Sample preparation of 8-hydroxy-2’-deoxyguanosine with solid phase extraction methodology based on molecular imprinting polymers and conventional silica based phases

Department of Physics, Chemistry and Biology

Bachelor’s Thesis

Sample preparation of 8-hydroxy-2’-deoxyguanosine with solid

phase extraction methodology based on molecular imprinting

polymers and conventional silica based phases

Nina Bergman June 14, 2011

LITH-IFM-G-SE-11/2468

Link¨opings University Department of Physics, Chemistry and Biology 581 83 Link¨oping

Department of Physics, Chemistry and Biology

Sample preparation of 8-hydroxy-2’-deoxyguanosine with solid

phase extraction methodology based on molecular imprinting

polymers and conventional silica based phases

Nina Bergman

Thesis work done at arbets- och milj¨omedicin, the University Hospital in Link¨oping

June 14, 2011

Supervisor: Per Leanderson Examiner: Stefan Svensson

Link¨opings University Department of Physics, Chemistry and Biology 581 83 Link¨oping

Abstract

The aim of this study was to develop methods for sample preparation for 8-OHdG in blood plasma samples with different solid phase extraction techniques using HPLC with an elec-trochemical detector. The solid phase extraction cartridges used were Chromabond® _C18,

Oasis® _{MAX, and three types of SupelMIP™ cartridges for chloramphenicol, riboflavin, and}

nitroimidazoles. The SupelMIP™ cartridges are based on molecularly imprinted polymers-technique. The separation of 8-OHdG in samples extracted from blood plasma was carried out with a Thermo Quest Hypersil Division ODS column (250 mm × 4 mm, 3µm I.D.) and methanol:buffer (10:90, v/v) as mobile phase. Recovery and selectivity was studied for the different solid phase extraction methods. The highest recovery was obtained using the Chromabond C18 cartridge with a recovery of 92%, and CV coefficient 9.5% (n = 4). 8-OHdG could not be extracted on MIP-cartridges for chloramphenicol or riboflavin, but was retained on MIP columns for nitroimidazoles, and the highest recovery was 49%.

Acknowledgments

I would like to thank my supervisor Per Leanderson for all my newfound skills that I obtained during the project, and also the rest of the employed staff at Arbets- och milj¨omedicin. I would also like to thank my examiner Stefan Svensson who always took time to help and support me during the whole project. Thank you for all the rewarding discussions about everything. Furthermore, I want to thank Roger S¨avenhed and Martin Josefsson at Link¨opings University who taught me so much about analytical chemistry. Your courses provided me with enough knowledge to work on this project.

I would also like to thank Johan Thim and Kristin Bergman for their support, Aleksandra Kyslychenko for being a good friend, and Lorentz Larsson for helping me coming in contact with Arbets- och milj¨omedicin. And last, but not least, Geertruida van Maldegem for introducing me to the world of chemistry.

Abbrevations

8-OHdG 8-hydroxy-2�-deoxyguanosine

CSSA citric acid, sodium acetate trihydrate, sodium hydroxide, acetic acid DNA deoxyribonucleic acid

HDV hydrodynamic voltammogram

HPLC high-performance liquid chromatography MAX mixed-mode anion-exchanger

MIP molecularly imprinted polymers ROS reactive oxygen species

Contents

Contents iv

1 Introduction 1

1.1 8-Hydroxy-2’-Deoxyguanosine . . . 1

1.2 Solid Phase Extraction . . . 1

1.3 Molecularly Imprinted Polymers . . . 3

1.4 High Performance Liquid Chromatography . . . 4

1.4.1 Stationary and Mobile Phases . . . 5

1.4.2 Electrochemical Detector . . . 5

1.5 The Aim of This Project . . . 5

2 Materials and Methods 6 2.1 Instrumentation . . . 6

2.2 Chemicals . . . 6

2.3 High Performance Liquid Chromatography . . . 7

2.3.1 The Electrochemical Detector . . . 7

2.4 Sample Preparations . . . 7

2.4.1 Aqueous Samples . . . 7

2.4.2 Blood Plasma Samples . . . 7

2.5 Traditional Solid Phase Extraction Methods . . . 7

2.5.1 Octadecyl Modified Silica Phase . . . 7

2.5.2 Mixed-Mode Anion-Exchanger . . . 9

2.6 Commercial Molecularly Imprinted Polymers . . . 9

3 Results 11 3.1 HPLC Method Development . . . 11

3.1.1 Analytical Recovery . . . 12

3.2 Traditional Solid Phase Extraction . . . 13

3.2.1 Octadecyl Modified Silica Phase . . . 13

3.2.2 Mixed-Mode Anion-Exchanger . . . 14

3.3 Commercial MIP Cartridges . . . 15

4 Discussion 17 4.1 HPLC Method Development . . . 17

4.2 Octadecyl Modified Silica Phase . . . 17

4.4 Commercial Molecularly Imprinted Polymers . . . 18 4.5 Conclusions and Future Work . . . 19

Bibliography 21

Appendices 24

A Preparation of the CSSA Buffer 24

B Tables 25

B.1 Volume Table for SPE with C18 . . . 25 B.2 Repeated Tests . . . 26

Chapter 1

Introduction

1.1

8-Hydroxy-2’-Deoxyguanosine

Reactive oxygen species (ROS) are formed in our cells all the time. This is a normal, important, and necessary process for the killing of invading microorganisms[1]. Normally, our bodies have a defense system (antioxidants) that keeps our levels of ROS balanced. However, carcinogenic substances like tobacco smoke, or ionisation radiation, amongst other things, can also produce ROS[2]. An overproduction can lead to a state of oxidative stress. This stress is linked to an elevated risk of oxidative damage to our DNA when the defense system is out of balance. Oxidative damage to DNA can contribute to several diseases such as cancer. However, our bodies try to repair the damaged DNA by enzyme reactions[3].

One of the most reactive oxidants is the hydroxyl radical, and it can attack the nu-cleotides in our DNA strands, proteins, and lipids[2–4]. The hydroxyl radical can be generated, e.g., after radiolysis of water caused by ionisation radiation, or after metal-driven dissociation of hydrogen peroxide. When the hydroxyl radical interacts with the nucleotide guanosine in the DNA, 8-hydroxy-2’-deoxyguanosine (8-OHdG) is formed[3]. After DNA excision repair, 8-OHdG is transported out of the cell and into the circu-lation, where it can be used as a biomarker for oxidative stress and DNA damage[5]. It can be measured in urine and blood plasma samples using high-performance liquid chromatography (HPLC)[6–8], gas-chromatography-mass spectrometry (GC-MS), liq-uid chromatography-mass spectrometry-mass spectrometry LC-MS/MS[6, 9], and HPLC with enzyme-linked immunosorbent assay (ELISA)[7]. The urinary concentrations in healthy people has been estimated to 14.7± 4.6 (n = 15) nmol/l[9] and the concentra-tions in plasma to 1–70 pg/ml[10]. In cancer patients, the concentraconcentra-tions are significantly higher[9]. In Figure 1.1, the chemical structure of 8-OHdG is displayed.

1.2

Solid Phase Extraction

When the analyte (8-OHdG in this study) is present in a complex matrix, or in low concentrations for trace analysis, the sample preparation is important, and liquid-liquid extraction (LLE) and solid phase extraction (SPE) can be helpful tools to isolate the

Figure 1.1: Chemical structure of 8-OHdG.

analyte from the matrix, concentrate, and simplify the chromatographic work[11]. The SPE technique has many advantages, such as high reproducible recovery, concentration of the analyte, requires less organic solvent than liquid-liquid extraction, and is a simple technique to use[12]. The SPE cartridge contains a reservoir where the sample is loaded, frits that function both as filters and to retain the sorbent[13], and a sorbent bed that often contains the same silica based material as the stationary phase in HPLC[14, pp. 278–279]; see Figure 1.2.

Figure 1.2: This figure shows the SPE cartridge. The sample is loaded at the top of the cartridge into something called the reservoir, the frits function as filters, and the sorbent is where the sample is adsorbed.

The analyte in the sample is either adsorbed to the sorbent material, and can be removed selectively after the interfering compounds have been washed away, or the analyte could pass through the sorbent while the interfering compounds are adsorbed to the sorbent[12].

Because of the low concentration, and overlapping peaks due to the complex matrix, it is difficult to make a quantitative analysis of 8-OHdG today, so sample preparation techniques such as SPE could be used to great advantage. For SPE analysis, C18 car-tridges, mixed mode anion exchange (MAX) cartridges[15], and strong cation exchange (SCX) cartridges[10], have been used to separate 8-OHdG from its matrix in previous studies. To approach the problem with low concentrations using HPLC, an electrochem-ical detector is most often used for detection[15].

SPE contains several steps: conditioning, loading of the samples, washing, and eluting[11, pp. 556–583][12][16, pp. 2–4].

The conditioning is made to activate the sorbent and to get the maximal phase interface between the sorbent and the sample[12]. Methanol is often used followed by water or a buffer. The sample is then loaded and passed through the cartridge. For some cartridges, the sorbent is not allowed to run dry during these steps[16, pp. 2–4]. The flow rate is another variable that is important for some cartridges[17]. The washing step elutes weakly bonded compounds that can interfere in the chromatographic analysis, and the eluting step is where the analyte is eluted through the cartridge and collected[12].

The sorbents can be polar, non-polar, have ion exchange functionalities, or be of a mixed mode phase with several functionalities at the same time, which makes it possible to obtain a greater selectivity[12]. The method development in SPE is usually based on trial-and-error procedures[18].

The cartridges that are focused on in this study will be Chromabond® C18, Oasis® MAX, and commercial MIP cartridges. The C18 cartridge, which has a nonpolar sor-bent, is of the same type that is often used in reversed phase HPLC columns, where hydrophobic compounds will be adsorbed to the hydrophobic sorbent, and polar com-pounds will more easily pass through the cartridge. The Oasis® MAX cartridge contains a mixed mode of polymeric sorbents with reversed phase (C18) and anion functionalities. Quarternary amines provides the strong anion effect[7]. It is used to extract acidic com-pounds, with pKa 2–8[7, 8], from plasma, aqueous, or urine samples. Since 8-OHdG is a weak acid with its OH groups[15], and has pKa 6–8[19, 20], the Oasis® MAX cartridge might be a good choice. According to a previous study, the Oasis® MAX cartridges made it possible to remove interfering peaks, but with lower recovery than cartridges containing only C18[6].

1.3

Molecularly Imprinted Polymers

A more recent sample preparation technique is molecularly imprinted polymers (MIP), which can provide higher selectivity and a high recovery because of the selective binding sites that are imprinted in the polymer[21, 22]. This provides a lower background for complex matrices in HPLC analysis, which makes it possible to reach lower detection limits[22].

A template molecule, that resembles the target molecule as closely as possible, is mixed with functional monomers. By adding a crosslinking monomer and an initiator, a polymerization is made around the molecule. This polymerization forms a three-dimensional complex around the template molecule. When removing the template molecule, there is an imprinting of the molecule that is very specific to the target molecule by size, shape, and to the chemical functionalities[22–24]. The process is described in Figure 1.3.

When choosing the functional monomers it is important to know what type of in-teraction that is desired because the monomers decide what type of inin-teractions there are between the target molecule and the monomers[24]. Two commonly used functional monomers are methacrylic acid (MAA) and 4-vinylpyridine (4-VP), both of which forms

Figure 1.3: Construction of a MIP. The functional monomers are mixed with the template molecule. A crosslinker and initiator are added to form a three-dimensional complex around the template molecule. The template molecule is then removed and leaves a molecular imprint.

non-covalent bondings to the template molecule[25]. Other types of monomers can also be used to form other interactions, such as hydrogen-bonding, Van deer Waals inter-actions, or ionic bonding, for example[23]. Common crosslinking agents are ethylene glycol dimethacrylate (EDMA), divinylbenzene, or N,N’-methylene bisacrylamide. A common initiator for polymerisation is 2,2’-azobis(isobutyronitrile) (AIBN). The poly-mer can be packed into either ordinary SPE cartridges, or they can be connected directly to the HPLC instrumentation as a pre-column[23], or packed in special syringes (called MEPS), or in pipette tips[26]. Attempts to make a MIP for 8-OHdG in urine samples has been published using HPLC with UV detection[9]. In their study, guanosine was used as template molecule, acrylamide and 4-Vinylpyridine was used as functional monomers, and N,N’-methylene bisacrylamide as crosslinking agent. The initiator used was AIBN, and dodecanol was used as solvent.

There are many commercial MIP cartridges for specific compounds, or groups of compounds, but no cartridge for just 8-OHdG seems to be available today. During the research for this project, no records has been found of investigations of commercial MIP cartridges designed for different compounds that has been tested with 8-OHdG.

1.4

High Performance Liquid Chromatography

The HPLC system is most often used for less volatile compounds, and separates the compounds depending on their polarity in relation to the mobile phase and stationary phase used[11, pp. 556–583][14, pp. 313–329]. The HPLC system uses a pump system to obtain enough pressure to make the mobile phase pass through the column at a specific flow rate. The column contains a stationary phase of fine particles or polymers to provide a large surface area to get high-resolution separations[11, pp. 556–583]. The analyte is injected with a syringe into the injection valve and then introduced into the column together with the mobile phase flow by the pressure from a pump. The analytes are separated on a column, and will pass through the column with the mobile phase and give different retention times depending on the polarity for the analytes. The eluting compounds are recognised by a detector which produce a signal that is converted by a chromatographic software to a chromatogram with peaks whose areas corresponds to the amount of the eluting compound.

1.4.1

Stationary and Mobile Phases

Reversed phase chromatography, where the mobile phase is more polar than the station-ary phase, is the most commonly used technique[11, 14]. The eluent strength becomes higher the less polar the mobile phase is[14]. The stationary phase does normally contain silica particles bonded to carbon chains of different lengths, usually octadecyl (C18), and different functional groups such as nitrile and phenyl, can also be used[27, pp. 177–140]. The carbon chains makes the stationary phase more hydrophobic. Mobile phases of-ten used in reversed phase are water or buffers mixed with methanol, acetonitrile, or tetrahydrofurane (THF), in different ratios[27, pp. 174–177].

1.4.2

Electrochemical Detector

An electrochemical detector can be used for analytes that can be oxidized or reduced, and this can be a very sensitive detection principle. Moreover, the detector is sensitive to both temperature and flow rate[11, pp. 556–583]. The sample that is eluted from the column reaches the detector, where a potential is placed across two electrodes. Compounds that are oxidable or reducible requires different potentials to cause a reaction. The minimal potential that is necessary can be investigated in a hydrodynamic voltammogram (HDV). Normally, at the top of the HDV there is an optimum signal to noise ratio, and for potentials near this optimum, the analyte of interest is fully oxidized (or reduced)[28].

1.5

The Aim of This Project

The aim of this project was to develop a method for sample preparation of 8-OHdG in blood plasma samples, using traditional solid phase extraction and molecular imprinting polymers for 8-OHdG, followed by analysis with HPLC-ECD. The following questions were considered:

• What recovery can be obtained from different SPE and MIP cartridges? • Could a commercial MIP provide a higher selectivity than traditional SPE? • Is it possible for us to make a molecularly imprinted polymer for 8-OHdG that can

Chapter 2

Materials and Methods

In this chapter, methods for sample preparation and concentration of 8-OHdG for two traditional SPE cartridges, a C18 and a MAX cartridge, are considered. Furthermore, three different MIP cartridges, designed for chloramphenicol, riboflavin, or nitroimida-zoles, are also tested. To investigate what had an effect on the systems, the different steps used in SPE were varied and analysed. Aqueous samples spiked with 8-OHdG were used in the beginning, and blood plasma samples were tested when the different steps seemed to work. A concentration of the sample was desired because of the low amount found in blood plasma, so a concentration factor of ten was used in most of the analysis.

2.1

Instrumentation

The chromatography system used was a HPLC-Jasco X-LC™ 3185 PU equipped with Rheodyne 7125 (Cotati, California, USA) manual injector. The columns used were either Grace Smart RP 18 3u (100 mm × 4.6 mm I.D.) or Thermo Quest Hypersil Division ODS (250 mm _{× 4 mm, 3µm I.D.) with an injection volume of 10 µl. The mobile phase} was methanol:CSSA buffer (10:90, v/v) with a flow rate of 1.0 ml/min using Grace Smart and 0.6 ml/min using Thermo Quest columns, respectively. The detection system used was an ESA Coulochem II electrochemical detector. The samples were evaporated in a Savant SpeedVac.

2.2

Chemicals

Ammonium acetate and sodium acetate trihydrate were purchased from Merck (Darm-stadt, Germany); sodium hydroxide and di-sodiumhydrogenphosphate from Riedel-de Ha¨en (Seelze, Germany); 8-hydroxy-2’-deoxyguanosine (8-OHdG), methanol, hexane, and acetonitrile, from Sigma Aldrich (Spruce street, St.Louis, MO, USA); citric acid from May & Baker (Dagenham, England). Blood plasma was obtained from Link¨opings Uni-versity Hospital, SPE cartridges used were Chromabond® C18 ec, 45 µm, 1ml/100mg, Oasis® MAX (186000367) 3cc/60 mg 30 µm, 0.25 meq/g. Commercial MIP cartridges

were SupelMIP™ SPE cartridges made for chloramphenicol, 55 µm, 25 mg/10 ml (53240-U), riboflavin, 25mg/10 ml (53236-U) 58 µm, or nitroimidazoles 62 µm, 50 mg/3ml.

2.3

High Performance Liquid Chromatography

To obtain an adequate separation of the peaks in the plasma samples, several mobile phases were made with 6%, 8%, and 10% methanol/CSSA buffer, 2% acetonitrile/CSSA buffer, 0.1% acetic acid and 6% methanol/CSSA buffer, and analysed under isocratic conditions. The specifics of how the CSSA buffer was made can be found in Appendix A. Two columns were tested: Grace Smart and Thermo Quest. Different flow rates were also tested.

2.3.1

The Electrochemical Detector

To obtain a HDV for 8-OHdG, the voltage on the detector was set to different values between 100–600 mV, and samples with the same concentration were injected into the HPLC-system as the current (nA) and peak areas were registered.

2.4

Sample Preparations

2.4.1

Aqueous Samples

A stock solution of 8-OHdG with a concentration of 10 µmol/l was made in water. From this solution, a working solution with the concentration of 100 nmol/l was prepared in mobile phase, or buffer.

2.4.2

Blood Plasma Samples

To precipitate the proteins, 110 µl of cold 5% trichloroacetic acid was added to 1000 µl of both a spiked plasma sample containing 100 nmol/l 8-OHdG and an unspiked plasma sample. The samples were vortexed for 15 seconds and centrifuged at 14100g for 2 min-utes. The supernatant was removed from the pellet and transferred to Eppendorf tubes. The supernatant was then mixed with an equal amount of 1 mol/l ammonium acetate with pH 5.25.

2.5

Traditional Solid Phase Extraction Methods

2.5.1

Octadecyl Modified Silica Phase

Several measurements with different conditions were made to develop a sample prepa-ration method with C18 cartridges to see which conditions were the most suitable. This was accomplished by trying to separate the peaks and calculating the recovery for 8-OHdG in aqueous and blood plasma samples. The conditions investigated were different SPE cartridges, washing, and eluents. According to a previous study, 8-OHdG in urine

samples has been successfully extracted using C18 cartridges[29]. Using their proposed method as a starting point, several tests were carried out with different variables; see Table B.1 in the Appendix for specifics. Tests were made to see if acetonitrile or hexane could work as washing solution, and if acetonitrile could be used as eluent.

The final method obtained was the following: The C18 cartridges were first pre-conditioned with 1000 µl methanol followed with 1000 µl 25 mmol/l phosphate buffer with pH 5.5. The sample, 1000 µl for aqueous samples and approximately 700 µl for blood plasma samples, was then loaded into the cartridge and washed with 300 µl of the 25 mmol/l phosphate buffer and 300 µl water. The cartridges were dried with nitrogen gas and placed in a SpeedVac at 40°C for 15 minutes. Elution was performed with 2× 300 µl methanol. The samples were dried in a SpeedVac at 40°C for 1 hour, resuspended in 100 µl mobile phase, shaken at 2400 shakes/minute for 2 minutes, and centrifuged at 14100g for one minute. The evaporation step is made to concentrate the amount of 8-OHdG by a factor ten in the samples. See Figure 2.1 for a sketch of the proposed SPE method.

Figure 2.1: The proposed SPE method using C18 cartridges. First the cartridge is conditioned with methanol and phosphate buffer. The sample is then loaded into the cartridge. The cartridge is washed with phosphate buffer and water. The 8-OHdG is then eluted with methanol.

Blood plasma samples were tested with the final method for C18. Hexane was used as a third washing step to try to dispose of peaks that elute late, and to get a shorter analysis time. Repeated extractions were made with five blood plasma samples spiked with 100 nmol/l 8-OHdG to investigate the variation between the samples. These samples were prepared using 250 µl plasma for each sample, the remaining chemicals were rescaled to match the amount of blood plasma.

Another extraction, using 500 µl spiked blood plasma and chemicals rescaled to match this volume, was injected and analysed four times, to investigate the variation within the samples.

2.5.2

Mixed-Mode Anion-Exchanger

In a previous study[6], an Oasis® MAX cartridge has been used successfully to remove interfering peaks. Different pH values of the samples and washing solutions were tested to see if pH has an effect on the loading and washing systems. Aqueous samples were made with 100 nmol/l 8-OHdG in 25 mmol/l phosphate buffers with pH 2, 4, 5, 7, and 9. For each sample, 1000 µl was passed through one of five different MAX cartridges. Each cartridge was then washed with 1000 µl of the 25 mmol/l phosphate buffer which has the corresponding pH. Tests using hexane and methylene chloride as washing solutions, and methanol:acetic acid (98:2, v/v) and pure methanol as eluents, were made.

To see if the speed when the sample is loaded and passed through the cartridge that provides the first fraction is important, one sample was forced through the cartridge at a high speed, and one was allowed to pass through the cartridge slowly. The first fractions were collected and analysed. Investigations were also made to see if the concentration of the buffers had any effect. This was accomplished by making new samples of 100 nmol/l 8-OHdG: one in 1 mmol/l ammonium acetate buffer and one in 25 mmol/l phosphate buffer.

The final method obtained was the following: The cartridges were conditioned with 2000 µl methanol and 2000 µl Milli-Q water. The blood plasma samples were loaded into the cartridge and washed respectively with 1000 µl phosphate buffer, pH 5, and 1000 µl methylene chloride. The samples were then eluted with 1000 µl of methanol:acetic acid (98:2, v/v). The samples were evaporated in a SpeedVac at 40°C for 1 hour, and then resuspended in 100 µl mobile phase consisting of methanol:CSSA buffer (8:92, v/v), shaken at 2400 shakes/minute for 2 minutes, and centrifuged at 14100g for one minute.

2.6

Commercial Molecularly Imprinted Polymers

To see if a greater selectivity could be obtained for 8-OHdG, three different MIP car-tridges were tested: SupelMIP™ SPE cartridges for chloramphenicol, riboflavin, or nitroimidazoles. Gravity flow was used during all sample loading on these cartridges.

As a first step, aqueous solutions spiked with 100 nmol/l 8-OHdG were tested, and later blood plasma samples were tested on the nitroimidazole cartridge. The steps below follows those proposed in the data sheets for the respective cartridge; see [30–32], respectively.

SupelMIP™ SPE Chloramphenicol. The cartridges were conditioned with 1000 µl methanol, 1000 µl milli-Q water, and the samples were loaded into the cartridges. During these steps, the cartridges were not allowed to dry. 8-OHdG was eluted with 2_{× 1000 µl} methanol:acetic acid:water (89:1:10, v/v/v) and the first fraction for each sample was analysed on the HPLC to see which amount of 8-OHdG that was retained in the car-tridge.

SupelMIP™ SPE Riboflavin. The cartridges were conditioned with 1000 µl methanol, 1000 µl milli-Q water, and the cartridges were not allowed to dry during these steps. 8-OHdG was then eluted with 3 _{× 1000 µl acetonitrile:water(70:30,v/v).} The first fractions were collected and analysed on the HPLC.

SupelMIP™ SPE Nitroimidazoles. The cartridges were conditioned with 1000 µl toluene, 1000 µl acetonitrile, and 1000 µl 10 mmol/l ammonium acetate with pH 6. The samples, 1000 µl 100 nmol/l 8-OHdG in 10 mmol/l ammonium acetate buffer at pH 6, were then loaded into the cartridge. The cartridges were washed with 1000 µl milli-Q water, 2×1000 µl hexane, and 1000 µl heptane:toluene (3:1, v/v). Between each washing step, a vacuum was applied. 8-OHdG was then eluted with 2_{×1000 µl acetonitrile:water} (60:40 v/v) with 0.5% acetic acid. The samples were evaporated in a SpeedVac at 40°C over the night, and resuspended in 100 µl mobile phase consisting of methanol:CSSA buffer (10:90, v/v).

Tests were made with different conditioning solvents, similar to those used for the nitroimidazole cartridge, on the riboflavin and chloramphenicol cartridges, and attempts to elute 8-OHdG with methanol, or acetonitrile, with 0.5% acetic acid were made on the nitroimidazole cartridge to see if the evaporation time in the SpeedVac could be reduced with a different eluent. To see if the sample loading volume was important, one blood plasma sample was diluted with 100 µl 1 mmol/l ammonium acetate buffer and one with 500 µl 1 mmol/l ammonium acetate buffer.

Chapter 3

Results

3.1

HPLC Method Development

In the first part of the study, a HPLC method for analysis of 8-OHdG was set up. Different variables such as potential, mobile phases, different columns, and flow rates were investigated. The potentials were plotted against the noted peak areas in an HDV graph; see Figure 3.1. The information from the HDV shows that the optimal signal to noise ratio is near 300 mV. This potential gives a complete oxidation of 8-OHdG and was therefore used in all further experiments unless otherwise noted.

Figure 3.1: An HDV for 8-OHdG. The potentials were set between 100-600 mV and plotted against the peak areas. The optimum signal to noise ratio is near 300 mV.

Using a mobile phase consisting of methanol:CSSA buffer (6:94, v/v) on the Grace Smart column, and the flow rate set to 1 ml/min, resulted in a retention time of 4.6 min-utes. However, when blood plasma samples were used, there was an interfering peak in the chromatogram at almost the same retention time as 8-OHdG. Attempts to ob-tain a better separation of the peaks by changing the flow rate, and changing the mobile

phase to either acetonitrile:CSSA buffer (2:98, v/v), or acetic acid:methanol:CSSA buffer (0.1:5:94.9, v/v/v), had almost no effect on the Grace Smart column. A mobile phase with acetonitrile:CSSA buffer (2:98, v/v) resulted in a retention time of 5.59 minutes and no improvement of the separation compared with 6% methanol. When changing the mobile phase to methanol:CSSA buffer (10:90, v/v), using Thermo Quest Hypersil Division ODS, instead with a flow rate of 0.5 ml/min, the 8-OHdG peak was separated from the interfering peak, although the retention time of 8-OHdG was 18.2 minutes. These chromatographic conditions were therefore used in all further analysis of blood plasma samples. A Chromatogram of 8-OHdG, separated on the Thermo Quest column using C18 SPE, is shown in Figure 3.2.

Figure 3.2: Chromatogram of one blood plasma sample, spiked with 100 nmol/l 8-OHdG. The peaks are separated on the Thermo Quest Hypersil Division ODS (250 mm × 4 mm, 3µm I.D.) column, at a flow rate of 0.5 ml/min, and mobile phase consisting of methanol:CSSA buffer (10:90, v/v).

Another attempt that was made, was to change the potential by changing cell E1 to 150 mV. This had no significant effect on which compounds, interfering or not, that were detected.

3.1.1

Analytical Recovery

In the absence of an internal standard, the recovery of 8-OHdG is calculated using the peak area ratio of the sample using the SPE method of interest, to one known reference sample spiked with 100 nmol/l 8-OHdG in buffer. Since the evaporation and resuspendation enrich the samples, the peak area of the reference samples were multiplied by the corresponding concentration factor. The analysis of the aqueous reference samples were carried out on the same day, and under the same conditions, as for the blood plasma samples.

3.2

Traditional Solid Phase Extraction

3.2.1

Octadecyl Modified Silica Phase

The C18 method provided a high recovery, and 8-OHdG did not pass through in the first fraction or with the first washing step of the cartridges. Washing with hexane gave no loss in recovery in aqueous samples, but tests with blood plasma did not show any improvement of washing away late eluting peaks in the chromatogram. Washing with acetonitrile resulted in a poor recovery and could perhaps be used as eluent instead. In plasma samples, using a C18 cartridge not only gave a high recovery, but also re-moved a considerable amount of background in the latter part of the chromatogram; see Figure 3.3.

Figure 3.3: Chromatogram of one spiked blood plasma sample (100 nmol/l 8-OHdG) using solid phase extraction on a C18 SPE cartridge, and one diluted blood plasma sample (blood plasma:1 mol/l ammo-nium acetate buffer, 1:1, v/v) without SPE. The black line shows the spiked blood plasma sample, using SPE with a C18 cartridge, and the purple line shows the blood plasma without SPE. The retention time for 8-OHdG in the spiked blood plasma sample was 4.7 min. The HPLC column used was Grace Smart RP 18 3u (100 mm × 4.6 mm I.D.) with a flow rate of 1.0 ml/min, and mobile phase consist-ing of methanol:CSSA buffer (8:92, v/v). The peaks are not separated under these chromatographic conditions.

In Table 3.1, the peak areas for each step of the extraction and calculated recovery for aqueous samples spiked with 100 nmol/l 8-OHdG is shown; see Table B.1 in Appendix for specifics of the different tests. In blood plasma samples, hexane was tested as a washing step to see if the peaks that elute late could be avoided so a shorter analysis time could be obtained, but the results showed no significant difference to chromatograms without hexane. The results of the analysis of the blood plasma samples are described below.

Table 3.1: Areas in mVs and recovery for aqueous solutions of 100 nmol/l 8-OHdG using SPE with a C18 cartridge. The volume of the spiked aqueous sample used in these tests is 900µl, so the concentration factor is nine. The table shows the peak areas of the reference sample with 100 nmol/l 8-OHdG before extraction; in the first fraction; in wash 1–3; in the eluent after extraction, evaporation, and redilution of the sample in 100 µl mobile phases; and the calculated recovery for each test.

Test Ref. sample Frac 1 Wash 1 Wash 2 Wash 3 Elution Recovery (%)

1a 150 - - - 833 166 12 1b 150 - - - - 1274 94 2a 156 - - 2 - 2305 164 2b 156 - - - - 2000 142 2c 156 - - - - 1536 109 3a 163 - - - - 1631 111 3b 163 - - - - 1742 119 3c 163 - - - - 1648 112 3d 207 - - - - 2191 118

To examine the variation in measurements, five blood plasma samples were prepared with concentration 100 nmol/l 8-OHdG. Assuming that the measurements follow a nor-mal distribution, the confidence interval

64.9 < µ < 75.0 (%)

was obtained for the recovery µ using a confidence degree of 95%. See Table B.2a in Appendix. In this case, the variational coefficient CV was calculated to be 5.8%. Similarly, the same sample with the concentration 100 nmol/l was injected four times to investigate how much the measurements varies within the same sample. Assuming a normal distribution again, a 95% confidence interval for the recovery µ was calculated to be

70.8 < µ < 95.9 (%)

The variational coefficient obtained was 9.5%. See Table B.2b in Appendix.

3.2.2

Mixed-Mode Anion-Exchanger

The results from the analysis of the first fractions of the aqueous samples at different pH values, was that everything was trapped in the cartridges independently of pH value. In further aqueous solutions, pH 5 was chosen. When using a syringe to force the sample through the cartridge, 17% passes through without bonding to the cartridge, and can be seen in the first fraction. The concentration of the buffer had no detectable effect on the sample loading. In one of the tests, where hexane was used in the washing step, there was a loss of 7% 8-OHdG that passed through with the hexane. Using methylene chloride as the washing solvent instead caused no loss, so this was used in further analysis as the washing step. Eluting one aqueous sample with methanol and 0.5% acetic acid gave the highest recovery (109%). The sample eluted with pure methanol resulted in 85% recovery in an aqueous sample. Using the final method for blood plasma samples, the

recovery obtained was 65%. A chromatogram of a spiked blood plasma sample extracted with a MAX cartridge is shown in Figure 3.4.

Figure 3.4: Chromatogram of one spiked blood plasma sample with 100 nmol/l 8-OHdG using solid phase extraction on a MAX cartridge. The HPLC column used was Thermo Quest Hypersil Division ODS (250 mm × 4 mm, 3µm I.D.) with a flow rate of 0.6 ml/min, and mobile phase consisting of methanol:CSSA buffer (8:92, v/v). There is a ridge at the end of the chromatogram, which indicates that not all of the nonpolar compounds had been removed during the SPE.

3.3

Commercial MIP Cartridges

At first, aqueous solutions spiked with 100 nmol/l 8-OHdG were tested. To the riboflavin cartridge, 62% 8-OHdG was retained. and to the chloramphenicol cartridge, 50% was retained. However, when the cartridges for nitroimidazoles were used, 100% of the 8-OHdG was retained. To see if the riboflavin and chloramphenicol cartridges could work with conditions similar to those used with the nitroamidozole cartridge, the same pre-conditioning procedure was used for those cartridges. On the chloramphenicol cartridge, only 29% was retained, and on the riboflavin cartridge, 45% was retained. Since it did not work sufficiently well, it was decided to do further analysis only on the nitroimidazole cartridge. Tests to elute 8-OHdG with acetonitrile:water (60:40, v/v) and 0.5% acetic acid, methanol with 0.5% acetic acid, and acetonitrile with 0.5% acetic acid, gave the recoveries 44%, 86%, and 47%, respectively; see Table 3.2.

The use of methanol with 0.5% acetic acid also reduced the evaporation time from one night to one hour, so all further elutions were made with methanol and 0.5% acetic acid since it also provided the highest recovery in the aqueous samples. After the first washing step with water, there was a small loss of 8-OHdG where 17% seemed to follow

Table 3.2: Areas in mVs and recovery for aqueous samples with the concentration 100 nmol/l 8-OHdG extracted on the nitroimidazole MIP cartridge. Peak areas of the reference sample before the extraction; in the first fraction; in wash 1–3; in the eluent after extraction, evaporation, and redilution of the sample in 100 µl mobile phase; and the calculated recovery. In test A, acetonitrile:water (60:40,v/v) with 0.5% acetic acid was used to elute 8-OHdG. In test B and C, methanol and 0.5% acetic acid, and acetonitrile and 0.5% acetic acid, respectively, were tested.

Test Ref. sample Frac 1 Wash 1 Wash 2 Wash 3 Elution Recovery

A 152 - 26 209 6 668 44%

B 152 - - 89 14 1310 86%

C 152 - - 196 6 708 47%

the water through the cartridge in one of the samples. In the second and third washing step, there is too much loss, so these washing steps are eliminated. In blood plasma samples, the samples that were diluted with 100 µl and 500 µl 1 mmol/l ammonium acetate buffer gave the recoveries 13% and 49%, respectively. A chromatogram using nitroimidazole cartridge can be found in Figure 3.5

Figure 3.5: Chromatogram of one blood plasma sample spiked with 100 nmol/l 8-OHdG using solid phase extraction with MIP nitroimidazole cartridges. The column used was Thermo Quest Hypersil Division ODS (250 mm × 4 mm, 3µm I.D.) with at a flow rate of 0.5 ml/min, and mobile phase consisting of methanol:CSSA buffer (10:90, v/v).

Chapter 4

Discussion

4.1

HPLC Method Development

In the beginning of the project, the HPLC column used was Grace Smart RP 18 3u (100 mm × 4.6 mm I.D.). Different mobile phases and flow rates were tested to see if sufficient separation of the peaks could be obtained in blood plasma samples. This was not successful due to the complex matrix. Many of these tests are left out of the report since they did not contribute to any important progress.

According to a previous study[10], 8-OHdG has been separated from plasma samples using a Develosil C30 column. There was no access to any type of HPLC column with a C30 stationary phase, so other columns were tested instead. One such column was the Thermo Quest Hypersil Division ODS (250 mm × 4 mm, 3µm I.D.), which made it possible to separate the peaks with an approximate retention time of 18 minutes using a flow rate of 0.5 ml/min. The time of analysis was quite long for the blood plasma samples, about 35 minutes per analysis. One possible solution could be to use gradient separation, and increase the concentration of methanol later in the chromatogram. This could decrease the time of analysis. However, this method will also require some time for the mobile phase system to stabilise between each new analysis.

4.2

Octadecyl Modified Silica Phase

The C18 cartridge gave a quite high recovery, about 70%, but the results show a vari-ation, even within the same sample. The recoveries in the aqueous samples are a lot higher than 100% in some cases, which could be an effect from the variation. To provide an accurate analysis in future work, an internal standard is needed. One possible reason for the variation is the manual injection, so an auto injector should be able to provide a lower variation. Another reason can be because of the low amount of blood plasma that was used in the tests, but since there was a lack of blood plasma during this analysis there was no possibility to use larger amounts for each extraction. The pipettes may also have contributed to some quantity of the variation.

Since there was no access to a suitable internal standard during the analysis in this project, it is difficult to make any quantitative analysis before this problem with the

variation is solved. The chromatograms using SPE with C18 in blood plasma, shows that a large portion of the interfering matrix has been removed; see Figure 3.3.

4.3

Mixed-Mode Anion-Exchanger

When sample preparation was made with the Oasis®MAX cartridge, the pH values were varied, but it was found that pH had no effect on the bonding of 8-OHdG in aqueous solutions. All 8-OHdG was retained in the sorbent in the cartridge independently of pH value. This indicates that only the reversed phase effect is active. A better result could possibly had been obtained in plasma samples if also the anion effect had been activated by treatment of the cartridge and sample with an alkaline solution. This was not tested. If a lower pH, less than 2.5, had been used, some of the interactions to the proteins in the matrix could perhaps had been broken. In the analysis, approximately pH 5 was used. Hence, only the reversed phase effect was active. A combined effect, from both reversed phase and anion exchanger effects, should probably provide a higher recovery.

In the chromatograms for the MAX cartridge, there is a ridge at the end; see, e.g., Figure 3.4. This indicates that the SPE with the MAX cartridge does not work suf-ficiently well to remove the nonpolar compounds. This causes the analysis to take approximately 10 minutes longer compared to SPE made with the C18 cartridge, since the base line takes longer to stabilise between the different analyses. The recovery is less than that obtained for C18, but the peaks are still separated and there are no interfering peaks at the same retention time as 8-OHdG.

To summarise, the MAX cartridge provided 65% recovery for blood plasma samples, and the time of analysis was longer compared with that after C18 cartridges due to the time it takes for the baseline to stabilise after each analysis.

4.4

Commercial Molecularly Imprinted Polymers

The surprising result when testing the commercial MIP columns, was that one of them actually worked for 8-OHdG. Since MIP cartridges are tailor-made to match a specific compound, or a group of similar compounds, it is surprising that recovery of 8-OHdG in plasma samples was high when using the nitroimidazole cartridge. The discovery that elution with methanol and 0.5% acetic acid could be made instead of using the proposed acetonitrile:water (60:40, v/v) solution is very useful since it reduces the evaporation time from one night to one hour. The water does not seem to have a great effect on the elution process, but this does not explain why the tests with acetonitrile and 0.5% acetic acid did not work as well as an eluent.

The other two cartridges tested (riboflavin and chloramphenicol) did not work suffi-ciently well, even for aqueous samples. The reason for this might be that the molecules are so different, both in size and other properties, that 8-OHdG does not bind to the sorbent in these cartridges. Therefore, these cartridges were not used for blood plasma samples.

Since the knowledge of how the cartridges work and what they are made of is lacking, it is very difficult to proceed with developing a method that works better than the one suggested previously. The loading and washing step had a recommended flow rate of less than 0.5 ml/min, or to use gravity flow if possible. In this study, gravity flow was applied. This process was time consuming and complicated due to the fact that the sorbent was not allowed to dry during the conditioning of cartridge. It is also not clear if everything was eluted when using this eluent, so further analysis is necessary to ensure that everything is extracted from the cartridge so following samples are not contaminated by leftover substances if the cartridges are used more than once.

Another phenomenon that was somewhat surprising was the fact that blood plasma samples that were diluted provided a higher recovery than the corresponding undiluted ones. One possible explanation could be that in the undiluted sample, proteins and other compounds interfere more when 8-OHdG is attached to binding sites in the cartridge.

The chromatograms for blood plasma samples using the nitroimidazole cartridge shows that all of the interfering compounds have been removed; see Figure 3.5. Com-pared with the chromatogram where the MAX cartridge was used, this one is better due to the lack of the ridge at the end, and therefore has a shorter time of analysis. But still, the C18 method is the best choice since it provides the highest recovery.

4.5

Conclusions and Future Work

The recovery for all methods were lower in blood plasma samples than in the aqueous samples, which was expected. One possible reason for this is that the complex matrix probably takes binding sites from 8-OHdG in the cartridges. Another reason for the loss of recovery in blood plasma samples is that 8-OHdG could be affected by the protein precipitation when trichloroacetic acid was used. One test was made to try to ultra-filtrate the blood plasma instead, using a membrane with a molecular weight cut off at 10000 dalton, but the filter broke because of the centrifugal forces. A different type of filter is necessary, but was not available at the time. There should be more analysis made with filtration, or alternative precipitation techniques, to see if this has an effect on the recovery.

The C18 cartridge was the most straight-forward sample preparation method to work with. The flow rate had no significant effect and the cartridge was allowed to run dry between the steps, so the laboratory work was less demanding. The MAX cartridge was more time consuming since the flow rate of the sample loading had an effect on the recovery. Furthermore, the method requires several chemicals that are toxic and bad for the environment, and it was necessary to carry out the extractions under a fume hood. The nitroimidazole MIP cartridge was the most difficult to work with for several reasons. The cartridge was not allowed to run dry between steps, and for some parts of the extraction it was necessary to apply a vacuum. Since the flow rate was very low, using only the gravity, the process also took a longer time.

If there would have been time to produce an in-house MIP specifically designed for 8-OHdG, it should have been possible to achieve a higher selectivity for 8-OHdG and to obtain a greater recovery.

Since there was variation, an internal standard, and an auto injector, is probably needed in further analysis. The variation could also have been affected by the low volumes of blood plasma used during those tests. Investigations using a larger amount of blood plasma may reduce the variation caused by the the pipettes. An automation of the SPE method is another alternative to reduce the variation between the samples by lowering the operator errors[16, pp. 243–244]. The fact that the variation within the same sample was higher than the variation between the samples remains unexplained. All of the repeated tests were carried out during the same day and with the same procedure.

The recovery calculation is also uncertain since the peak areas of the reference solu-tions are multiplied by the concentration factor. One should probably use a reference solution of the same concentration as the enriched sample instead of using a multiplica-tion factor. By doing this, it is possible to avoid extrapolating informamultiplica-tion.

The contents in blood plasma also varies from person to person, so there is a pos-sibility that the method may vary between different people. To be able to make a quantitative analysis in patients, the limit of quantification has to be determined to see if the proposed method with C18 could work with the chromatographic conditions obtained. A calibration curve should be derived for spiked blood plasma samples, so that the concentration can be estimated in unknown blood plasma samples. If it turns out not to be possible to detect low enough concentrations using SPE with C18 in blood plasma samples, one should look into constructing a MIP to remove more of the matrix and reach a lower detection limit.

To successfully construct a MIP for 8-OHdG requires a lot of time and planning. The first step is to decide what chemicals that should be used. A larger study of the current literature is needed to see what type of crosslinking agent, functional monomers, initiator, solvents, and template molecule, that can be used. A template molecule must be chosen to be as similar as possible to the target molecule, which in this case is 8-OHdG. Possibly one can use guanosine, which was suggested in a previous study[9], since 8-OHdG is too expensive. The functional monomer and crosslinking agent must be picked carefully. One must consider what types of interactions that are desirable is this case, to obtain a high selectivity for 8-OHdG in blood plasma samples. When it is decided what chemicals to use, a suitable cartridge, pipette, pre-column, or similar object, has to be chosen to contain the molecular imprinted polymer. After this, it is possible to calculate the amounts of chemicals that are needed. When the MIP is completed, one can proceed with method development, e.g., testing different conditioning steps, washing steps, and eluents.

Bibliography

[1] Moslen, M.T., Reactive oxygen species in normal physiology, cell injury and phago-cytosis, Advances in Experimental Medicine and Biology 366 (1994), 17–27. [2] Klaunig, J.E. et al, The role of oxidative stress in chemical carcinogenesis,

Envi-ronmental Health Perspectives 106 (1998), Suppl. 1, 289–295.

[3] Valavanidis, A. et al, 8-hydroxy-2�-deoxyguanosine(8OHdG): A Critical Biomarker of Oxidative Stress and Carcinogenesis, Journal of environmental science and health. Part C, Environmental carcinogenesis & ecotoxicology reviews 27 (2009), no. 2, 120–139

[4] Moriwaki, H. et al, Effects of mixing metal ions on oxidative DNA damage mediated by a Fenton-type reduction, Toxicology in Vitro 22 (2008), no. 1, 36–44.

[5] Kimura, S. et al, Evaluation of Urinary 8-Hydroxydeoxyguanine in Healthy Japanese People, Basic and Clinical Pharmacology and Toxicology 98 (2006), no. 5, 496–502. [6] Matayatsuk, C., Quantitative determination of 8-hydroxy-2�-deoxyguanosine as a biomarker of oxidative stresss in thalassemic patients using HPLC with an electro-chemical detector, Journal of Analytical Chemistry 63 (2008), no. 1, 52–56.

[7] Oasis® technical note, Oasis® MAX product and generic method information, Wa-ters Corporation 2003, http://www.younglin.com/brochure_pdf/waWa-ters/MAX. pdf; verified 2011-05-22.

[8] Oasis® care and use manual, Oasis mixed-mode ion-exchange cartridges and 96-well plates, Waters Corporation 2010, http://www.waters.com/waters/support. htm?locale=en_us&lid=10076790&cid=511442&type=USRM; verified 2011-05-22. [9] Zhang, S.W. et al, Molecularly imprinted monolith in-tube solid-phase

microex-traction coupled with HPLC/UV detection for determination of 8-hydroxy-2’-deoxyguanosine in urine, Analytical and Bioanalytical Chemistry 395 (2009), no. 2, 479–487.

[10] Koide, S. et al, Determination of human serum 8-hydroxy-2’-deoxyguanosine (8OHdG ) by HPLC-ECD combined with solid phase extraction (SPE ), Journal of Chromatography B 878 (2010), no. 23, 2163–2167.

[11] Harris, D.C., Quantitative Chemical Analysis, 7th ed., W.H. Freeman, New York 2007.

[12] Biotage®, QuickStart guide to SPE, Biotage 2011, http://www.biotage.com/ DynPage.aspx?id=49597; verified 2011-05-22.

[13] Applied Separations, Inc. 2010, Solid Phase Extraction Frits and Filters, http: //www.appliedseparations.com/SPE/frits/default.asp; verified 2011-05-25. [14] Simonsen, F., Analysteknik: instrument och metoder, Studentlitteratur, Lund 2005. [15] Zhang, S.W. et al, Determination of urinary 8-hydroxy-2’-deoxyguanosine by cap-illary electrophoresis with molecularly imprinted monolith in-tube solid phase mi-croextraction, Chinese Chemical Letters 21 (2010), no. 1, 85–88.

[16] Thurman, E.M. et al, Solid-Phase Extraction: Principles and Practice, John Wiley & Sons, New York 1998

[17] Argonaut technical note, Method Development in Solid Phase Extraction using ISO-LUTE SCX and SCX-3 SPE Columns for the extraction of Aqueous Samples, Dr. W´eber Consulting Kft. 2001, http://www.weber.hu/PDFs/SPE/Tn106_SCX_SPE. pdf; verified 2011-05-22.

[18] Poole, C.F. et al, Contributions of theory to method development in solid-phase extraction, Journal of Chromatography A 885 (2000), no 1–2, 17–39.

[19] Arnett, S.D. et al, Enhanced pH-mediated stacking of anions for CE incorporating a dynamic pH junction, Electrophoresis 28 (2007), no. 20, 3786–3793.

[20] Qian, K. et al, Preparation and application of a molecularly imprinted polymer for the determination of trace metolcarb in food matrices by high performance liquid chromatography, Journal of Separation Science 33 (2010), no. 14, 2079–2085. [21] Komiyama, M. et al, Molecular Imprinting: From Fundamentals to Applications,

Wiley-VCH Verlag, 2003 Weinheim.

[22] SupelMIP data sheet 407075, Molecularly imprinted polymers for the highly selective extraction of trace analytes from complex matrices, Supelco Analytical 2009, http://www.sigmaaldrich.com/etc/medialib/docs/Supelco/ General_Information/t407075.pdf; verified 2011-05-23.

[23] Las´akov´a, M. et al, Molecularly imprinted polymers and their application in solid phase extraction, Journal of Separation Science 32 (2009), no. 5–6, 799–812. [24] Turiel, E. et al, Molecularly imprinted polymers for sample preparation: A review,

Analytica Chimica Acta 668 (2010), no. 2, 87–99.

[25] Hwang, C.C. et al, Chromatographic characteristics of cholesterol-imprinted poly-mers prepared by covalent and non-covalent imprinting methods, Journal of Chro-matography A 962 (2002), no. 1–2, 69–78.

[26] Altun, Z. et al, Increasing sample preparation throughput using monolithic methacrylate polymer as packing material for 96-tip robotic device, Journal of Liquid Chromatography and Related Technologies 29 (2006), no. 10, 1477–1489.

[27] Meyer, V.R., Practical High-Performance Liquid Chromatography, 5th ed., John Wiley & Sons, West Sussex 2010.

[28] ESA technical note 70-6348, Hydrodynamic Voltammograms: Generation, Expla-nation, and Optimization of Applied Potentials, ESA inc., http://www.esainc. com/download/?id=131; verified 2011-05-22.

[29] Sabatini, L. et al, A method for routine quantitation of urinary 8-hydroxy-2-deoxyguanosine based on solid-phase extraction and micro-high-performance liquid chromatography/electrospray ionization tandem mass spectrometry, Rapid Commu-nications in Mass Spectrometry 19 (2004), no. 2, 147–152.

[30] Supelco instruction set (data sheet), SupelMIP™ SPE - Chloramphenicol, Sigma-Aldrich Co. 2008. http://www.sigmaaldrich.com/etc/medialib/docs/Supelco/ Product_Information_Sheet/t706024.pdf; verified 2011-05-22.

[31] Supelco instruction set (data sheet), SupelMIP™ SPE - Riboflavin (Vitamin B2), Sigma-Aldrich Co. 2007. http://www.sigmaaldrich.com/etc/medialib/docs/ Supelco/Product_Information_Sheet/t706022.pdf; verified 2011-05-22.

[32] Supelco instruction set (data sheet), SupelMIP™ SPE - Nitroimidazoles, Sigma-Aldrich Co. 2009. http://www.sigmaaldrich.com/etc/medialib/docs/Supelco/ Datasheet/t709075.pdf; verified 2011-05-22.

Appendix A

Preparation of the CSSA Buffer

Total volume is 1l. Citric acid:

C = 12.5 mmol/l, M = 210.14 g/mol, m = 2.62 g, and n = 12.47 mmol. Sodium acetate trihydrate:

C = 25 mmol/l, M = 136.08 g/mol, m = 3.4 g, and n = 24.99 mmol. Sodium hydroxide:

C = 30 mmol/l, M = 40.00 g/mol, m = 1.2 g, and n = 30.00 mmol. Acetic acid:

Appendix B

Tables

B.1

Volume Table for SPE with C18

Table B.1: Solvent and volume table for different experiments with SPE. The variables were tried out to get the best conditions and recovery of 8-OHdG in aqueous samples for the C18 cartridge. A is 900 µl methanol, B is 300 µl water, C is 300 µl 25 mM phosphate buffer with pH 5.5, D is 100 µl mobile phase with methanol:CSSA buffer (8:92, v/v), E is 900 µl 25 mM phosphate buffer with pH 5.5, F is 300 µl acetonitrile, G is 300 µl hexane, and H is 300 µl methanol.

Test Conditioning Wash 1 Wash 2 Drying Wash 3 Elution Resuspendation

1a A + E C B Yes F 2_{× H} D 1b A + E C B Yes - 2_{× H} D 2a A + E C B Yes G 2× H D 2b A + E C B Yes - 2_{× H} D 2c A + E C B Yes - 2_{× F} D 3a A + E C B Yes G 2× H D 3b A + E C B Yes G 2× H D 3c A + E C B Yes - 2_{× H} D 3d A + E C B Yes - 2× H D

B.2

Repeated Tests

Table B.2: Repeated tests for the C18 cartridge with 100 nmol/l 8-OHdG in blood plasma samples.

Between samples Peak area Recovery (%)

930 68

967 70

896 65

1045 76

971 71

(a) Analysis of five different blood plasma samples with the concen-tration 100 nmol/l.

Within samples Peak area Recovery (%)

1200 87

1270 92

1090 79

1027 75

(b) Analysis of the same blood plasma sample four times. Concen-tration 100 nmol/l.