Quantitative Bioanalysis - DiVA portaluu.diva-portal.org/smash/get/diva2:171664/FULLTEXT01.pdf ·...

ACTA

UNIVERSITATIS

UPSALIENSIS

UPPSALA

2008

Digital Comprehensive Summaries of Uppsala Dissertationsfrom the Faculty of Science and Technology 412

Quantitative Bioanalysis

Liquid separations coupled to targeted massspectrometric measurements of bioactivecompounds

BJÖRN ARVIDSSON

ISSN 1651-6214ISBN 978-91-554-7135-4urn:nbn:se:uu:diva-8581

�� !"� �""# �� !"$!% &� �� & �� &'��( �� )�� *��(

��

�� ( �""#( +�� ( ,�-�� & �� ( �� ( �� .!�( /! ��( ��( 01�2 34#53!5%%.54!�%5.(

'��&�� -�� & �� &�� &�� & �� -�� ( 0 �� & �� &� �� ) �� & �� &�� ( 6��-�� & �� & �� & ��&�� &�� &� �� & �� ) �� &-�� &� �� ( �� &��) � �� & ��

�� )�� 7 �� 5�� &�� -�� &� �� & �� (0 �� &��

�� &�� & �� &&�� (8� & �� & ��

�� & �� 5�� & �� )�� )�� &&�� &� �� ( �� 5�� & � ��5�)�� -�� &�� ( �� & �� 5�� & � �� -�� &�� & ��( �� )�� )�� &�-�� (�� & �� &��

)�� & �� -�� -�� & �� &�� &��(

� �� *�� )�� ,�-�� *�� 9�� -�� 5&5&�� +�� :��

�� !��" � �� " !�� " �# $%%" �� " �&'($)*+ ��" ��

; �<=� �� ""#

0112 !/%!5/�!.01�2 34#53!5%%.54!�%5.��$�$��$��$��5#%#! >��$??��(@�(��?��A��B��$�$��$��$��5#%#!C

To past, present and future friends

Papers included in this thesis

This thesis is based on the following papers, which are referred to in the text according to their Roman numerals: I Arvidsson, B., Allard, E., Sjögren, E., Lennernäs, H., Sjöberg, P.,

Bergquist, J. Online capillary solid phase extraction and liquid chro-matographic separation with quantitative tandem mass spectrometric detection (SPE-LC-MS/MS) of Ximelagatran and its metabolites in a complex matrix Submitted

II Amirkhani, A., Rajda, C., Arvidsson, B., Bencsik, K., Boda, K.,

Seres, E., Markides, K.E., Vecsei, L., Bergquist, J. Interferon-beta af-fects the tryptophan metabolism in multiple sclerosis patients. EUROPEAN JOURNAL OF NEUROLOGY 12 (8): 625-631 AUG 2005

III Arvidsson, B., Johannesson, N., Citterio, A., Righetti, P.G.,

Bergquist. J. High throughput analysis of tryptophan metabolites in a complex matrix using capillary electrophoresis coupled to time-of-flight mass spectrometry JOURNAL OF CHROMATOGRAPHY A 1159 (1-2): 154-158 AUG 3 2007

IV Pettersson, A., Amirkhani, A., Arvidsson, B., Markides, K.,

Bergquist, J. A feasibility study of solid supported enhanced microdi-alysis ANALYTICAL CHEMISTRY 76 (6): 1678-1682 MAR 15 2004

V Elhamili, A., Wetterhall, M., Arvidsson, B., Sebastiano, R., Righetti,

P.G., Bergquist, J. Rapid CE-TOF-MS separations of peptides and proteins using a monoquaternized piperazine compound (M7C4I) for capillary coatings accepted in ELECTROPHORESIS nov 2007 IN PRESS

Reprints were made with kind permission from the publishers.

Papers not included in this thesis:

� Tärning, J., Bergqvist, Y., Day, N. P., Bergquist, J., Arvidsson, B., White, N. J., Ashton, M., Lindegårdh, N. DRUG METABOLISM AND DISPOSITION 34 (12): 2011-2019 2006

� Dobrovolsky, V.N., Bowyer, J.F., Pabarcus, M.K., Heflich, R.H.,

Williams, L.D., Doerge, D.R., Arvidsson, B., Bergquist, J., Casida, J.E. Effect of arylformamidase (kynurenine formamidase) gene inac-tivation in mice on enzymatic activity, kynurenine pathway metabo-lites and phenotype BIOCHIMICA ET BIOPHYSICA ACTA-GENERAL SUBJECTS 1724 (1-2): 163-172 JUN 20 2005

� Sjögren, E., Arvidsson, B., Allard, E., Bergquist, J., Lennernäs, H.

Quantitative enzymatic kinetic study of Ximelagatran and its me-tabolites in pig liver, Manuscript

� Hallberg, E., Arvidsson, B., Nordborg, A., Bergquist, J., Irgum, K.

Peptic digestion of native S6wt in different environments using hy-drogen-deuterium exchange, Manuscript

� Allard, E., Arvidsson, B., Sjögren, E., Bergquist, J., Sjöberg, P.,

Analytical aspects of analyzing small doubly charged molecules with electrospray ionization mass spectrometry, Manuscript

Other Publications not included in this thesis:

� Arvidsson, B. and Bergquist, J. The Analysis of Kynurenine and its Metabolites, 1st chapter in the book “Kynurenines in the Brain: From Experi-ments to Clinics”, Nova Science Publishers, Inc. 2005

Authors contribution

Paper I – I planned, performed and wrote the paper together with E. Allard. Paper II – Assembled and optimized the analytical experimental setup and performed the experimental analysis. Wrote the initial draft of the paper and participated in completing it with co-authors. C. Rajda interpreted the clini-cal data. Paper III – Planned and performed initial experiments together with N. Johannesson. I completed the experimental analysis by optimizing and vali-dating the method. I wrote the paper. Paper IV – Assembled and optimized the analytical experimental setup and helped in the planning of the LC-MS analysis and interpretation of the LC-MS data. Paper V – Participated in planning and optimization of experimental ana-lytical parameters, and the supervision of the experimental analysis per-formed by A. Elhamili.

Contents

1. Introduction...............................................................................................13 1.1 Historical perspective .........................................................................14 1.2 Analytical chain of events ..................................................................15

2. Applications ..............................................................................................16 2.1 Today’s challenges .............................................................................16 2.2 Endogenous compounds.....................................................................16

2.2.1 Tryptophan and its metabolites...................................................17 2.2.2 Neuropeptides .............................................................................17

2.3 Exogenous compounds.......................................................................18 2.3.1 Pharmaceuticals and drugs .........................................................18

3. Strategy .....................................................................................................19

4. Sample treatment ......................................................................................21 4.1 Protein precipitation ...........................................................................21 4.2 Microdialysis......................................................................................22 4.3 Solid phase extraction ........................................................................23

5. Liquid separation ......................................................................................24 5.1 Liquid chromatography ......................................................................24

5.1.1 Reversed-phase chromatography................................................25 5.1.2 Mobile phase composition..........................................................26 5.1.3 Miniaturization ...........................................................................27

5.2 Capillary electrophoresis....................................................................28 5.2.1 Modification of the silica wall ....................................................30

6. Ionization ..................................................................................................31 6.1 Electrospray ionization.......................................................................32

6.1.1 Hyphenation of LS and MS with ESI .........................................34 6.1.2 Matrix effects..............................................................................36

7. Mass spectrometry ....................................................................................37 7.1 Mass analyzers ...................................................................................38

7.1.1 Quadrupole mass analyzer ..........................................................38 7.1.2 Time-of-flight mass analyzer......................................................39

7.2 MS/MS ...............................................................................................39

8. Quantitation ..............................................................................................41 8.1 Calibration methods ...........................................................................41

8.1.1 External standard calibration method .........................................41 8.1.2 Internal standard calibration method ..........................................42 8.1.3 Method of standard addition.......................................................44

9. Validation..................................................................................................46

10. Concluding remarks ................................................................................49 10.1 Future aspects ...................................................................................50

11. Acknowledgements.................................................................................51

12. Summary in Swedish ..............................................................................52

References.....................................................................................................56

Abbreviations

ACN Acetonitrile CE Capillary electrophoresis CID Collisionally induced dissociation CRM Charge residue model CSF Cerebrospinal fluid Da Dalton dc Direct current EOF Electroosmotic flow ESI Electrospray ionization i.d. Inner diameter IEM Ion evaporation model IS Internal standard LC Liquid chromatography LOD Limit of detection LOQ Limit of quantitation LS Liquid separation m/z Mass-to-charge ratio MeOH Methanol MRM Multiple reation monitoring MS Mass spectrometry MS/MS Tandem mass spectrometry Q Quadrupole mass analyzer QC/QA Quality check or Quality assurance QQQ Triple quadrupole mass analyzer rf Radio frequency RPLC Reversed-phase liquid chromatography S/N Signal-to-noise ratio SPE Solid phase extraction TIC Total ion current TOF Time-of-flight

13

1. Introduction

Analytical chemistry is the science of developing and improving methods for detection and determination of artificial and naturally occurring components in our surroundings and environment as well as within ourselves, in our tis-sues and body fluids. There are two ways of approaching an analytical prob-lem, either qualitatively or quantitatively. Qualitative analysis seeks to estab-lish the presence and identity of a species within a given sample, while quan-titative analysis seeks to establish the relative and/or absolute abundance of the species. The analytical chain of events consists of several equally impor-tant steps for the outcome of the analysis; planning, sampling, sample treat-ment, separation, detection, evaluation and interpretation. In any of these steps there are risks of making or obtaining errors that will complicate the analysis. With proper knowledge, planning and experience these errors can be minimized or even eliminated. The aim within method development for a research analyst is to achieve a robust and reproducible system setup with the ability to detect and determine low abundant species in a challenging matrix without making those errors.

For a long time focus has been on hardware, improving the instrumenta-tion and technology used for analysis as well as simplifying the use of the software to handle it. This has generated advanced fully automated platforms for detection of low abundant molecules in extremely complex samples of huge sample sets at fast speed, without almost any effort. With these easily managed analytical systems and high degree of sample throughput, with each sample generating lots of data, less demand are put on the analyst per-forming the analysis. However, this can possibly lead to mistakes by less experienced operators, complicating the interpretation of results and thus obstructing the chance of accurate and precise conclusions. Yesterday, we asked ourselves whether we obtained the correct answers. Today, with tools and instruments so sophisticated and generating ever in-creasing amounts of complex multidimensional information, the main focus should instead be whether we ask the correct questions? Data collected today will give rise to discoveries tomorrow.

14

1.1 Historical perspective To understand the advances made within the research field covered in this thesis, one has to take a look in the rear mirror and reflect upon the historical events leading to where we are today 1:

1897 J.J. Thomson discovers the electron and determines its mass-to-charge ratio (m/z) (Noble prize in 1906)

1912 J.J. Thomson constructs the first mass spectrometer (MS) (called pa-rabola spectrograph) 2. He observes negative and multiply charged ions as well as metastable ions.

1948 A.E. Cameron and D.F. Eggers publish a design and mass spectra for a linear time-of-flight (TOF) mass analyzer 3. W. Stephens pro-posed the concept of this analyzer in 1946 4. A. Tiselius receives the Nobel Prize for his contributions both to electrophoresis and absorption analysis, and especially for his dis-covery of the complex nature of proteins in blood serum 5.

1953 W. Paul and H.S. Steinwedel describe the quadrupole mass analyzer (Q) and the ion trap in a patent. (W. Paul received the Noble prize in 1989)

1958 Bendix introduces the first commercial linear TOF instrument 1967 F.W. McLafferty and K.R. Jennings introduce the collision-induced

dissociation (CID) procedure 6, 7. 1968 Finnigan introduces the first commercial quadrupole mass spec-

trometer 1973 R.G. Cooks, J.H. Beynon, R.M. Caprioli and G.R. Lester publish the

book Metastable Ions, a landmark in tandem mass spectrometry 8. 1974 First liquid chromatography/mass spectrometry (LC/MS) coupling

by P.J. Arpino, M.A. Baldwin and F.W. McLafferty 9. 1978 R.A. Yost and C.G. Enke build the first triple quadrupole (QQQ)

mass spectrometer 10. 1982 Finnigan and Sciex introduce the first commercial triple quadrupole

mass spectrometers 1984 M. Yamashita and J.B. Fenn demonstrates the electrospray ioniza-

tion (ESI) as a mass spectrometric technique for small molecules 11. 1987 R.D. Smith describes the coupling of capillary electrophoresis (CE)

with mass spectrometry 12. 1988 J. B. Fenn demonstrates the use of electrospray ionization for ana-

lyzing proteins above 20 000 Dalton (Da) 13. This concept was first proposed by M. Dole in 1968 14.

1994 M. Wilm and M. Mann describe the nanoelectrospray ionization in-terface (by then called the microelectrospray source) 15.

15

1.2 Analytical chain of events An analytical method is most often performed in several continuous steps, sometimes called the analytical chain of events (see Figure 1). This chain consists of several parts and independent processes all with the aim to en-hance the possibility for an accurate interpretation of results. The chain of events can look very different depending on the application or the intended purpose of the analysis, but in general it consists of; experimental design and planning, sampling and handling, sample treatment, separation of sample components, detection, evaluation, interpretation of the results, and valida-tion.

In developing successful analytical methods including the above men-tioned events, the analyst should start from the end by asking the questions making the foundation of the analysis; how much do we have of a certain species in our sample? What calibration method should be used? What is in the sample besides the analyte of interest? How can the response be evalu-ated? What detector should be used? How can a separation be performed? Do we need several separation steps? Can we clean and prepare the sample? How should the sampling be performed? What experiments should be per-formed? And so on, all necessary for the strategy of the method develop-ment. With this thesis, I will try to pinpoint the means, as well as the challenges, for optimization and considerations affecting the outcome of method devel-opment within quantitative mass spectrometric analysis of targeted bioactive compounds.

Figure 1. A schematic picture showing the analytical chain of events. Each step in the analytical chain is a piece of a puzzle.

16

2. Applications

Research in analytical chemistry usually involves fundamental research, like improvement of tools and chemical reactions that enhance the analytical possibilities, before further development. The scope for most of these im-provements is to be able to address new kinds of applications, new ways to monitor or identify species, which will generate new knowledge from the questions asked.

2.1 Today’s challenges Analysis of biological samples will always be challenging, due to their great diversity and complexity. Furthermore, they are very susceptible to degrada-tion and must be handled with care. Aside from containing peptides and proteins, samples usually include cells, salts, carbohydrates, and phospholip-ids etc., all making the analysis more demanding. Today’s challenges within life science and bioanalysis lies in the increasing need for tools to discover and monitor upcoming and present disease states, environmental pollutants, administrations of pharmaceuticals, and so on. Analyzing complex samples, like blood plasma, cerebrospinal fluid (CSF), tissue or urine, is a significant challenge even with today’s advanced instrumentation. It is not as much up to the sophisticated instruments and technologies, or wide range of separa-tion materials and solutions, as to the skill and experience of the researcher performing the method development.

The papers presented in this thesis focus on different relevant applica-tions of bioactive compounds, like for instance, pharmaceutical metabolism in the liverI, neurotransmitter metabolites involved in the pathogenesis of neurological disordersII,III , proteins and peptidesV , and signal substances like neuropeptidesIV.

2.2 Endogenous compounds The word endogenous means “arising from within”, and is used within life sciences as a term for molecules produced and active in the body of a living organism. Constantly, different processes takes place in our brains, ranging from the production or pathogenesis of cells, to signal substances, or neuro-

17

transmitters, traveling as couriers to another destination, within or between cells, in order to deliver their chemical messages. Many of these processes can be monitored, and since most of them will leave traces in the cerebrospi-nal fluid surrounding and protecting the brain, analysts can sample the CSF and perform analyses to obtain answers about these processes by monitoring it over a set duration of time. Some of these compounds will also be present in the blood, increasing the availability, since sampling of blood is much simpler than CSF.

2.2.1 Tryptophan and its metabolites In Paper II and III, two different approaches has been developed in order to monitor and quantify the amino acid tryptophan and some of its metabolites. These molecules are known to be of great interest in neurological diseases, especially in neurodegenerative disorders, such as multiple sclerosis 16, Alz-heimer’s disease 17, Parkinson’s disease 18, 19, schizophrenia 20, epilepsy 21, Down’s syndrome 22, Huntington’s disease 23, anxiety and depression 24. Paper II was performed in order to study the effect of administered beta-interferon on multiple sclerosis patients naive to the treatment and those under long-term treatment, all sampled at different times after administra-tion. These results were compared to samples obtained from a healthy con-trol group, also quantified with the same analysis method. In Paper III, there was less focus on the effects of the molecules and more emphasis on the miniaturization of the analysis, but of course, for the sake of finding new tools to study these effects.

2.2.2 Neuropeptides Neuropeptides is a group of molecules used by the central and peripheral nervous system to transfer chemical messages within and between individual cells. They are generally large and bulky molecules (500 – 5000 Da), present in vivo at low concentrations, ranging from pM (10-12 mol/L) to μM (10-6 mol/L). Analysis of neuropeptides is challenging due to rapid neurotransmis-sion processes, their low levels of concentration and that they are present in complex matrices, with large contents of other endogenous species. In Paper IV, an enhanced microdialysis method, introducing the use of a solid sup-port, was evaluated as a sample pretreatment method, and gave larger recov-eries compared to ordinary microdialysis.

18

2.3 Exogenous compounds The word exogenous means “produced outside” and is a term used in life sciences for molecules that is active within a living organism but originates from elsewhere, for instance pharmaceuticals and drugs.

2.3.1 Pharmaceuticals and drugs It is commonly known today that liquid chromatography coupled to tandem mass spectrometry is the preferred methodology for quantitation of targeted exogenous (and endogenous) compounds, like drug and metabolites 25-27. The selectivity gained by separation together with mass spectrometric meas-urements is necessary to distinguish between similar molecules like drugs and their metabolites, as well as the ability to monitor trace levels of the analytes.

One of the challenges to overcome when metabolites are to be analyzed in a sample is that these are derivates from a parent drug, and is assumed to share the same fragment ions as the parent drug 28. Furthermore, the metabo-lites can also be produced through biotransformation of the parent drug and will interfere and complicate the quantitation 29, 30. In Paper I, an analytical method with several separation steps performed on-line is described. This method was developed using compounds especially chosen to complicate the method development, where they share different polarities in pair, and origi-nates from the same parent drug, ximelagatran, a thrombin inhibitor. Bio-transformation of ximelagatran is carried out in the liver by hydrolysis of the ethyl ester and reduction of the hydroxyamidine group.

19

3. Strategy

“However beautiful the strategy, you should occasionally look at the re-sults.” - Winston Churchill

Development of an analytical method usually, and preferably, starts by set-ting up experimental conditions and criteria for the method. At this stage of the development, these are most often guidelines for approaching the chal-lenge. Knowledge about the sample as well as the analytes to be quantified is used to choose appropriate sample treatment, separation, detection, evaluation and validation techniques. Strategy includes the planning of the means to validate the method. With the proper strategy you will know dur-ing development what experiments to perform and the precision and accu-racy of the results necessary to advance the process. The validation part of the strategic assessment includes just that, planning, experimentation and documentation.

Time is a relevant variable and factor in almost every aspect of analytical chemistry; the time to develop a method, to maintain the representation of a specific time when sampling, the time spent with sample preparation, the time the analytes spend in the analytical column, the dwell time set for the mass spectrometric acquisition, the total time for the throughput cycle of an analytical run, the time available on the instrumentation, timing of the hy-phenation of the different parts of the analytical set-up, etc.

When dealing with strategic planning of a new method, the most impor-tant thing is to understand the intended purpose of the method; what is it supposed to monitor? In what levels? matrix? Budget and time? All neces-sary information to be able to fit the purpose. Is the purpose of the method to identify or discover compounds in the specific matrix, the criteria for quali-tative analysis should be followed. If the purpose instead aims for investi-gating, or monitoring the levels of certain compounds in the intended matrix, a quantitative approach should be performed. Furthermore, is the method supposed to have a targeted approach, knowing what compounds to quantify, going for absolute quantitation, or is it the relative responses between sam-ples or batches over time that is the purpose, so a relative quantitation should be performed.

One of the main reasons for the popularity of combining LC with MS is the ease of interfacing them on-line. Assembling the separate parts of the

20

analytical system in an on-line fashion grants the advantage of being able to automate the process of acquisition. Automation saves time, labor and money, as well as gives the advantage of running the instrumentation during hours where it is not supervised and it is also supposed to yield more reliable and reproducible results than manually operated set-ups. However, the need for precise planning is still necessary, if not even more important. In table 1 several aspects to consider are listed, just to show how complex the strategic planning is. The included aspects in the table have been chosen for analysis of molecules in a complex mixture with liquid separation and mass spec-trometry.

Table 1. Considerations regarding the strategy for method development. The vari-ables mentioned is limited to the analytical techniques mentioned in this thesis. Event Example of considerations Aim of analysis

Qualitative analysis (identification, structure elucidation) Quantitative analysis (relative- or absolute quantitation) Off-line analysis or Automated

Sampling Handling, Storage, Direct treatments Sample treatment

Precipitation (acids, organic modifier, salts), Depletion, Ultra filtration (cut-off), Derivatization (masking, ionization, separation), Dilution, Microdialysis (cut-off, selectivity, flow rate, gradient), Liquid-liquid extraction, Solid phase extraction (stationary phase, mobile phase, on-line or off-line)

Separation

Liquid chromatography (Flow rate, normal-phase or reverse phase, length of column, Temperature, inner diameter, stationary phase, particle size, pore size, mobile phase organic modifier, pH, ionic strength, buffer additive, acid, ion-pairing agent) Capillary electrophoresis (Length of capillary, Temperature, Inner diameter, Applied voltage, buffer addi-tives, organic modifier, pH, ionic strength, acids, modification of the inner wall)

Ionization

ESI, nano-ESI, Atmospheric pressure chemical ionization, Atmospheric pressure photoinitiated ionization, Matrix assisted laser desorption ionization

Mass analyzer

Quadrupole, Time-of-flight, Ion trap, Fourier transformation ion cyclo-tron resonance MS, Tandem mass spectrometry

Evaluation

External standard calibration, Internal standard calibration, Standard addition method, Derivatization techniques

Validation/Quality checks

Sensitivity, Selectivity, Precision (Repeatability, Ruggedness, Repro-ducibility), Accuracy, Linearity, Robustness, Limit of detection, limit of quantification, Column stability, Analyte stability (ambient temperature, refrigerated, in matrix, freeze-thaw cycles)

21

4. Sample treatment

Sampling must be performed with care, minimizing, or eliminating the risks of contaminations, losses due to adsorption, and/or poor stabilities. For quan-titation of species within biofluids it is of outermost importance that the measured results correspond to the situation in vivo at the time of sampling. Therefore, proper handling, sample treatment and storage are necessary to avoid problems like decomposition or unwanted chemical reactions.

There are several ways to treat a sample, all with the same goal; to trans-form the analytes of interest into the most suitable form and produce a com-patible environment for optimal detection. This might involve extracting the analyte from a complex matrix, enrich a very dilute sample to achieve higher concentration of target analyte, removing or masking interfering species or even chemically transforming the analyte to a more suitable or more easily ionizable or detectable form. All sample treatment conditions must be care-fully considered due to risk of degradation or adsorption of analyte, which will give rise to poor recovery and so disturb the accuracy and precision of the analysis. The extent and type of treatment must be chosen to suit the properties of the analyte as well as the analytical demands of every event in the analytical procedures chosen.

Different techniques have been utilized to pretreat the biological samples in the papers presented, including protein precipitation II, microdialysis IV, extraction and enrichment I. Sometimes multiple steps are necessary to achieve set goals of operation.

4.1 Protein precipitation A large portion of the molecular content in biofluids is high abundant large proteins that will complicate both the separation and the ability to ionize target analytes in the ESI. One way to get rid of these disturbing proteins is by precipitation. This is a rapid and easy to use technique that denatures the proteins, giving a precipitate which can be separated by centrifugation and later removed. Precipitation can be carried out with strong acids (perchloric acid, trichloroacetic acid or hydrochloric acid), organic modifiers miscible with water (methanol or acetonitrile), and salts, or even by boiling the sam-ple. One major drawback is that precipitation lacks selectivity, meaning that there are great risks of loosing analytes due to adsorption to precipitate or

22

sensitivity to the precipitation technique used. A more selective approach could be to use commercial depletion kits which are produced to get rid of specific compounds in biofluids. However, an obvious risk is that non-specific losses occur in those kits why they have to be used with precaution.

4.2 Microdialysis Microdialysis is a gentle membrane technique that uses a gradient for extrac-tion in aqueous solutions. In the process, two functions control the extrac-tion; first small molecules can diffuse across a semi-permeable membrane that has a certain pore size excluding large molecules, and second, the ex-traction is elevated by the use of an extraction solvent gradient across the membrane. It is possible to use microdialysis in vivo, using a probe that can be inserted into a living host, with fluid pumped through it. The dialysate, the collected fluid from the outlet of the probe, can later be analyzed by any liquid detection technique. One advantage is that protein- and enzyme free aqueous samples are obtained, on the other hand samples are often diluted and needs to be enriched.

In 1966 Bito et al. 31 performed the first static microdialysis experiment where they implanted semi-permeable sacs in dogs, and measured electro-lytes in the extracellular fluid (ECF) of the brain. The first in vivo experi-ments utilizing continuous perfusion was performed by Delgado et al. 32 in 1972 with the use of a probe that permitted collection of samples from mon-keys. This probe was later improved by Ungerstedt and Pycock in 1974 33. Microdialysis has since been a popular tool and is recognized as an estab-lished technique conducted on routine basis in many laboratories for neuro-science, pharmacokinetics and biotechnology.

Controlling the extraction efficiency is considered to be the key issue for microdialysis, and the variables include temperature, perfusion flow rate, membrane area, material, cut-off, sample and perfusion medium composi-tion. As with other analytical techniques, the aspect of time is important and in microdialysis it is the perfusion time. With a lower flow rate the equilib-rium time will be increased generating higher extraction performance.

In Paper IV we investigated the possibilities of including a solid support with specific analyte affinity in the microdialysis fluid. This enhanced the extraction efficiency, as well as the selectivity of the microdialysis. The se-lected analytes will interact with the particles acting as the solid support making those carriers (see Figure 2). When later collected it is easy to get rid of unwanted compounds and solvents without loosing analytes. By elu-tion of the analytes from their carriers, enrichment can be performed, as well as the possibility to dissolve them in a proper solvent for further analysis.

23

Figure 2. The basic principle of solid supported microdialysis.

4.3 Solid phase extraction A commonly used technique for cleaning and/or pretreating biological sam-ples is solid-phase extraction (SPE) 34, 35, which uses modified silica materi-als, similar to the ones used in reversed phase chromatography. Knowledge of the retention mechanisms is of great use. The SPE can easily be coupled on-line to liquid chromatography system, by the use of small ordinary LC columns packed with small particles. On-line coupling improves the sensi-tivity and selectivity, together with decreased risk of analyte losses, due to mistakes or contamination 36.

In Paper I SPE was used on-line in a column switching set-up to achieve on-column trapping. Column switching is the means where the direction of flow of the mobile phase is changed by valves, so that the effluent from one column is passed to another column during a defined period of time 37. Ram-steiner lists five objectives of column switching 37; (1) increased chroma-tographic retention and selectivity, (2) enrichment of sample, (3) protect sensitive detectors from contamination, (4) prevent destabilization of chro-matographic equilibrium from matrix, and (5) to achieve further objectives or a combination of objectives within one chromatographic network.

24

5. Liquid separation

Separation is the means to transform a mixture of compounds into isolated entities, preferably containing only one compound each. Hence, liquid sepa-ration (LS) is the means to perform this transformation of a dissolved liquid mixture into distinct “zones” of its individual compounds within a liquid environment. Two techniques, liquid chromatographyI,II,IV, and capillary electrophoresisIII,V has reached a dominant status within liquid separation for analytical purposes. These two separation techniques perform different ap-proaches for the separation of solutes. In LC, the liquid mixture of com-pounds is pumped through a separation media exposing the individual com-pounds to different retention mechanisms making them travel with different velocity, and CE makes the sample compounds migrate through a capillary by the force of an electric field separating them according to their individual charge and size.

5.1 Liquid chromatography “For ordinary column chromatography unknowable variables exist in-volving everything from the nature of the complex multi-site surface to the details of the method of packing the column.” - J. C. Giddings 1954 38

Chromatography is a separation process that performs separation of solutes due to mass transfer between two different phases. One of the phases is static and held in place, referred to as the stationary phase, and the other, the mobile phase which contains the dissolved sample, moves through or past this stationary phase, usually driven by a mechanical pump. The exposure of two chemically distinct phases causing the individual sample components to partition between them, trying to achieve equilibrium, is called retention. Most common, the stationary phases are either hydrophilic with a strong affinity for polar solutes, called normal-phase, or hydrophobic, attracting non-polar or weakly-polar solutes, referred to as reversed-phase. In either case, the corresponding mobile phase has a contribution of opposite polarity, making parts of the two phases each others counterpart and thus generates a selectivity factor, depending of their respective strength.

25

5.1.1 Reversed-phase chromatography Reversed-phase chromatography (RPLC) has become the more popular of the two due to the increased interest in bioanalysis, since RPLC uses sol-vents that are more compatible with biological samples and molecules. An-other aspect of the increased popularity of RPLC is the availability of a wide variety of highly pure, chemically stable packing materials, increasing the choices for separation, and thus improving the selectivity of the method. Manufacturers constantly develop new stationary phases with higher me-chanical strength, better reproducibility, faster mass transfer kinetics, higher efficiency, broader pH range stability and increased selectivity control 39. Reversed-phase chromatography is also less sensitive to polar impurities.

The separation performed with RPLC depends on hydrophobic interac-tions between solute molecules in the mobile phase and the hydrophobic surface of the stationary phase. By adsorption-desorption processes of the solutes between the two phases, retention of molecules is achieved. This generates different travel times for the solutes in the column, and thus sepa-rates them from each other. The polarity of the mobile phase is supposed to be high enough to ensure binding of the solutes to the stationary phase. On the other hand, the mobile phase polarity needs to be low enough to keep the solutes dissolved. Several retention models have been explained in the litera-ture and will be briefly mentioned. The partition model states that the reten-tion factor of analytes is correlated with their partition coefficients between water and octanol and directly affected by their solubility in the mobile phase 40, 41. The adsorption model, however, explains the interactions be-tween the solute molecules and the bonded layer ligands 42, 43, while the sol-vophobic theory is based on that the retention of a compound depends on its size and on the surface tension of the mobile phase 44.

The simplest definition of chromatographic performance (CP), could ap-proximately be explained by:

CP = aA + bB + cC + dD + … + zZ (1) Where the terms A, B, C, D, … Z, represents forces that can exist between molecules, in solution or immobilized, affecting the chromatography, and a, b, c, d, …z, are variables to each of these forces representing attraction or repulsion that contributes with positive or negative effects on the CP. Exam-ples of forces that affects the chromatographic performance includes; disper-sion forces, dipole forces, dipole-induced forces, hydrogen bond interac-tions, electrostatic interactions, charge-transfer interactions, size and shape interactions. Some of these forces are usually referred to as van der Waals interactions, and the energies represented by these forces are much smaller (0-10 kJ/mol) than the energies in chemical reactions and those involved in the formation of chemical bonds (>100 kJ/mol). Despite low energies, these

26

forces control the chromatographic interactions 45. Great advances has been made in the theory of column dynamics by some of the pioneers of chroma-tography including Giddings, van Deemter, Golay, Knox, Horvath and Guiochon 46-51.

During method development there are several things to consider that will affect the separation; column length and its inner diameter, stationary phase particle porosity, diameter and uniformity, diffusion coefficients, capacity factor, molar absorption and concentration of the solute, molar mass, viscos-ity of the mobile phase, inner diameter of connecting capillaries and the flow resistance of the column 52, 53. RPLC columns are most often packed with small, spherical, microporous silica particles with functionalized alkyl groups (for example C18) that act as the stationary phase for the separation. These particles are typically a few micrometers (2-10 μm) in diameter, pore sizes of approximately a few hundred Angstrom (100-300 Å), and these di-mensions should yield a large surface area (several hundred square meters per gram) allowing high solvent loadings. The particles should also be inert and have a regular shape with a narrow diameter range. The choice of pack-ing material, such as type and dimensions will determine the appropriate flow and injection volume 54. By reducing the particle diameter, the total surface area in the column will increase, generating an increased separation efficiency and retention of analytes, but raises the demand for higher pres-sures to maintain the velocity of the flow. Even the resolution increases by the decreased particle size, due to more stable flow through the column and that the distance between the particles decreases, making the distance that the solute must diffuse shorter.

5.1.2 Mobile phase composition The mobile phase liquid should dissolve the sample, not react with the sta-tionary phase, have low viscosity, and preferably not be expensive or toxic. Several parameters can be adjusted in the mobile phase to improve separa-tion or selectivity; choice of organic modifier and the mobile phase composi-tion (volume %), pH, ionic strength, viscosity, fluid temperature, ion-pairing agents, competing ions etc. It is extremely important to maintain sample solubility throughout the loading processes in order to avoid precipitation of the sample on the column. In RPLC the organic modifier volume ratio and strength is responsible for the attraction of hydrophobic solutes to the mobile phase, leading to decreased retention times for the analytes. Most preferred and commonly used organic modifiers are methanol (MeOH) and acetoni-trile (ACN), each with its own advantages and disadvantages. Acetonitrile has lower viscosity, which yields better kinetics generating sharper peaks, lower back pressure putting less demand on the column or permits higher flow rates. Acetonitrile also has higher elution strength and is less polar than methanol, which yields lower solvent consumption. On the other hand is

27

methanol less toxic than acetonitrile. When mixing the mobile phase the addition of the organic modifier will change the pH 55.

Gradient elution is performed to achieve adequate separation resolution for polar compounds eluting early, and to keep acceptable retention time for late eluting hydrophobic compounds, during a single analytical run. In table 2 advantages and disadvantages in using gradient elution is explained. This is most common performed by increasing the elution strength of the mobile phase making the retention of analytes decrease, by increased level of the organic modifier used for elution over a defined time. There are other ways to perform gradient elution, such as changing the temperature 56-59, flow rate 60, 61, ion-pairing reagents, pH, or ionic strength, for example.

Table 2. Comparison of advantages and disadvantages in using gradient elution compared to isocratic elution.

Preferred elution Reason Gradient elution62

A gradient span a wide range of sample retention

A gradient increase the number of peaks eluted within a certain time interval

A gradient can remove sample impurities with strong elution

A gradient is advantageous in multidimensional separation

A gradient separate large molecules with more ease

A gradient is well suited for mass spectrometry Isocratic elution63

A gradient is complicated to develop

A gradient is slower due to equilibrium time64

A gradient is difficult to transfer between laboratories65

A gradient can cause baseline problems

A gradient need more pure solvents

To evaluate whether a method for separation should be carried out with gra-dient or isocratic elution, start by making an initial slow gradient run. This initial separation will provide some important answers needed to go further with optimization; approximate retention range for the analytes, approximate elution composition for isocratic elution, separation of more polar com-pounds, identification of late-eluting compounds within a reasonable time and faster separation of the total sample advancing the method development substantially. The optimization of gradient elution is far more complex than in isocratic elution, as more parameters are involved in the selectivity 66-68.

5.1.3 Miniaturization There are and have been two main reasons for miniaturization of separation columns; the ability to achieve improved separation efficiencies and to inter-

28

face LC with MS in a way to make use of the advantages gained. Another aspect is to minimize the pollution of the environment by using smaller vol-umes of solvents, since they often are toxic. For a long time this advantage was not obvious, but when capillary columns were introduced 69, 70 chroma-tographer could see the use. As the column inner diameter (i.d.) decreased, it was observed that efficiencies increased, described as a consequence of the entire packing bed being affected by the column wall 71. The gain in reduc-ing the dimensions of the column was not only to the efficiency, but also to the flow rate, which was in the optimum range for coupling with electros-pray ionization. Decreasing the column dimension from a normal bore col-umn of 4.6 mm to a narrow bore column of 320 μm reduces the flow rate from approximately 1 mL/min to between 1-10 μL/min. In addition, the re-duced flow rate further increases the sensitivity with the use of electrospray with more than 8000 times, if the same amount of material is injected 72.

In Paper I, II and IV capillary column was used for the analytical separa-tions. All capillary columns used in the papers presented are packed in house 73. In brief, the particles are slurry mixed with trichloroethylene and injected into a reservoir connected to an empty capillary. With the outlet of the capil-lary secured with frits and screeners, the slurry is inserted into the capillary by delivering 0.1 mL/min of acetonitrile from a pump. This procedure is performed until the pressure reaches 300 bar, which is then maintained for one hour using constant pressure mode followed by a slow pressure decrease through the column. During the packing procedure, the column is kept in an ultrasonic bath for even distribution of particles. It is important to make the capillary columns long initially, to achieve as long part as possible with a uniform bed of particles. This is the part of the capillary closest to the outlet, and the uniformity decreases with the distance.

5.2 Capillary electrophoresis Electrophoresis is defined as the transport of charged molecules induced by an electric field in an electrolyte rich environment. Fundamental studies were first performed in the late 19th century, and important contributions were made by the Swedish chemist Arne Tiselius in the 1930´s 5 and further explored by Stellan Hjertén 74, as a technique used for separation of multi-charged macromolecules due to different migration times caused by their charge and size. Tiselius contributions in both electrophoresis and chroma-tography made him acknowledged as a Nobel laureate in 1948.

29

Figure 3. Schematic picture of a capillary electrophoresis system in its most simple form. Capillary electrophoresis, first reported in 1981 75, is a miniaturized form of electrophoresis that performs the migration of charged species within narrow fused silica capillaries (25-75 μm), filled with background electrolytes. The separation is achieved since the molecules will migrate at different velocities through the capillary determined by their charge-to-size ratio. The technique has become popular due to high efficiency separations, the ability to inject small volumes of crude samples and its simplicity in instrumentation. To perform CE separation, the fused silica capillary is filled with an appropriate buffer solution and the sample is introduced at the inlet. On both ends of the capillary an electrode is attached and a high potential (10 – 30 kV) is ap-plied, generating the electric field. Charged species in the sample will mi-grate towards the electrode of opposite charge with an electrophoretic veloc-ity determined by its charge-to-size ratio. By the interactions of the buffer electrolytes to the silica wall together with the applied electric field, an elec-troosmotic flow (EOF) if performed. The EOF in a bare silica capillary is usually (at pH > 4) greater than the electrophoretic velocity of the solutes, it is generally chosen to have a direction from the injection to the detection side of the capillary and will thus ensure that all species present in the sam-ple eventually passes the detector. The migration time of the analytes are determined by both the EOF and the electrophoretic velocity of each analyte respectively (see Figure 3).

30

One of the gained effects by the forces creating and upholding the EOF is that the flow profile in CE is laminar. Compared to pressure driven flow systems, like LC, which have a parabolic flow profile, the laminar profile yields much higher separation efficiencies 76.

5.2.1 Modification of the silica wall When employing capillary electrophoretic separation in bioanalysis, espe-cially in separation of positively charged molecules, like proteins and pep-tides, the analyte shows a tendency to interact strongly with the negatively charged silanols on the fused silica wall. These adsorptions are caused by hydrophobic interactions, electrostatic interactions and/or hydrogen bonding. The effect of such forces will give rise to band broadening as well as tailing, memory effects and changes in the EOF, leading to decreased repeatability and separation efficiency 77, 78. Furthermore, these interactions have been shown to increase with increasing size of solutes. There are several ways to minimize these challenges; by adjusting the property of the (1) solution buffer or (2) the capillary wall. The solution can be adjusted by changing its properties due to pH, ionic strength and buffer additives, most often to their extremes. The capillary wall on the other hand can be changed by adsorption of polymers and surfactants or by covalent bonding of a coating to the si-lanol groups.

In Paper III and V a coating called M7C4I was employed since it had been proved to generate a stable reversed EOF79. Even the stability of the coating is one of the major necessities when coating the inner surface of the capillary, since a potential bleeding of the surface will cause suppression and high background signals generated in the electrospray 80, 81. The M7C4I forms a covalent bond with ionized silanols at alkaline pH and the free dia-mine interacts with the wall by hydrogen bonding via the tertiary amino group. In addition, the quaternary amino group forma an ionic bond with neighboring dissociated silanols 82, 83. In Paper III the M7C4I coating proved useful for the analysis of small endogenous compounds, a separation that was not possible without the coating. The surface generated repeatable and stable EOF, even though the analysis was challenged by the injection of crude samples. The coating also granted necessary efficiency to analyze trace levels of the analytes, down to low nanomolar concentrations. In Paper V, the M7C4I showed the same stability for larger molecules like peptides and proteins, generating rapid separations with high efficiencies. Further-more, the coating could even handle large intact proteins up to 669 kDa, as well as separation of digested bovine serum albumin (BSA).

31

6. Ionization

“The power of an analytical methodology which would combine the benefits of two well-established and complementary techniques, high-performance liquid chromatography (HPLC) for the effective separation of solutes and mass spec-trometry (MS) for the sensitive detection and identification of unknown organic structures, is obvious.” -Patrick J. Arpino 1979 84

Approximately 30 years ago, at the dawn of interfacing liquid chromatogra-phy with mass spectrometry, some criteria were set for the requirements of an ideal LC/MS interface 84; the interface should not limit the operating conditions of the liquid chromatography separation or the mass spectromet-ric detection, it should be unaffected by different solvent systems working under different flow conditions, programmed solvent composition gradients, or possible impurities in the solvents. It should contain low dead volumes to maintain the chromatographic profile, yield a large dynamic range over sev-eral orders of magnitude for low detection limits and transfer as much of the effluent solutes as possible to the inlet of the mass spectrometer at a constant level and independent of the chemical nature of solvents and solutes. Pref-erably the performance of the interface should remain the same for the transmission of volatile, nonvolatile, and thermally labile solutes, as well as ensure that the solutes will not be chemically modified by the interface in an uncontrolled manner.

To be able to separate and detect analytes in a mass spectrometer, the ions need to be transferred into the gas-phase. There are several ways to accomplish this transition between phases and many research groups, as well as instrumental manufacturers, have developed different ion-sources for this purpose. Very few, however, shows the ability to perform this task in a re-producible and effective fashion without either uncontrolled cluster forma-tion, fragmentation, discrimination of analytes, gas-phase interactions, ad-duct formations, electrochemical reactions or other physical or chemical interferences complicating the formation of ionized analytes.

One of the most popular and commonly used ionization techniques for liquids today is electrospray ionization, which has been used through the experiments presented in this thesis. ESI has reached its dominating status for its ability to perform stable and “soft” ionization at atmospheric pressure

32

for a wide range of flow rates and solvent compositions and its ability to produce multiply charged ions of large molecules. ESI interfaces also pos-sess a versatile construction and are easily coupled to different liquid separa-tion techniques, such as LC and CE. Even with all these advantages and merits, the ESI still has much to prove before it fulfills the ideal attributes mentioned by Arpino 84. There are other common ionization techniques per-formed at atmospheric pressure (API), like atmospheric pressure chemical ionization (APCI) 85, atmospheric pressure photoinitiated ionization (APPI) 86, 87 and some variants of the electrospray ionization, like for instance; pneumatically assisted Ionspray for both microflows 88 as well as high flow rates without the need of flow splitting 89, and Thermospray (TSP) 90. Apart from these, another popular ionization technique is matrix assisted laser de-sorption ionization (MALDI) 91 which in contrast to the already mentioned techniques, is most often performed under vacuum.

In the infancy of hyphenation between liquid separation techniques and mass spectrometry, many fascinating as well as interesting suggestions were made; for instance the moving belt interface by McFadden et al. 92, which was the first commercial LC-MS interface, the direct liquid introduction (DLI) 93, which as the term implies, introduces, after splitting the effluent (1-5%) from the LC directly into the MS and the monodisperse aerosol gen-erator for introduction of liquid chromatographic effluents (MAGIC) 94, which utilizes an aerosol beam separator to protect the MS from poorly desolvated aerosol droplets. The work performed in the papers presented in this thesis I-V is exclusively performed with ESI and will therefore be the only technique covered in de-tail.

6.1 Electrospray ionization The Nobel laureate 2002, J. B. Fenn, reported the multiple charge ion forma-tion in 1988 95, just a few years after initially introducing electrospray as a source for applications with mass spectrometry, together with Yamashita 11, the so called ESMS. At the same time, independently of the work by Yama-shita and Fenn, but very similar, Alexandrov et al. 96 developed a method called “extraction of ions from solution at atmospheric pressure, EIS AP”. In this thesis the term electrospray ionization (ESI) will be used for further discussion. In electrospray ionization, the requirements for successful formation of ions depend on two steps; (1) the transfer of solutes from the liquid to the gas phase, and (2) to add a charge to these analytes if they are not already in a charged state 13, 25, 97, 98. To achieve this, an electric field is applied between

33

the liquid delivering capillary outlet and the inlet of the mass spectrometer. This electric field is generated by applying a potential of a few kilovolts (3–5 kV), positive or negative, on either the liquid capillary outlet or the mass spectrometer inlet, and the other acting as the counter electrode necessary to obtain the electric field. Electrospray existed long before its application to mass spectrometry, in the beginning of the last century Zeleny 99 showed that, as a result of the applied electric field, and at appropriate conditions, the liquid will disperse into electrically charged droplets. However, the first record of liquid surface disruption, was actually recorded by Bose already in the 18th century 100. By generation of the electric field, the ions in solution will start to migrate, creating a charge distribution, separating them by their respective charge. If a positive potential is applied on the capillary outlet, called positive ion mode, cations would migrate towards the surface and the electrostatic forces between these cations and the counter electrode (of oppo-site charge) causes a bending of the liquid surface into a cone shape, through destabilization of the surface imposed by the charge distribution, which is commonly referred to as Taylor cone 101. When the forces creating the liquid deformation overcome the surface tension of the liquid, it results in the emis-sion of charged droplets and will continue as a consequence of excess charges, performing a liquid jet stream, or spray (see Figure 4).

Figure 4. Positive mode electrospray Driven by electrostatic attraction, these charged droplets will during their path towards the counter electrode generate an electrical current. They will also be subject to continuous solvent evaporation, decreasing their size, and thus, increasing their charge per volume ratio. When this ratio becomes

34

stronger than the Rayleigh stability limit 102, the droplets will undergo cou-lombic fission, emitting smaller charged droplets. The process continues with additional evaporation, further increasing the charge to volume ratio, generating new, even smaller charged droplets, and so on, until they finally are transferred to the gas phase.

The mechanism for the production of gas-phase ions from these small charged droplets is not completely known, but there are two dominating theories, the Charge residue model (CRM) and the Ion evaporation model (IEM) (see Figure 5.). CRM was first described by Dole 14 as the successive multi-step droplet fission by solvent evaporation resulting in droplets con-taining only one ion, and by its desolvation the ion is released as a free gas-phase ion. The IEM was introduced by Iribarne and Thompson 103 and de-scribes the same droplet fission process, but with the transition from liquid to gas-phase being forced by coulombic repulsion “pulling” the charged ion from the droplet.

Figure 5. A comparison of the two dominating theories for generation of gas phase ions, the Ion evaporation model (IEM) and the Charge residue model (CRM)

6.1.1 Hyphenation of LS and MS with ESI As previously mentioned, the transition of solutes within a liquid solution to become gas phase ions, generates some interesting obstacles to be solved by the analyst. These obstacles are the result of several factors necessary for the upholding of spray stability. Modern instrumentation is becoming more so-phisticated and constantly improves the ability to form and maintain stable and efficient electrospray ionization. To achieve this, several variables needs

35

to be in control during operations; the applied spray potential, the distance to the counter electrode, nebulizing gas flow rate, and the solution flow rate, viscosity and dielectric constant 104, 105. Working with reversed-phase liquid chromatography prior to the ionization, stability of the spray is more attain-able since a large quantity of the liquid exiting the capillary outlet contains a moderately polar organic solvent such as methanol or acetonitrile, lowering the surface tension. With higher amount of aqueous solvent, the higher the surface tension, resulting in increased difficulties in adjusting electrospray variables without risk of loosing stability. As mentioned, the electrospray stability depends on a good mix of used solvents for transport and separa-tion, but also the purity and recent preparation of these solvents. Enke states that the amount of excess charge in the electrospray is directly proportional to the current measured and inversely proportional to the flow rate 106. This amount of excess charge is the limiting factor of how many ions that are detectable from the electrospray, meaning that analyte concentration below the excess charge concentration is proportional to the response. Knowledge of this excess charge concentration gives the advantage of determining the range of quantitation for a given current and flow rate 107. With analyte con-centrations above the upper limit created by the amount of excess charge, saturation occurs through competition between ionic species. This is the dynamic range limitation in the upper concentration region. Since the amount of excess charge is determined by the difference in concentration of cations and anions in solution, it can be controlled and optimized in method development. The dynamic range in electrospray ionization is linear over three to four orders of magnitude limited by saturation in the upper region and background interferences in the lower region. Several groups states that saturation generally occurs at an analyte concentration higher than 10-5 M 27,

108-111. This value varies with different organic solutions 110, limitations in ion transport into the vacuum stages of the mass spectrometer 112, the amount of excess charge and efficiency of droplet charging 27.

The hyphenation of CE-MS began in the late 1980´s 12, 113-115 and there are three major techniques to achieve it; (1) direct coupling, (2) liquid junc-tion or (3) sheath flow interface116. The latter is the most used, and also cho-sen for the experiments performed in Paper III and V. Since CE flow rates are much lower (200-400 nL/min) than the optimum flow for electrospray (1–10 μL), addition of liquid needs to be added. This is solved by the intro-duction of a so called sheath liquid, which is driven by an external pump, and coupled to the electrospray interface. This additional liquid can be used to optimize the electrospray conditions independently of the separation buffer.

36

6.1.2 Matrix effects Matrix effects are commonly a result of variations within the electrospray ionization and its efficiency due to the presence of coeluting compounds causing either suppression or enhancement effects. The matrix effect was first described by Tang and Kebarle 27 who reported decreased signal re-sponse of certain organic bases with increasing amounts of other organic bases. Some state that the effect is due to accessibility of the charged droplet surface in the electrospray 117. Others have proven the effect to be dependent on the polarity of the compounds 118, reporting that the more polar the ana-lyte is, the more susceptible to suppression is its electrospray response. Both of these cases are familiar to the point that in reversed-phase chromatogra-phy positive ion mode electrospray, the core of the droplets will be of polar nature, and as a consequence, the surface will contain less polar analytes, having the advantage of being ionized to a higher degree. A second imposed suppression effect has been reported by Sterner et al. 119, showing that larger molecules will suppress the response of smaller molecules. One approach to compensate for the suppression caused by the matrix is to have an internal standard, isotopically labeled or structural analog, coeluting with the analyte 120.

With the increased selectivity gained by using modern mass spectrome-ters, some misleading consequences has evolved through overestimation of system performance; for instance, that sample clean-up is unnecessary, minimization, or even elimination of sample separation, and that the separa-tion column has been used for sample loading purposes only 121.The solution composition will also affect the response and several reports have shown that organic analyte responses are higher in solutions with higher amount of or-ganic solvents 122-125. This effect is probably due more efficient desolvation and improved spray stability due to the decreased surface tension. Zhou and Hamburger showed that this effect is positive up to approximately 80% or-ganic modifier and above that the signal decreases most likely due to that the surface tension being to low causing difficulties in achieving spray stability 124. Volatile buffers containing a significant concentration of ionic species to act as charge carriers is commonly used to achieve the electrospray. These ionic species can be ionic analyte, added electrolyte, and/or charged products of the electrochemical reaction. In positive ion mode the charge carrier can be a cations added to the solution in the form of a neutral salt, such as am-monium acetate. Furthermore, if ionic buffer species ion-pairs with the ana-lyte to strongly, then the analyte ions may be prevented from carrying the excess charge on the droplet surfaces decreasing the electrospray response 126.

37

7. Mass spectrometry

Mass spectrometry is the means to separate and detect gas-phase ions ac-cording to their mass-to-charge ratio. The mass spectrometer is a powerful tool used in almost any analytical process today. It can be utilized for identi-fication of unknown compounds, quantitation of target species and to eluci-date the structural and chemical properties of a sample, for example. Mass spectrometry has reached a status as a necessity in a wide range of areas for a diversity of applications, for instance; forensic analysis, such as identifica-tion and quantitation of toxic compounds and drugs, like illegitimate steroids in athletes or animals used in racing, pharmaceutical analysis, to monitor the biotransformation of a prodrug in a patient, chemical analysis, the chemical composition of natural and artificial compounds, environmental analysis, such as pollutants, contaminants in fish as well as genetic damages obtained from environmental causes, and geological analysis, such as determine the age, origin and exposure history of specimen in archeological samples. The instrumentation consists of several parts; an ion source that enables the for-mation of ions from solution, a mass analyzer that focuses the ion beam and separate the ions, most often a collision cell where ions collide with gas molecules and undergo fragmentation, preferably a second filtering mass analyzer for selection, and last, a detector. Apart from the ion source, which commonly works in atmospheric pressure, the mass spectrometers works at low vacuum pressure, uphold by one or several pumps, to enable control of ions trajectories within the system. The data collected by the mass spectrometer are presented as mass spec-tra. These spectra are reports of ion abundances for each m/z value within a chosen region for a specific time frame, usually normalized to the most abundant ion. Mass spectral peak height is a representation of the number of ions present and the width represent the spread of m/z values according to the instruments resolving power. This resolving power, or resolution, of the mass spectrometer is defined as the ability of the mass spectrometer to dis-tinguish between compounds of nearly identical masses, by separating their peaks. Mass accuracy is the calculated deviation of a measured mass to its supposed value. Multiply the measured mass-to-charge value with its amount of charges and compare with its molecular weight to get the mass measurement error. Mass accuracy is generally described as parts-per-million (ppm). By presenting several mass spectra collected over a period of time, according to their total amount of detected ions per spectra (represent-

38

ing a time frame), a mass chromatogram is achieved. This is often called total ion current (TIC) and is especially interesting when the mass spec-trometer is hyphenated with a separation technique, such as liquid chroma-tography or capillary electrophoresis. A duty cycle is a measurement of the number of detected ions as a ratio to the number of total ions present. With a low duty cycle there is relatively low sensitivity and the opposite for a high duty cycle. Mean free path is the average distance that an ion or molecule will travel before colliding with another ion or molecule. This mean free path is determined by the temperature and pressure of the gas as well as the size of the molecules. This mean free path must be greater than the distance between the inlet and the detector of the mass spectrometer. A high mean free path insures predictable and reproducible fragmentation, high sensitivity and reliable mass analysis. In the mass spectrometer there is an electric field drop, ranging from the inlet and down to the detector. Due to this field drop, ions are pulled through the system by the field. This can also be used to ac-celerate ions within the mass spectrometer.

7.1 Mass analyzers The main goal of a mass analyzer is to sort and separate ions according to their mass-to-charge ratio. To perform this task there are several kinds of mass analyzers, with their respective advantages and limitations. In this the-sis quadrupole mass filter I,II,IV 127 and time-of-flight mass analyzer III,V 128 were used, but there are other mass analyzers commonly used on the market today, like for instance ion traps 129 and Fourier transform ion cyclotron resonance (FT-ICR) mass spectrometers 130.

7.1.1 Quadrupole mass analyzer The quadrupole instruments 131, 132 are the most common of the mass spec-trometers today and for the last 30 years. The quadrupole itself is a trans-mission filtering device that consists of, as the name says, four conductive poles, or rods, acting in pairs. Mass sorting with quadrupoles depends on ion motion generated from direct current (dc) and radio frequency (rf) electric fields. This field is obtained by applying and alternating positive and nega-tive dc potential, to the paired rods, and adjusting the frequency rf voltage. The created electric field, which results in a three-dimensional, time-varying field where ions “in resonance” with the field reaches the other end, while others are filtered out, due to unstable trajectory within the quadrupole. The result is that if a positive ion enters the quadrupole it will be attracted and drawn to the closest rod with a negative charge. If the potential of this rod changes before the ion reaches it, the ion will change its direction and if these potential changes continuous the ions will follow an oscillating path

39

through the quadrupole as long as it is “in resonance” with the corresponding selectivity. This allows for selection of particular ions able to pass through. By adjusting the amplitude, a stepwise selection of ions that are allowed through the quadrupole is achieved. Such a selection from low to high m/z values a mass spectrum is generated within a given period of time, often called a cycle.

The quadrupole is limited by its ability to separate ions of different m/z values due to poor resolution. Ions with larger number of charges are more difficult to separate from each other since they get closer together and are difficult to distinguish from each other. The resolving power increases with mass, due to the fact that the higher mass ions have lower velocities and spend longer time in the mass analyzer and experience more rf-cycles.

7.1.2 Time-of-flight mass analyzer The time-of-flight mass spectrometer operates by the principle that ions are accelerated into a field-free region, where they move towards the detector with a velocity determined by their m/z value. The acceleration results in that an ion has the same kinetic energy as any other ion with the same charge. By measuring the time it takes for each ion to travel a defined length, the length of the flight chamber, the m/z value can be calculated. Ions with low m/z will have higher velocity and will thus reach the detector first. Theoretical advan-tages with TOF analyzers include the ability for fast and practically simulta-neous acquisition of spectra, high resolving power, high mass accuracy and no upper mass limit 132. As, for the simultaneous acquisition of complete spectra it is limited by the cycle time which is defined by the flight time of the ions with highest m/z values.

The choice of a proper mass analyzer requires understanding of the alter-natives. The mass spectrometer should match the analytical needs and some variables to consider are; the mass range, resolving power, mass accuracy, ion transmission and scan speed.

7.2 MS/MS The term tandem mass spectrometry (MS/MS), means that at least two mass analyzers are used in the same acquisition of ions. The immediate advantage of using more than one mass analyzer is improved selectivity. In MS/MS, the two stages of mass filtering are separated by a part generating fragmenta-tion of ions passing through the first selective stage. These fragments of ions are then separated and sorted in the second stage before reaching the detector. Today, some instruments can achieve a third stage of selection, MS/MS/MS or MS3, or even higher order of selectivity, thus requiring n amounts of analyzing steps, usually abbreviated MSn.

40

Fragmentation of charged molecules can occur as a consequence of ac-celeration and collision with surrounding gas molecules. This kind of frag-mentation is often called collision induced dissociation 133 and can be ob-tained, either up-front of the MS or in the collision cell in between two mass analyzers 134. Up-front fragmentation can be used for declustering in the region between the ion source and the vacuum inside the mass spectrometer. However, it is difficult to control this fragmentation and unwanted fragmen-tations often occur for small labile molecules. When fragmentation is per-formed in a collision cell it can be selective and optimized through control of selected parent ion(s), collision energy accelerating the molecules, amount of collision gas and choice of collision gas.

Table 3. Most common scan types available Technique Scan type Description MS Full scan Scanning for analytes over a defined mass window

Single Ion Monitoring (SIM) Selection of a distinct, narrow mass interval represent-ing a single compound

MS/MS

Product Ion scan (PI) Selection of specific parent ion in the first mass analyzer, fragmentation, followed by scanning for generated fragments in the second mass analyzer

Precursor Ion scan or Parent ion scan Selection of specific fragment ion in the second mass analyzer due and followed by scanning for a potential parent ion in the first mass analyzer

Multiple Reaction Monitoring (MRM) Selection of specific parent ion in the first mass analyzer, fragmentation and selection of a specific fragment ion in the second mass analyzer

In Paper I, II and IV a triple quadrupole instrument was used for the analy-sis. In a triple quadrupole instrument, three quadrupoles are arranged in a linear fashion. The first (Q1) and third quadrupole (Q3) are used for mass selection, operating in both dc and rf potentials, while the second quadrupole (Q2) is used for fragmentation and operates only with fixed rf voltage. In Table 3 some common scan types that can be performed using a triple quad-rupole instrument are mentioned and described briefly.

The use of multiple reaction monitoring will dramatically increase the molecular specificity of the quantitative method 135. This is due to the high selectivity yielding almost no background signal, generating higher signal-to-noise ratio (S/N).

41

8. Quantitation

“True genius resides in the capacity for evaluation of uncertain, hazard-ous, and conflicting information.” - Winston Churchill

Quantitative studies can be performed with two distinct approaches; either relative quantitation or absolute quantitation. The first, aim to study the ef-fect, may it be increase or decrease, of one or more analytes over a period of time. The acquired data are compared to each other for the interpretation and no definite levels can be concluded. The other, the absolute quantitation, aims to study the amount of one or more analytes at a given time. The ac-quired data are compared to those of known amounts for the interpretation, and specific concentrations can be quantified. Of the papers included in this thesis that presents quantitative resultsI,II,IV, Paper I and II uses absolute quantitation and Paper IV uses relative quantitation.

8.1 Calibration methods The aim in absolute quantitative mass spectrometry is to correlate the ac-quired response with the quantity of the analyte of interest. There are several ways to achieve this calibration and the most common ones are external standard calibration, internal standard calibrationI and standard addition methodII.

8.1.1 External standard calibration method The term external comes from that the quantitative results acquired are cor-related to known results from injected standards acquired separate from the unknown samples. Standards are prepared with known amounts at increasing concentrations. A fixed volume of these standards are injected onto the ana-lytical system and the acquired peak areas are integrated. By plotting the response area against its respective concentration, a calibration curve can be achieved (see Figure 6A). From this graph a mathematical relationship can be calculated giving the response (Ranalyte) as a function (f(x)) of the concen-tration of injected samples (canalyte) (Equation 2).

42

Ranalyte = f(canalyte) (2)

By the injection of a sample with unknown concentration with the same vol-ume as the injected standards, the measured response will follow the func-tion obtained. Since the volumes are the same, there will be proportionality between the acquired response and its quantity as long as the function is linear and that the acquired response is within the range of the standards. A value outside of the calibration range is not proved to follow the response function and should not be calibrated according to the model.

The external standard calibration method should only be used when there are almost no variations in instrument performance or sample preparation. Common problems with this method are that variations in the injections are not compensated for and that it is difficult to prepare a relevant blank sample that mimics the matrix.

8.1.2 Internal standard calibration method The term internal comes from that quantitative results acquired are corre-lated to fixed amounts of internal standard (IS) added to the sample as a ref-erence. Several standards should be prepared with increasing concentrations, and fixed amount of IS added. After the injection of spiked standards into the analytical system the peaks are integrated. A calibration curve (see Fig-ure 6B.) can be achieved by plotting the ratio between the analyte peak area and its respective IS peak area on the Y-axis against the concentration of the analyte on the X-axis. A mathematical relationship can be calculated to de-scribe the response ratio (Ranalyte/RIS) as a function (f(x)) of analyte concen-tration (canalyte) (Equation 3).

Ranalyte/RIS = f(canalyte) (3)

The response function needs to be linear for use in quantitation. To calculate the concentration of unknown samples, the response area from the analyte peak is divided with the peak area of the measured IS and the ratio gives the quantity according to the response function from Equation 3.

The IS should be chosen due to its similarity in chemical and physical properties to the analyte (see Table 4). This is desired so that it will experi-ence the same treatment (ideally at the same time) as the analyte within the analytical system. The more similar the IS, the larger the possibility that both compounds experience the same losses in the system and the ratio between the compounds will remain unchanged during the analysis giving the advan-tage of neglecting these variations. Examples of variations compensated by the use of IS are those inflicted by sample preparation, like dilution or con-centration of the sample due to evaporation and adsorption, and instrumental

43

variations, like injection volume and matrix effects 136, 137. To make the most of the IS, it should be added to the sample as early as possible in the analyti-cal chain to obtain maximum accuracy.

Figure 6. Calibration models used for linear regression quantitation; (A) external standard calibration and, (B) internal standard calibration. Table 4. Desired properties of an internal standard used for quantitation The ideal IS should .. .. be pure

.. have similar chemical and physical properties as the analyte of interest

.. have similar retention time as the analyte in the separation

.. not be naturally present in the sample

.. not be contaminated by the analyte

.. have proven stability for the analytical conditions

.. be present at a concentration that lies within the linear range when added to the sample

.. at least have a difference of 4 in mass from the analyte

There are at least two approaches to find a suitable internal standard; a com-pound related to the analyte, or a structural analogue labeled with stable isotopes (2H, 13C, 15N, 18O). The latter is preferred since it will experience the same treatment as the analyte in the analytical process to a higher extent. A isotopically labeled IS will react in the same way to extraction and sample treatment, have almost the same retention in the separation 138, experience the same ionization conditions in the electrospray, and undergo the same fragmentation conditions in the MS/MS. There is a possibility that analyte

44

and deuterium labeled IS analogue will differ in retention time, probably due to decreased hydrophobicity 139. In addition, co-eluting IS may show unex-pected behavior in the ionization, which generates enhancement and sup-pression effects 140, 141.

In Paper I we used internal standards labeled with stable isotopes of all four analytes. Two of the IS differed with 4 Da, while the other two differed with 7 Da. The IS in these experiments did not show any sign to differ in retention, compared to its analyte. An important issue in dealing with IS is that the total amount of sample increase with each addition of IS. In Paper I, where four analytes and four IS were used, together with a complex sample composition, it was extremely necessary to monitor saturation or enhance-ment effects in the ionization due to co elution. The addition of IS should be within the linear range and amounts used should be proportional to the num-ber of IS used.

8.1.3 Method of standard addition Working with quantitation of analytes in biological fluids is challenging since it is difficult to find suitable blanks. However, the method of standard addition compensates for these matrices. In this method, the sample is di-vided into several (at least 3) sub samples, and all except one is spiked with increasing amounts of a standard solution. The level of added standard is usually within 25 – 100 % of expected analyte concentration and in any case, it must be within the linear range. The response (peak area) of all sub sam-ples are plotted against their added amount of standard, where the sample not spiked is referred to as zero (see Figure 7.). The mathematical correlation of added concentrations still follows Equation 2, with the addition that extrapo-lation is performed. This extrapolation is performed by setting the response (Ranalyte) to zero in Equation 2, and calculate the concentration.

Since extrapolation is performed, the method is considered less precise than the other calibration methods that use interpolation. Another drawback with the method is the demand for larger amounts of sample, which could complicate the use when dealing with animal models. By combining stan-dard addition with further addition of a suitable internal standard increases the accuracy in results due to variations in sample preparation or instrumen-tal error.

45

Figure 7. The extrapolation of the linear relationship between samples in standard addition method for quantitation. At the point where the response is zero, it is possi-ble to calculate the amount of analyte initially present in the sample. The quantitation in Paper II is performed using standard addition. If the amount of sample volume is sufficient this method performs good quantita-tion of analytes in complex matrices. The sample volume is by far the most limiting in this method, especially if some sample treatment is performed as well. In Paper II protein precipitation was used to get rid of the protein in the samples. The volume also limits the amount of sub samples possible for the calibration.

46

9. Validation

Validation is the means to assure a minimum level of accepted quality that a method will generate according to its intended purpose. It is supposed to guarantee evidence that the interpreted quantitative conclusion is correct and that the results are obtained accordingly with in advance set performance criteria. To achieve the desired repeatability of a method there are some considerations during its development; linearity in the calibration model for the range where the target analyte has its expected concentration, precise, accurate, known and controlled variance of response, selectivity that guaran-tees detection of expected analytes, ability to detect as low concentrations as necessary for the application, stability of analytes in sample matrix before and during the analysis and, of course, robustness of results, generating good quality data for analyses performed over a long period of time for a given instrumental set-up, even with small differences in pH, ionic strength, mo-bile phase composition, ambient temperature or chromatographic retention in different laboratory environments.

Table 5. Definitions and terminology in validation Parameter Description Sensitivity

The change in signal response due to a change of analyte abundance in the sample

Selectivity

The ability of an analytical method to find (and quantify) the analyte of interest in the presence of other compounds

Precision Repeatability

A measure of closeness of results obtained with repeated analysis of identical samples by the same conditions

Ruggedness

A measure of closeness of results obtained with repeated analysis of identical samples with variations of within-laboratory conditions

Reproducibility

A measure of closeness of results obtained with repeated analysis of identical samples with variations of between-laboratory conditions

Accuracy A measure of closeness of obtained results to their true values Linearity

The ability to obtain results which are directly proportional to the analyte abundance in a sample

Robustness

The capacity to tolerate, or remain unaffected, by deliberately induced variations of conditions (within-laboratory)

Limit of Detection (LOD) The lowest concentration possible to detect (S/N > 3) within the limits of precision and accuracy

Limit of Detection (LOD) The lowest concentration possible to quantitate (S/N > 10) within the limits of precision and accuracy

47

Validation of an analytical method should not be considered a one-time test just to show the reliability of the method for a referee or an agency. On the contrary, validation tests should be performed continuously for as long as the method is in use, or throughout the lifetime of a drug product. The method should be reproducible in such a case that it can be performed by another operator, with an equivalent equipment, in another location or laboratory and through time 142. Quality checks (QC), or quality assurance (QA) tests should be performed continuously during method development, as during routine analysis. Documentation of system and method performance is extremely valuable for the discovery of complications in the analysis or when re-validating the method. Re-validation should be performed at all times when some of the variables affecting the analysis are changed.

Table 6. How-to perform validation of methods Parameter Performed experiments Sensitivity

Create a calibration curve within the linear range with at least 5 concen-trations (n=3). Plot concentration against response and the slope corre-spond to the sensitivity

Selectivity

Inject blank matrix and identify; (1) co eluting compounds at the same retention time as the analytes, (2) if response is gained corresponding to analyte

Precision Repeatability

Make several injections (n>3) in the low, medium and high concentra-tion regions of the linear range

Ruggedness

Make several injections (n>3) in the low, medium and high concentra-tion regions of the linear range with the within-laboratory variations, like another operator or instrumentation

Reproducibility

Make several injections (n>3) in the low, medium and high concentra-tion regions of the linear range at a different laboratory, by another operator and instrumentation

Accuracy

Comparison of calibration curves with identical standards at different occasions. Accuracy tests can be performed as either (1) short-term, at consecutive days, or as (2) long-term accuracy by comparison of calibra-tions obtained with several weeks in between.

Linearity

Create a calibration curve within the linear range with at least 5 concen-trations (n=3). The obtained results should be proportional to the in-crease in abundance (R2>0.999)

Robustness

Make a small change in the method, inject samples and compare against previously obtained results.

Limit of Detection (LOD) Inject samples of decreasing concentration until S/N close to (but still higher than) 3 within the limits of precision and accuracy. No extrapola-tion should be performed.

Limit of Detection (LOD) Inject samples of decreasing concentration until S/N close to (but still higher than) 10 within the limits of precision and accuracy. No extrapo-lation should be performed.

Depending on its intended purpose certain aspects of the method should be evaluated before and during analysis, like for instance linearity, accuracy,

48

precision, sensitivity, selectivity, limit of quantitation (LOQ) and ruggedness 143. Table 5 presents the definitions for different validation approaches nec-essary to prove the method and in Table 6 there are suggestions on how such tests can be performed. In brief, preparation of standards with increasing concentrations should be performed in the intended sample matrix. The amount of added standard should not exceed 5% (v/v) of the total volume. A calibration curve can be produced by the injection of prepared standards and an injected blank matrix sample. From the calibration curve several evi-dences can be observed; the linearity over the concentration interval, the repeatability of responses at each concentration and the respective retention times, sensitivity in response due to the slope of the linear interval, accuracy as a comparison with previous calibration curves, and potential interfering co eluting species, or memory effects, by the results of the blank matrix sample. All injections should be performed in random order with at least three repli-cates of each concentration.

To be able to test the method for robustness it should be provoked by de-liberately change one or more variables. Robustness testing are not yet man-datory according to ICH 143, but the US Food and Drug Administration (FDA) demands it for registration of drugs in the USA 142. The analytical column, by itself, represents several variables that are crucial for the robust-ness of the method; resolution, capacity factor, tailing factor and the analysis time 144. Robustness testing has for a long time been considered a validation procedure performed when the method is developed. However, some con-sider it a test to be performed during development and optimization phase 145, while others consider it a development test 146. Examples of variables to change when using LC includes; mobile phase composition, like pH, ionic strength, buffer additive, acids, organic modifier, column attributes like, type, batch, manufacturer and the temperature over the column, flow rate and if gradient is used the initial and final composition as well as the gradi-ent slope. In the case of CE as chosen separation; Buffer composition, like pH, ionic strength, buffer additives, acids, organic modifier, capillary attrib-utes like batch and manufacturer, temperature over the capillary, applied field strength, rinse time and injection pressure, for example.

49

10. Concluding remarks

The key to a successful analysis lies within the control of the analytical process, as well as, optimization and validation of all the steps involved. Without the proper evidence of sufficient analytical performance the inter-pretation of results will be challenged or even give rise to misleading results. It is also important to keep in mind that the priority of an analytical method is to fit the purpose, meaning that the time spent, costs and effort put into method development should be in line with the scope of interest from the researcher or the customer.

This thesis demonstrates enhancements made in several of the steps in-volved in the analytical chain. These ranges from introducing a solid support to microdialysis, in the form of added particles of a specific affinity for im-proved selectivity, to complete automated analytical systems with several separation steps, proved to handle complex samples with a wide range in analyte polarity. Additionally, with the deactivation of the inner silica wall, it has been shown that capillary electrophoretic separations were possible for analytes such as small endogenous polar compounds and up to huge proteins with a mass of 669 kDa.

The aim for all projects has been to accomplish methods that can handle complex samples with good repeatability and accuracy of the results. In the papers presented, several different biological matrices have been investi-gated for this purpose; pig liver extract, blood plasma and cerebrospinal fluid. In addition, elevated complexity due to a wide range of target polari-ties, ranging from polar to less polar, has also been investigated since it chal-lenges the selectivity in the methods used.

Another aim has been to develop methods with the ability to quantify the analytes present in the samples. Several calibration methods has been used for this purpose, from relative recovery investigations, to absolute quantita-tion with standard addition, and the use of internal standards labeled with stable isotopes, making them structural analogues. Even some effort to achieve quantitation with capillary electrophoresis was tried, but limited by the loading capacity of the technique. With the addition of an extraction step in-line would possibly overcome this limitation.

50

10.1 Future aspects

“The best way to predict the future is to invent it” - Alan Kay

I think future bioanalysis will make use of electrodriven separation tech-niques, like capillary electrophoresis, if absolute quantitation can be per-formed. The ability to handle small volumes of crude samples is extremely tempting, as well as fast separations, giving high throughput, with narrow efficient peaks with great sensitivity in electrospray ionization. By assem-bling several techniques in-line, for multiple step analysis, like extraction, enrichment or cleaning, the technique will be able to handle larger sample loadings, elevating the performance and making quantitation possible. Elec-trically driven flow is also more suitable than pumped for miniaturized ana-lytical systems like chips or μTAS, which will be more routinely, used in future analysis.

51

11. Acknowledgements

In only five years, I have crossed path with many people that has influenced and impressed me, and I have experienced more than I had ever imagined. For this I will be forever grateful. Supervisors and mentors: Jonas, for granting me the opportunity to play around in the lab and believing in my crazy ideas even if they sometimes have been lacking in focus, Magnus, for being there when I needed it the most, Ardeshir, my first mentor and my scientific idol, and a special thanks to Per, for always taking the time for discussions and for giving valuable scientific advises.

Collaborators: Ardeshir, Cecilia, Andreas, Joel, Niklas, Emma, Anna, Nina, Erik A, Erik S, Anisa and Magnus Support: Barbro, Bosse and Yngve, for all the administrative and technical assistance. Officemates: Leif, Gustav and Erik, for keeping up the mood and for all the fun “scientific” discussions. Collegues: Past and present colleagues at Analysen in Uppsala, as well as in Umeå. Friends: Torpeder, Barbafarsor, Innebandyfolk, Gourmetklubben, Smurfer, Lextalionis smurfer, Enola crew and Discgolfers My Family: Mom and Dad, for your endless support and love, my brother, Håkan, for encouragement, inspiration and mentorship, and Märtha, for keeping me above the surface and feeding me with love and positivism, I am lucky to be loved by you! Tack! Björn Arvidsson Uppsala 19/3 2008

52

12. Summary in Swedish

Analytisk kemi inbegriper läran och vetenskapen om att utveckla och för-bättra analysmetoder för att upptäcka och identifiera såväl artificiella som naturligt förekommande ämnen i vår omgivande natur (hos växter och i sjö-ar), såväl som inom våra egna kroppar (i våra vävnader och kroppsvätskor). Analytisk kemi är ett brett område som överbrygger ämnesgränser och åter-finns inom många discipliner; såsom biologi, farmaci, medicin och självfal-let alla de kemiska områdena. En analytisk problemställning kan angripas på (minst) två olika sätt; kvalitativt eller kvantitativt. Kvalitativ analys söker närvaron, och identiteten hos en eller flera molekyler i ett studerat prov. Kvantitativ analys söker den relativa eller absoluta mängden av ett eller flera ämnen i ett prov. Vid utveckling av nya analysmetoder följer man en fler-stegsprocess som kallas den analytiska kedjan. Denna består av ett flertal moment som alla bidrar till resultatet och den slutliga tolkningen av en genomförd analys. Inom denna process är samtliga av dessa steg lika värde-fulla för att slutresultatet skall bli så sanningsenligt som möjligt. De om-nämnda stegen är försöksplanering, provtagning, hantering och förändring av provet, separation av olika provkomponenter, detektion av dessa kompo-nenter, utvärdering av komponenternas respons, och slutligen tolkning av insamlade resultat. Dessa steg följer varandra i ovan nämnda ordning och i samtliga steg finns möjligheten och risken att misstag begås, något som i slutändan försvårar, om inte omöjliggör tolkningen av (de insamlade) resul-taten. Med en noggrann försöksplanering, stor kunskap och en rejäl portion erfarenhet kan dessa eventuella misstag eller felaktiga handhavanden kon-trolleras, minimeras, och helst av allt elimineras. På samma sätt eftersträvar den analytiska kemisten att utveckla analysmetoder som ger repeterbara resultat och tål små förändringar av metoden utan att därigenom generera osanna tolkningar. Gärna skall metoderna också erbjuda möjligheten att kunna detektera och haltbestämma ämnen som förekommer i mycket små mängder, i prover som också innehåller stora mängder av andra ämnen som komplicerar analysen med sin närvaro. Igår frågade vi oss om vi ur våra analyser erhållit de korrekta svaren. Idag, med tillgång till sofistikerad och känslig teknologi, avancerad mjukvara som samlar och jämför stora datamängder och en god kunskap om proverna vi analyserar, handlar det mer om huruvida vi har ställt de korrekta frågorna.

53

I denna avhandling presenteras metodutveckling och forskningsresultat som bidragit till förbättrade möjligheter att haltbestämma låga halter av moleky-ler i mänskliga kroppsvätskor. De molekyler som analyserats kan vara an-tingen endogena, vilket betyder att de är skapade i kroppen, eller exogena, som betyder att molekylerna är aktiva i kroppen, men har sitt ursprung från någon annanstans, som exempelvis läkemedel eller droger.

I artikel I var syftet att skapa en automatiserad metod, där alla delarna inom den analytiska kedjan skulle vara ihopkopplade och synkroniserade med varandra. Fördelar med att sammankoppla flera analyssteg är exempel-vis minskad risk för att förlora material från provet inom systemet, samt kontrollerade vätskeflöden, volymer och tider för varje moment som ger ökad repeterbarhet av analysmetoden. Som modellsystem under utveckling-en av analysmetoden valdes ett antikoagulerande läkemedel, ximelagatran, två av dess intermediära nedbrytningsprodukter, hydroxyl-melagatran och etyl-melagatran, samt dess aktiva substans, melagatran. Detta modellsystem var speciellt utvalt för att ge en stor utmaning för metodutvecklingen, då dessa fyra molekyler har mycket olika polaritet parvis, något som försvårar separation och detektion. Dessutom, som med alla nedbrytningsprodukter, finns möjligheten till att något inom analysförfarandet påverkar någon/några av molekylerna så pass att dom bryts ned till att likna någon av de andra molekylerna strukturmässigt. Således försvåras analysen av att molekylerna konstant måste kontrolleras för att säkerställa dess sanna ursprung. Metoden visade sig framgångsrik och gav goda möjligheter till att detektera och halt-bestämma låga halter av samtliga fyra molekyler i ett upparbetat leverprov som använts för att simulera de naturliga förhållandena i människokroppen vid intag av läkemedel.

I artikel II förädlades en metod som tidigare skapats inom forskargrup-pen, för att analysera och haltbestämma aminosyran tryptofan och några av dess nedbrytningsprodukter i prover erhållna från ett ungerskt sjukhus. Pro-verna var tagna från patienter med multipel skleros (MS) vilka behandlats med en ny medicinering, som läkarna tror ger en minskad nedbrytning av nervbanornas myelinskidor, dvs. det normala sjukdomsförfarandet för MS-sjuka. Tyvärr har en förmodad bieffekt varit att man upptäckt en ökad de-pression hos dom behandlade patienterna. Tryptofan och dess nedbrytnings-produkter har tidigare bevisats bidra till depression vid minskad produktion av en av dess nedbrytningsprodukter, serotonin. Detta ämne är också inblan-dat i uppkomsten av andra neurologiska störningar, som exempelvis ätstör-ningar, PMS, schizofreni, epilepsi och Alzheimers sjukdom. Genom att stu-dera metabolismen och dess förhållanden hos patienter som medicineras för första gången och patienter som behandlas under ett års tid och jämföra des-sa med friska kontrollindivider hoppades vi kunna utröna huruvida seroto-ninnivåerna minskat pga av den nya medicineringen. Metoden visade sig fungera mycket bra och alla prover kunde analyseras och nedbrytningspro-dukterna haltbestämmas med god precision och noggrannhet.

54

I artikel III skapades en metod för detektion av specifika molekyler ur små volymer blod utan att först rena provet, något som brukar visa sig ödes-digert för analysmetoder då det orenade provet kan orsaka okontrollerade flöden och sätta igen kapillärerna. För att säkerställa och bevisa metodens duglighet jämfördes den med vattenprover preparerade med tillsatser av samma molekyler för att urskilja om metoden genererade samma resultat oavsett provsammansättning. Vid separationsanalys av extremt små volymer används ofta en teknik som kallas kapillärelektrofores, och som namnet an-tyder utnyttjas elektrofores för att driva molekylerna i en tunn kapillär mot detektorn. Kapillären som nyttjas som separationsmedia har en inner diame-ter motsvarande tjockleken på ett hårstrå, vilket gör att med en längd på un-gefär en halvmeter så är den totala volymen inte mer än 4 mikroliter, vilket är ungefär samma volym som i en droppe vatten. Elektrofores är en separa-tionsteknik som genom att applicera en stor potential, mellan 5 000 och 30 000 volt, över en kapillär fylld med separationsmedia och provet, skapar ett elektriskt fält över kapillären. Detta leder i sin tur till att molekylerna vill förflytta sig mot den sida av det elektriska fältet som har motsatt laddning som de själva, med en hastighet som påverkas av kraften på deras laddning och storlek. På grund av dessa rörelser kan man således separera molekyler-na i provet och detektera dem. Eftersom man i separationsmediumet har en stor mängd andra elektrolyter som också är laddade, skapas en kraft som kallas elektroosmotiskt flöde (EOF) och som driver hela vätskan åt en rikt-ning bestämd av valet av dess elektrolyter. Detta motverkar att molekyler i provet som har samma laddning som elektroden vid detektorn skulle gå för-lorade genom att de vandrat bort från detektorn. Metoden visade prov på mycket god känslighet och påverkades inte av att blodprovet inte renades innan analys.

I artikel IV undersöktes möjligheterna till att ytterligare förbättra en ex-traktionsteknik som sedan ett tag tillbaka använts rutinmässigt inom neuro-vetenskapliga och farmakokinetiska områden. Tekniken som förädlades kal-las mikrodialys och är en miniatyrisering av den välkända dialystekniken som används inom sjukvården för att rena blodet på patienter med nedsatt njurfunktion. Vid mikrodialys förs en sond in i en vävnad och en extrak-tionsvätska pumpas kontinuerligt genom denna sond och fångar upp moleky-ler som selektivt passerar från vätskan i vävnaderna genom ett membran och följer extraktionsvätskan ut ur vävnaden för vidare analys. Selektiviteten i tekniken ligger i valet av membran på sonden, samt extraktionsstyrkan på lösningen som pumpas genom systemet. För att ytterligare förbättra denna teknik tillsattes partiklar till extraktionslösningen i syfte att höja selektivite-ten för specifika molekyler genom att nyttja deras affinitet för ett specifikt material. Största nackdelen med mikrodialys har tidigare varit att provlös-ningarna blir utspädda, vilket försvårar senare detektion, men i detta fall med partiklarna som bärare av de molekyler man vill detektera är det möjligt att låta avdunsta provet utan att förlora sina molekyler. Man kan också enkelt

55

skölja bort oönskade molekyler som tagit sig genom membranet men inte har affinitet för partiklarna, också utan att förlora de molekyler man är intresse-rad av. Metoden visade en stor förbättring av extraktionen och lovar gott som uppreningsteknik.

I artikel V, användes samma separationsteknik som i artikel III, kapillär-elektrofores. Men i denna artikel var syftet att skapa tillräcklig separation av proteiner och peptider istället för små molekyler som i den förra artikeln. Dessutom har det tidigare visat sig svårt att genomföra denna separation av proteiner och peptider med kapillärelektrofores då dessa molekyler gärna vill interagera starkt med kapillärens inre yta. För att förhindra denna okontrol-lerade och oönskade interaktion behöver ytan på insidan av kapillären för-ändras. I artikel II hade redan en sådan förändring genomförts och denna applicerades även här. Normalt är ytan i en kapillär negativt laddad vid lågt pH vilket vanligtvis brukas för separation av dessa peptider och proteiner, och således bör då ytan antingen ”avladdas”, eller täckas genom interaktion med något som är positiv laddat, något som kallas deaktivering. Det senare visade sig fungera alldeles utmärkt och via ett samarbete med en italiensk forskargrupp som utvecklat en molekyl för denna typ av deaktivering lycka-des vi bevisa att den fungerade för detta ändamål i vår instrumentuppsätt-ning. Metoden visade god separation inom kort analystid för peptider och proteiner, samt god kompatibilitet med masspektrometri. Samtliga arbeten inkluderade i denna avhandling har haft som mål att kunna haltbestämma molekyler som är specifikt intressant för olika medicinska och kliniska intressen. Då dessa molekyler förekommer i biologiska vätskor, som normalt försvårar analyser, har utgångspunkten varit att även inkludera detta i målsättningen, att analysmetoden ska tåla hantering av provet och ändå generera repeterbara resultat.

56

References

(1) De Hoffman, E.; Stroobant, V. Mass spectrometry - Principles and Applications

2002. (2) Thomson, J. Rays of Positive Electricity and Their Application to Chemical Analy-

sis, Longmans Green, London 1913. (3) Cameron, A. E.; Eggers, D. F. Review of Scientific Instruments 1948, 19, 605-607. (4) Stephens, W. E. Physical Review 1946, 69, 691-691. (5) Tiselius, A. Transactions of the Faraday Society 1937, 33, 524-535. (6) Jennings, K. Int. J. Mass Spectrom. Ion Phys. 1968, 1, 227. (7) McLafferty. FW; Bryce, T. A. Chemical Communications 1967, 1215. (8) Cooks, R.; Beynon, J.; Caprioli, R. Metastable Ions, Elsevier, New York 1973. (9) Arpino, P.; Baldwin, M. A.; McLafferty. FW Biomedical Mass Spectrometry 1974,

1, 80-82. (10) Yost, R. A.; Enke, C. G. Journal of the American Chemical Society 1978, 100,

2274-2275. (11) Yamashita, M.; Fenn, J. B. Journal of Physical Chemistry 1984, 88, 4451-4459. (12) Olivares, J. A.; Nguyen, N. T.; Yonker, C. R.; Smith, R. D. Analytical Chemistry

1987, 59, 1230-1232. (13) Fenn, J. B.; Mann, M.; Meng, C. K.; Wong, S. F.; Whitehouse, C. M. Science 1989,

246, 64-71. (14) Dole, M.; Mack, L. L.; Hines, R. L. Journal of Chemical Physics 1968, 49, 2240. (15) Wilm, M.; Mann, M. Proc. R. Soc. 42nd ASMS Conference, Chicago, IL 1994, 770. (16) Sakurai, K.; Zou, J. P.; Tschetter, J. R.; Ward, J. M.; Shearer, G. M. Journal of

Neuroimmunology 2002, 129, 186-196. (17) Baran, H.; Jellinger, K.; Deecke, L. Journal of Neural Transmission 1999, 106,

165-181. (18) Miranda, A. F.; Sutton, M. A.; Beninger, R. J.; Jhamandas, K.; Boegman, R. J.

Neuroscience Letters 1999, 262, 81-84. (19) Ogawa, T.; Matson, W. R.; Beal, M. F.; Myers, R. H.; Bird, E. D.; Milbury, P.;

Saso, S. Neurology 1992, 42, 1702-1706. (20) Erhardt, S.; Blennow, K.; Nordin, C.; Skogh, E.; Lindstrom, L. H.; Engberg, G.

Neuroscience Letters 2001, 313, 96-98. (21) Yamamoto, H.; Murakami, H.; Horiguchi, K.; Egawa, B. Brain & Development

1995, 17, 327-329. (22) Baran, H.; Cairns, N.; Lubec, B.; Lubec, G. Life Sciences 1996, 58, 1891-1899. (23) Beal, M. F.; Matson, W. R.; Swartz, K. J.; Gamache, P. H.; Bird, E. D. Journal of

Neurochemistry 1990, 55, 1327-1339. (24) Orlikov, A. B.; Prakhye, I. B.; Ryzov, I. V. Biological Psychiatry 1994, 36, 97-102.

57

(25) Covey, T. R.; Lee, E. D.; Henion, J. D. Analytical Chemistry 1986, 58, 2453-2460. (26) Huang, E. C.; Wachs, T.; Conboy, J. J.; Henion, J. D. Analytical Chemistry 1990,

62, A713. (27) Tang, L.; Kebarle, P. Analytical Chemistry 1993, 65, 3654-3668. (28) Kostiainen, R.; Kotiaho, T.; Kuuranne, T.; Auriola, S. Journal of Mass Spectrome-

try 2003, 38, 357-372. (29) Jemal, M.; Xia, Y. Q. Rapid Communications in Mass Spectrometry 1999, 13, 97-

106. (30) Matuszewski, B. K.; Chavez-Eng, C. M.; Constanzer, M. L. Journal of Chromatog-

raphy B-Analytical Technologies in the Biomedical and Life Sciences 1998, 716, 195-208.

(31) Bito, L.; Davson, H.; Levin, E.; Murray, M.; Snider, N. Journal of Neurochemistry 1966, 13, 1057

(32) Delgado, J. M. R.; Defeudis, F. V.; Roth, R. H.; Ryugo, D. K.; Mitruka, B. M. Archives Internationales De Pharmacodynamie Et De Therapie 1972, 198, 9

(33) Ungerstedt. U; Pycock, C. Bulletin Der Schweizerischen Akademie Der Medizini-schen Wissenschaften 1974, 30, 44-55.

(34) Hopfgartner, G.; Husser, C.; Zell, M. Therapeutic Drug Monitoring 2002, 24, 134-143.

(35) Jemal, M. Biomedical Chromatography 2000, 14, 422-429. (36) Niederlander, H. A. G.; Koster, E. H. M.; Hilhorst, M. J.; Metting, H. J.; Eilders,

M.; Ooms, B.; de Jong, G. J. Journal of Chromatography B-Analytical Technolo-gies in the Biomedical and Life Sciences 2006, 834, 98-107.

(37) Ramsteiner, K. A. Journal of Chromatography 1988, 456, 3-20. (38) Giddings, J. C.; Eyring, H. Journal of Physical Chemistry 1955, 59, 416-421. (39) Gritti, F.; Guiochon, G. Journal of Chromatography A 2005, 1099, 1-42. (40) Dill, K. A. Journal of Physical Chemistry 1987, 91, 1980-1988. (41) Dorsey, J. G.; Dill, K. A. Chemical Reviews 1989, 89, 331-346. (42) Tchapla, A.; Colin, H.; Guiochon, G. Analytical Chemistry 1984, 56, 621-625. (43) Tchapla, A.; Heron, S.; Colin, H.; Guiochon, G. Analytical Chemistry 1988, 60,

1443-1448. (44) Horvath, C.; Melander, W.; Molnar, I. Journal of Chromatography 1976, 125, 129-

156. (45) Vitha, M.; Carr, P. W. Journal of Chromatography A 2006, 1126, 143-194. (46) Vandeemter, J. J.; Zuiderweg, F. J.; Klinkenberg, A. Chemical Engineering Science

1956, 5, 271-289. (47) Golay, M. J. E. Analytical Chemistry 1957, 29, 928-932. (48) Giddings, J. C. Journal of Chromatography 1960, 4, 11-20. (49) Knox, J. H.; Parcher, J. F. Analytical Chemistry 1969, 41, 1599. (50) Colin, H.; Guiochon, G. Journal of Chromatography 1977, 141, 289-312. (51) Horvath, C.; Lin, H. J. Journal of Chromatography 1978, 149, 43-70. (52) Meyer, V. R. Journal of Chromatography 1985, 334, 197-209. (53) Billiet, H. A. H.; Degalan, L. Journal of Chromatography 1989, 485, 27-50. (54) Abian, J.; Oosterkamp, A. J.; Gelpi, E. Journal of Mass Spectrometry 1999, 34,

244-254. (55) Roses, M. Journal of Chromatography A 2004, 1037, 283-298.

58

(56) Zhu, P. L.; Snyder, L. R.; Dolan, J. W.; Djordjevic, N. M.; Hill, D. W.; Sander, L. C.; Waeghe, T. J. Journal of Chromatography A 1996, 756, 21-39.

(57) Zhu, P. L.; Dolan, J. W.; Snyder, L. R. Journal of Chromatography A 1996, 756, 41-50.

(58) Zhu, P. L.; Dolan, J. W.; Snyder, L. R.; Hill, D. W.; VanHeukelem, L.; Waeghe, T. J. Journal of Chromatography A 1996, 756, 51-62.

(59) Zhu, P. L.; Dolan, J. W.; Snyder, L. R.; Djordjevic, N. M.; Hill, D. W.; Lin, J. T.; Sander, L. C.; VanHeukelem, L. Journal of Chromatography A 1996, 756, 63-72.

(60) Bidlingmaier, B.; Unger, K. K.; von Doehren, N. Journal of Chromatography A 1999, 832, 11-16.

(61) Lesins, V.; Ruckenstein, E. Journal of Chromatography 1989, 467, 1-14. (62) Jandera, P. Journal of Chromatography A 2006, 1126, 195-218. (63) Snyder L.R., G. J. L., Kirkland J.J. Wiley-Interscience, New York 1997, Chapter 8. (64) Dolan, J. W. Lc Gc North America 2003, 21, 968. (65) Dolan, J. W. Lc Gc North America 2002, 20, 940 (66) Neue, U. D.; Mazzeo, J. R. Journal of Separation Science 2001, 24, 921-929. (67) Jandera, P. Journal of Chromatography 1989, 485, 113-141. (68) Stadalius, M. A.; Quarry, M. A.; Snyder, L. R. Journal of Chromatography 1985,

327, 93-113. (69) Takeuchi, T.; Ishii, D. Journal of Chromatography 1982, 238, 409-418. (70) Takeuchi, T.; Ishii, D. Journal of Chromatography 1981, 213, 25-32. (71) Karlsson, K. E.; Novotny, M. Analytical Chemistry 1988, 60, 1662-1665. (72) Tomer, K. B.; Moseley, M. A.; Deterding, L. J.; Parker, C. J. Mass Spectrometry

Reviews 1994, 13, 431-457. (73) Amirkhani, A.; Heldin, E.; Markides, K. E.; Bergquist, J. Journal of Chromatogra-

phy B-Analytical Technologies in the Biomedical and Life Sciences 2002, 780, 381-387.

(74) Hjertén, S. Chromatography Review 1967, 9, 122. (75) Jorgenson, J. W.; Lukacs, K. D. Clinical Chemistry 1981, 27, 1551-1553. (76) Jorgenson, J. W. Analytical Chemistry 1986, 58, A743 (77) Bonvent, J. J.; Barberi, R.; Bartolino, R.; Capelli, L.; Righetti, P. G. Journal of

Chromatography A 1996, 756, 233-243. (78) Castelletti, L.; Verzola, B.; Gelfi, C.; Stoyanov, A.; Righetti, P. G. Journal of

Chromatography A 2000, 894, 281-289. (79) Sebastiano, R.; Gelfi, C.; Righetti, P. G.; Citterio, A. Journal of Chromatography A

2001, 924, 71-81. (80) Hernandez-Borges, J.; Neususs, C.; Cifuentes, A.; Pelzing, M. Electrophoresis

2004, 25, 2257-2281. (81) Puerta, A.; Axen, J.; Soderberg, L.; Bergquist, J. Journal of Chromatography B-

Analytical Technologies in the Biomedical and Life Sciences 2006, 838, 113-121. (82) Pontoglio, A.; Vigano, A.; Sebastiano, R.; Maragnoli, L.; Righetti, P. G.; Gelfi, C.

Electrophoresis 2004, 25, 1065-1070. (83) Righetti, P. G.; Gelfi, C.; Sebastiano, R.; Citterio, A. Journal of Chromatography A

2004, 1053, 15-26. (84) Arpino, P. J.; Guiochon, G. Analytical Chemistry 1979, 51, A682. (85) Horning, E. C.; Carroll, D. I.; Dzidic, I.; Haegele, K. D.; Horning, M. G.; Still-

wel.Rn Journal of Chromatography 1974, 99, 13-21.

59

(86) Nilsson, S. L.; Andersson, C.; Sjoberg, P. J. R.; Bylund, D.; Petersson, P.; Jornten-Karlsson, M.; Markides, K. E. Rapid Communications in Mass Spectrometry 2003, 17, 2267-2272.

(87) Robb, D. B.; Covey, T. R.; Bruins, A. P. Analytical Chemistry 2000, 72, 3653-3659. (88) Bruins, A. P.; Covey, T. R.; Henion, J. D. Analytical Chemistry 1987, 59, 2642-

2646. (89) Hopfgartner, G.; Wachs, T.; Bean, K.; Henion, J. Analytical Chemistry 1993, 65,

439-446. (90) Blakley, C. R.; Vestal, M. L. Analytical Chemistry 1983, 55, 750-754. (91) Karas, M.; Hillenkamp, F. Analytical Chemistry 1988, 60, 2299-2301. (92) McFadden, W. H.; Schwartz, H. L.; Evans, S. Journal of Chromatography 1976,

122, 389-396. (93) Yinon, J.; Hwang, D. G. Journal of Chromatography 1985, 339, 127-137. (94) Willoughby, R. C.; Browner, R. F. Analytical Chemistry 1984, 56, 2626-2631. (95) Wong, S. F.; Meng, C. K.; Fenn, J. B. Journal of Physical Chemistry 1988, 92, 546-

550. (96) Alexandrov, M. L.; Gall, L. N.; Krasnov, N. V.; Nikolaev, V. I.; Pavlenko, V. A.;

Shkurov, V. A. Doklady Akademii Nauk Sssr 1984, 277, 379-383. (97) Bruins, A. P. Journal of Chromatography 1991, 554, 39-46. (98) Kebarle, P. Journal of Mass Spectrometry 2000, 35, 804-817. (99) Zeleny, J. J. Phys. Rev. 1917, 10, 1. (100) Bose, G. Wittenberg 1745. (101) Taylor, G. Proc R Soc London A 1964, 280, 383. (102) Kebarle, P. 1997, 3-63. (103) Iribarne, J. V.; Thomson, B. A. Journal of Chemical Physics 1976, 64, 2287-2294. (104) Chen, D. R.; Pui, D. Y. H. Aerosol Science and Technology 1997, 27, 367-380. (105) Delamora, J. F.; Loscertales, I. G. Journal of Fluid Mechanics 1994, 260, 155-184. (106) Enke, C. G. Analytical Chemistry 1997, 69, 4885-4893. (107) Constantopoulos, T. L.; Jackson, G. S.; Enke, C. G. Journal of the American Society

for Mass Spectrometry 1999, 10, 625-634. (108) Cech, N. B.; Enke, C. G. Analytical Chemistry 2000, 72, 2717-2723. (109) Ikonomou, M. G.; Blades, A. T.; Kebarle, P. Analytical Chemistry 1990, 62, 957-

967. (110) Raffaelli, A.; Bruins, A. P. Rapid Communications in Mass Spectrometry 1991, 5,

269-275. (111) Zook, D. R.; Bruins, A. P. International Journal of Mass Spectrometry and Ion

Processes 1997, 162, 129-147. (112) Kostiainen, R.; Bruins, A. P. Rapid Communications in Mass Spectrometry 1994, 8,

549-558. (113) Caprioli, R. M.; Moore, W. T.; Martin, M.; Dague, B. B.; Wilson, K.; Moring, S.

Journal of Chromatography 1989, 480, 247-257. (114) Lee, E. D.; Muck, W.; Henion, J. D.; Covey, T. R. Journal of Chromatography

1988, 458, 313-321. (115) Smith, R. D.; Barinaga, C. J.; Udseth, H. R. Analytical Chemistry 1988, 60, 1948-

1952. (116) Smith, R. D.; Olivares, J. A.; Nguyen, N. T.; Udseth, H. R. Analytical Chemistry

1988, 60, 436-441.

60

(117) King, R.; Bonfiglio, R.; Fernandez-Metzler, C.; Miller-Stein, C.; Olah, T. Journal of the American Society for Mass Spectrometry 2000, 11, 942-950.

(118) Bonfiglio, R.; King, R. C.; Olah, T. V.; Merkle, K. Rapid Communications in Mass Spectrometry 1999, 13, 1175-1185.

(119) Sterner, J. L.; Johnston, M. V.; Nicol, G. R.; Ridge, D. P. Journal of Mass Spec-trometry 2000, 35, 385-391.

(120) Kitamura, R.; Matsuoka, K.; Matsushima, E.; Kawaguchi, Y. Journal of Chroma-tography B 2001, 754, 113-119.

(121) Annesley, T. M. Clinical Chemistry 2003, 49, 1041-1044. (122) Lu, W. Z.; Poon, G. K.; Carmichael, P. L.; Cole, R. B. Analytical Chemistry 1996,

68, 668-674. (123) Temesi, D.; Law, B. Lc Gc North America 1999, 17, 626. (124) Zhou, S. L.; Hamburger, M. Rapid Communications in Mass Spectrometry 1995, 9,

1516-1521. (125) Zook, D. R.; Forsmo-Bruce, H.; Briem, S. Chromatographia 2000, 52, S60-S64. (126) Cech, N. B.; Enke, C. G. Mass Spectrometry Reviews 2001, 20, 362-387. (127) Miller, P. E.; Denton, M. B. Journal of Chemical Education 1986, 63, 617-622. (128) Mamyrin, B. A. International Journal of Mass Spectrometry 2001, 206, 251-266. (129) Nourse, B. D.; Cooks, R. G. Analytica Chimica Acta 1990, 228, 1-21. (130) Dienes, T.; Pastor, S. J.; Schurch, S.; Scott, J. R.; Yao, J.; Cui, S. L.; Wilkins, C. L.

Mass Spectrometry Reviews 1996, 15, 163-211. (131) McLuckey, S. A.; Wells, J. M. Chemical Reviews 2001, 101, 571-606. (132) Burlingame, A. L.; Boyd, R. K.; Gaskell, S. J. Analytical Chemistry 1998, 70,

647R-716R. (133) Bruins, A. P. Mass Spectrometry Reviews 1991, 10, 53-77. (134) Loo, J. A.; Edmonds, C. G.; Udseth, H. R.; Smith, R. D. Analytica Chimica Acta

1990, 241, 167-173. (135) Desiderio, D. M. Life Sciences 1992, 51, 169-176. (136) Boyd, R. K. Rapid Communications in Mass Spectrometry 1993, 7, 257-271. (137) Rosing, H.; Man, W. Y.; Doyle, E.; Bult, A.; Beijnen, J. H. Journal of Liquid

Chromatography & Related Technologies 2000, 23, 329-354. (138) Wieling, J. Chromatographia 2002, 55, S107-S113. (139) Eltayar, N.; Vandewaterbeemd, H.; Gryllaki, M.; Testa, B.; Trager, W. F. Interna-

tional Journal of Pharmaceutics 1984, 19, 271-281. (140) Liang, H. R.; Foltz, R. L.; Meng, M.; Bennett, P. Rapid Communications in Mass

Spectrometry 2003, 17, 2815-2821. (141) Stokvis, E.; Rosing, H.; Beijnen, J. H. Rapid Communications in Mass Spectrome-

try 2005, 19, 401-407. (142) FDA US Food and Drug Administration 1994, CMC3. (143) ICH http://www.ich.org/ incorporated nov 2005. (144) Vander Heyden, Y.; Jimidar, M.; Hund, E.; Niemeijer, N.; Peeters, R.; Smeyers-

Verbeke, J.; Massart, D. L.; Hoogmartens, J. Journal of Chromatography A 1999, 845, 145-154.

(145) Dadgar, D.; Burnett, P. E.; Choc, M. G.; Gallicano, K.; Hooper, J. W. Journal of Pharmaceutical and Biomedical Analysis 1995, 13, 89-97.

(146) Fabre, H. Journal of Pharmaceutical and Biomedical Analysis 1996, 14, 1125-1132.

Acta Universitatis UpsaliensisDigital Comprehensive Summaries of Uppsala Dissertationsfrom the Faculty of Science and Technology 412

Editor: The Dean of the Faculty of Science and Technology

A doctoral dissertation from the Faculty of Science andTechnology, Uppsala University, is usually a summary of anumber of papers. A few copies of the complete dissertationare kept at major Swedish research libraries, while thesummary alone is distributed internationally through theseries Digital Comprehensive Summaries of UppsalaDissertations from the Faculty of Science and Technology.(Prior to January, 2005, the series was published under thetitle “Comprehensive Summaries of Uppsala Dissertationsfrom the Faculty of Science and Technology”.)

Distribution: publications.uu.seurn:nbn:se:uu:diva-8581

ACTA

UNIVERSITATIS

UPSALIENSIS

UPPSALA

2008

Quantitative Bioanalysis - DiVA portaluu.diva-portal.org/smash/get/diva2:171664/FULLTEXT01.pdf ·...

Documents

Transcript of Quantitative Bioanalysis - DiVA portaluu.diva-portal.org/smash/get/diva2:171664/FULLTEXT01.pdf ·...