 5000 rpm for 10 min. 100 μl of the supernatant was introduced into the SEC-ICP-MS system.
Additionally, a simple digestion procedure was employed in order to determine the molecular nature of the high molecular weight selenium species in the cell wall fraction of the species. Proteinase K hydrolysis was performed assuming that selenium species are present in proteins associated with the cell walls. The enzyme solution was prepared by adding about 0.03g
127
Protinase K in 5 ml of sodium acetate (0.02 mol l -1 pH 4.7) and Tris-HCl (0.03 mol l-1 pH 7.5) buffer solutions, respectively. The solution was thoroughly stirred and 100 µl of the solution was added to the 100 µl of the freeze-dried fraction containing a high molecular weight selenium peak. The above mixture was sonicated overnight and was subjected to SEC-ICP-MS analysis.
6.3.7 Gel electrophoresis on protein fraction
The high molecular weight selenium peak in the cell wall extract from Superdex 75 size exclusion runs were collected a total of five times. The resulting sample was concentrated using stream of nitrogen prior to separation by gel electrophoresis. Precast 8-16% Tris-HCl (Bio-RAD,
Hercules, CA) polyacrylamide gel was used to obtain a molecular weight profile for the sprout fraction of interest. Sample buffer was prepared by adding 50 µL β-mercaptoethanol to 950 µL
Laemmli sample buffer (BioRad Laboratories, Hercules, CA). SDS-PAGE molecular weight standards (Broad Range, Bio-RAD, Herucles, CA) covering 6.5-200 kDa was prepared through a
1:20 dilution with sample buffer. Concentrated sprout extract was then diluted 1:1 with SDS sample buffer (10% βME). The resulting solutions were heated at 95 ºC for 10 minutes followed by a 30 s centrifugation to remove any condensate. Marker (10 µL) and sample (20 µL) were then loaded on the gel and run at 90 V for 3 hours. Gel staining was performed for 12 hours with
Coomassie Brilliant Blue R-250 (Bio-Rad, Hercules, CA) followed by high destain (40% methanol/10%acetic acid) and low destain (10% methanol/6% acetic acid) for 6 hours each. The resulting gel was digitally photographed and bands of interest submitted for mass spectrometric analysis after tryptic digestion.
128 6.3.8 In-gel tryptic digestion
Three protein spots labeled as 1, 2, 3 were excised from the gel and destained for in-gel digestion of proteins by trypsin. Destaining was performed three times for 10 min with intermittent vortexing with 200 µl of 25 mM ammonium bicarbonate / 50% acetonitrile solution. Each time, the solutions were discarded. The gel sliced shrunk and became white. After that gel particles were dried for 20 min in vacuum centrifuge.
Dried gel spots were treated with trypsin after the reduction and alkylation as described elsewhere.13, 14
6.3.9 MALDI-TOF analysis
All MALDI-TOF MS experiments were done on a Bruker Reflex IV reflectron MALDI-TOF mass spectrometer (Bruker Daltonics, Billerica, MA, USA) equipped with a nitrogen laser as previously described. Peptide mass spectra were obtained in the positive ion mode at an acceleration voltage of 20 kV, extraction plate voltage of 17.1 kV and lens voltage of 10.1 kV by accumulating 300 laser shots. All three digests were analyzed under identical instrumental parameters. The instrument was calibrated externally using two standards. The data were acquired as positive ions in the reflectron mode. This provides the better resolution and mass accuracy compared to the linear mode.
The protein digest was de-salted using C-8 Zip tips and mixed with a-Cyano-4- hydroxycinnamic acid (HCCA) as matrix in a ratio of 1:2 then spotted on the MALDI plate.
After drying the sample at room temperature, it is ready for MALDI analysis.
6.4 RESULTS AND DISCUSSION
6.4.1 Total selenium concentration in samples and extracts
Initially, total selenium determination was performed on the freeze dried sprout sample. Results of the total selenium determination are summarized in Table 6.2. It can be seen that the selenium accumulated by the seedlings was proportional to the Se(IV) added to the growth media.
However, it was observed that exposure to selenite greater than 5.0 µg/ ml selenium significantly inhibited the growth of the seedlings.
6.4.2 Distribution of selenium in sprouts
As a first approach to selenium speciation, freeze dried sprout samples grown under different selenium levels were treated with the 0.03 mol l-1 Tris-HCl buffer (pH 7.5) solution. Tris-HCl is known to solubilize water soluble proteins from the plant tissue. It is clear from Figure 6.1 that selenium is associated with both low as well as high molecular weight fractions in the sprouts.
The fractionation profile shows that three peaks of significant intensities are found in the size exclusion chromatogram. The first peak of low intensity eluting at about dead volume (10 min) represents the association of selenium with the high molecular weight compounds presumably proteins as these species are soluble in alkaline media. Peaks eluting in the region of 19.5 min to
23.6 min indicates association of selenium with medium molecular weight fraction to low molecular weight fractions. These peaks may be due to the presence of low molecular weight proteins, peptides containing selenium amino acids and inorganic selenium. It is also observed that distribution of selenium among the high and low molecular weight fractions in the sprout samples does not change with the levels at which they were exposed to (data not shown). Only peak intensity changed with respect to the total selenium present in the sample. On this basis
130
further speciation studies were performed on the plants with the highest Se concentration (10 µg g-1).
≥ 70 kDa 6.5 kDa 25000 16 kDa ≤ 1.3 kDa
20000
15000
10000 Se Abundance, cps cps Abundance, Se 78
5000
0 010203040 Time, min
Figure 6.1: SEC-ICP-MS chromatogram of the Tris-HCl extract of the whole sprout tissue.
6.4.3 Distribution of selenium in various sub-cellular components
All the selenoproteins identified so far in animals are enzymes, with the selenocysteine residue responsible for their catalytic functions. They include the glutathione peroxidases (GPxs), the iodothyronine deiodinases, the thioredoxin reductases, and a selenophosphate synthetase. Most of the GPxs have been detected in the cytosols of the various animal tissues. Some selenoenzymes are also known to play important role as mitochondrial and membrane proteins.1
As stated earlier, there is limited information on the possible role and the presence of such proteins in the plant kingdom. In this study, sub-cellular fractionation of plant tissue is performed in order to locate the presence and any possible identification of the selenium containing proteins. Three fractions (cell wall, membrane and cytosol fractions) were obtained after the sprout tissue was subjected to the cellular fractionation procedure as described in the
131
experimental section. Each fraction was treated with the Tris-HCl solution in order to release any water soluble proteins from them and the extract was analyzed through SEC-ICP-MS. The results are depicted in the Figure 6.2(a-c). It can be noted from the figure that the elution profile of selenium in various sub-cellular fractions such as cell wall, membrane and cytosol fraction is very similar. However, it is of interest to note that the relative distribution of the selenium containing peaks in the cell wall fraction is different from other two fractions. The intensity of high molecular weight selenium containing species is higher relative to the other peaks in the cell wall fraction as compared to other two fractions (Figure 6.2a). This indicates the association of selenium to the cell wall of the plant tissue. Based on the peak area comparison, it is also seen that majority of the selenium is incorporated into molecules in the cytosol mainly in the medium to low molecular weight fraction (Figure 6.2b). These could be due to presence of low molecular weight selenium containing proteins, peptides or selenoamino acids.
132
Figure 6.2: SEC-ICP-MS chromatogram of the Tris-HCl extract of a) Cell wall fraction b) cytosol fraction and c) membrane fraction.
Since focus of this work was characterization of high molecular weight selenium containing species, the cell wall fraction of the plant tissue was further analyzed. The high molecular weight peak eluting at dead volume (≥ 70 kDa) (Figure 6.2 a) of the superdex column was collected 5 times for this purpose. Recent study in the lower organisms such as yeast and fungi have shown that selenium can be associated with the proteins of their cell wall.15, 16 In order to find the molecular nature of the high molecular weight peak in the cell wall fraction of bean sprouts, proteolysis was performed. The selenium containing fraction was treated with Proteinase
K. This is an endolytic protease known for cleaving carboxylic side of aromatic, aliphatic and hydrophobic amino acids. It can be seen from figure 6.3 that after the action of Proteinase K on
133
the high molecular weight peak, the conversion of the high molecular weight species into low molecular weight compounds, indirectly indicating the association of selenium to proteins of cell wall. In order to further confirm the association of selenium to proteins in cell wall; gel electrophoresis was also performed on this fraction.
1200
1000
800
600
400 Se Abundance, cps cps Abundance, Se 78
200
0 0 10203040 Time, min
Figure 6.2: SEC-ICP-MS chromatogram of the high molecular weight peak in the cell wall fraction (shown in black line) and after Proteinase K digestion (shown in blue line)
6.4.4 Gel electrophoresis on SEC fraction
Due to limited peak capacity, complete separation of compounds in complex biological matrices cannot be obtained by size exclusion chromatography.17 In such cases application of other kinds of separation techniques such as RP-IP, cation exchange or anion exchange and gel electrophoresis is usually followed on SEC fractions when greater resolution is desired for species identification. Multidimensional approaches are becoming increasingly popular in speciation studies, especially in cases when peak purity is critical for characterizing particular species by molecular mass spectrometry such as ESI-MS and ESI-MS-MS. In this study, gel
134
electrophoresis is used to confirm the molecular nature of the selenium containing peak as protein and its further separation and more accurate molecular weight determination.
Figure 6.4 shows the results of gel electrophoresis analysis of the SEC fraction. It can be seen that peak eluting at dead volume of the superdex column actually contained three protein bands in region of about 200-97 kDa, 67 kDa and 66-45 kDa. The high molecular weight of protein band is expected because of inability of superdex 75 column to separate proteins higher than 70 kDa.
Figure 6.4 : Representative picture of gel electrophoresis of fraction 1 (Right lane), left lane showing molecular weight standards.
6.4.5 MALDI-MS analysis
In order to confirm any selenium containing proteins present in the bands, each band was excised and in-gel tryptic digestion of the band was performed. Each digest was subjected to MALDI-
135
MS analysis. Since selenium exhibits characteristic isotopic patterns, possible presence of similar patterns (one or two Se present in a fragment) was searched for throughout the mass spectra.
Of the three spot, band labeled as 1 (Figure 6.4) was found to contain a peptide with possible selenium isotopic pattern with the molecular weight of 2120 kDa (Figure 6.5). Any other such selenopeptide could not be detected in the other two proteins bands. Further experiments involving LC-ESI-MS were performed in order to confirm the presence of this peptide and subsequent identification of any selenium containing protein. However, low levels and possible contamination by keratins during sample preparation compromised this effort.
Further experiments with a more concentrated fraction needs to be performed in order to determine the exact molecular mass and possible peptide mapping of selenoproteins. This will also help to elucidate the specific or non-specific incorporation of selenium in the mung bean indicating the essentiality of this element in the higher plant.
400 2120 350
300
250 2118 200 2122 Abundance 150 2116
100
50
0 2110 2115 2120 2125 2130 m/z
Figure 6.5: MALDI-MS spectra showing a possibility of selenium containing peptide obtained after tryptic digest of the protein band 1.
136
6.5 CONCLUSIONS
Subcellular fractionation of mung bean sprouts resulted in the localization of selenium containing proteins in the cell wall fraction of the plant tissue. The preliminary results obtained by the combination of SEC-ICP-MS and gel electrophoresis followed by MALDI-MS suggest possible incorporation of selenium to the proteins in this non-accumulating legume. The possible presence of a selenopeptide with the molecular mass of 2120 kDa was also indicated, but further proof is necessary. Subcellular fractionation on the level of sample preparation was found to be beneficial prior to the proteome analysis in the plant sample for the preliminary characterization of selenopeptides at ultratrace levels.
6.6 REFERENCES
(1) Behne, D.; Kyriakopoulos, A. Annual Review on Nutrition 2001, 21, 453-473. (2) Fu, L. H.; Wang, X. F.; Eyal, Y.; She, Y. M.; Donald, L. J.; Standing, K. G.; Ben- Hayyim, G. Journal Of Biological Chemistry 2002, 277, 25983-25991. (3) Gergely, V.; Kubachka, K. M.; Mounicou, S.; Fodor, P.; Caruso, J. A. Journal Of Chromatography A 2006, 1101, 94-102. (4) Infante, H. G.; O'Connor, G.; Rayman, M.; Wahlen, R.; Spallholz, J. E.; Hearn, R.; Catterick, T. Journal Of Analytical Atomic Spectrometry 2005, 20, 864-870. (5) Mounicou, S.; Meija, J.; Caruso, J. Analyst 2004, 129, 116-123. (6) Grant, T. D.; Montes-Bayon, M.; LeDuc, D.; Fricke, M. W.; Terry, N.; Caruso, J. A. Journal Of Chromatography A 2004, 1026, 159-166. (7) Montes-Bayon, M.; Grant, T. D.; Meija, J.; Caruso, J. A. Journal Of Analytical Atomic Spectrometry 2002, 17, 1015-1023. (8) Montes-Bayon, M.; LeDuc, D. L.; Terry, N.; Caruso, J. A. Journal Of Analytical Atomic Spectrometry 2002, 17, 872-879. (9) Easwari, K.; Lalitha, K. Biological Trace Element Research 1995, 48, 141-160. (10) Lalitha, K.; Easwari, K. Biological Trace Element Research 1995, 48, 67-89. (11) Ramos, I.; Esteban, E.; Lucena, J. J.; Garate, A. Plant Science 2002, 162, 761-767. (12) Munoz, A. H. S.; Kubachka, K.; Wrobel, K.; Corona, F. G.; Yathavakilla, S. K. V.; Caruso, J. A.; Wrobel, K. Journal Of Agricultural And Food Chemistry 2005, 53, 5138- 5143. (13) Hellman, U.; Wernstedt, C.; Gonez, J.; Heldin, C. H. Analytical Biochemistry 1995, 224, 451-455. (14) Rosenfeld, J.; Capdevielle, J.; Guillemot, J. C.; Ferrara, P. Analytical Biochemistry 1992, 203, 173-179.
137
(15) Dernovics, M.; Stefanka, Z.; Fodor, P. Analytical And Bioanalytical Chemistry 2002, 371, 473-480. (16) Polatajko, A.; Sliwka-Kaszynska, M.; Dernovics, M.; Ruzik, R.; Encinar, J. R.; Szpunar, J. Journal Of Analytical Atomic Spectrometry 2004, 19, 114-120. (17) Shah, M.; Caruso, J. A. Journal Of Separation Science 2005, 28, 1969-1984.
138
CHAPTER 7 | CONCLUSIONS
139 7.1 CONCLUSIONS
From the previous chapters it is clear that application of ICP-MS and its combined use with molecular mass spectrometric techniques in conjunction with various separation techniques has become a promising analytical tool in characterization of biologically important molecules in complex matrices. Easy on-line coupling with variety of separation methods such as HPLC and
GC have increased the scope of ICP-MS in various applications. The unparallel element selectivity and high sensitivity (irrespective of chemical form of the element) obtained by intervention of ICP-MS in these fields, not only simplifies but also speeds up the overall analysis and also the results achieved are impressive and unequivocal in terms of analytical performance characteristics. On the other hand, multidimensional approach at the detection end proves to be the most exciting analytical tool in positive and precise identification of analytes of interest.
Such approach has been successfully applied in this dissertation to understand the metabolism of certain elements in biological system and studying the fate of some anthropogenic molecules.
Likewise, the inclusion of ICP-MS along with molecular mass spectrometric techniques can benefit many biological and environmental applications where element selective detection can be exploited. However, in order to generate meaningful results from the complimentary use of the two techniques, it is required that the sample preparation is performed keeping in mind the differences in detection capability of the two methods. Signal suppression arising from matrix impurities, fragmentation products or unidentified constituents hinders the analysis of trace amount of species in the samples using molecular mass spectrometric techniques. Care in sample pretreatment, in terms of sample purification and selective extraction of analytes of interests from complex matrices, needs to addressed prior to the method development using combined use of these techniques. As discussed in chapter six, sub-cellular fractionationation and other
biochemical methods may be the possible approach to reduce the matrix components for successful application of the two state of the art analytical techniques for studying the metabolism of the certain element in complex biological system. Preliminary results show that such approach was found to be beneficial in studying the selenium incorporation into the proteins in the plant system. However, a more detailed analysis needs to be performed in order to extend our current knowledge and draw any conclusions on the specific or non-specific incorporation of selenium into proteins. Future work might involve more elaborate fractionationation procedure specific for the higher land plants. Fractionation of cytosol components into various sub-cellular organelles should also be considered in order to explore the possible role of selenium in plant system at cellular level. After the sample pre-treatment, SEC-ICP-MS followed by the gel electrophoresis was performed in order to further reduce the matrix component prior to the proteomics analysis by molecular mass spectrometric techniques. However with the increase in the analytical steps, possibility of contamination at each step is also increased. So tryptic digestion of the fraction collected should also be considered prior to the characterization of selenium containing proteins using molecular mass spectrometric techniques.
In summary, analytical procedures involving atomic and molecular mass spectrometry can be a very successful and powerful approach for problems associated with the biochemical speciation when intervention of biochemical and molecular methods is considered for effective isolation of compounds of interest from the complicated matrix components.
141