Pathogen Detection Technique using Nuclear Magnetic Resonance by Daniel Slater A thesis submitted to the Graduate Faculty of Auburn University in partial fulfillment of the requirements for the Degree of Master of Science Auburn, Alabama May 5, 2013 Keywords: NMR Relaxation, Biosensor, MRI Nanotechnology, Nanoparticles, Superparamagnetic Copyright 2013 by Daniel Slater Approved by Barton C. Prorok, Chair, Associate Professor of Materials Engineering Dong-Joo Kim, Associate Professor of Materials Engineering Ron Beyers, MR Physicist and Safety Officer ii Abstract Nuclear Magnetic Resonance (NMR) signal relaxation is influenced by aggregate formation when magnetic nanoparticles bind with target molecules. In this study, a magnetic resonance based nanosensor was developed for a clinical MRI system. The goal of this research was the detection of target polystyrene microspheres conjugation with superparamagnetic iron oxide (SPIO) nanoparticles in a rapid and highly sensitive multiplexed fashion. The binding effects would occur due to the interaction of the streptavidin coated SPIO nanoparticles with the biotinylated polystyrene microspheres (BPM). The SPIO nanoparticles were used as a proximity sensor that would augment molecular interactions. The T2 dephasing rate of nearby water molecule protons was demonstrated to be lower in the dispersed state rather than once aggregated thus providing an adjustment in transverse relaxation time, ?T2. An optimal concentration range of the SPIO nanoparticles was shown for its greatest sensitivity at affecting T2 with any modification in the nanoparticles dispersion. By utilizing the optimal SPIO concentration, a NMR based nanosensor was developed that would identify the presence of the BPM within a solution. The SPIO nanoparticle sensor shows promise for its diagnostic capability and its portability. iii Acknowledgments I would like to thank Dr. Prorok for giving me the opportunity to work on this project and for his continuous support. I would also like to thank Dr. Ron Beyers from the Auburn MRI Center for his patient involvement with the MRI data collection and reduction. I also want to thank Marrianne Sullivan and Naveed Sidiqui for their assistance with supportive ideas and their suggested improvements. Finally, I want to thank all my committee members for their guidance throughout this research. iv Table of Contents Abstract ...................................................................................................................................................................... ii Acknowledgments ..................................................................................................................................................... iii List of Tables .............................................................................................................................................................. vi List of Figures ........................................................................................................................................................... vii List of Abbreviations .................................................................................................................................................. ix Chapter 1 Introduction ............................................................................................................................................... 1 1.1 Synopsis ................................................................................................................................................ 1 1.2 Thesis Structure .................................................................................................................................... 6 Chapter 2 Literature Review ....................................................................................................................................... 8 2.1 SPIO Nanoparticles ................................................................................................................................. 8 2.1a Comparison Between SPIO Nanoparticle Core Sizes ................................................................ 9 2.2b SPIO Nanoparticle Stability and Compatibility ....................................................................... 10 2.2 Detection Mechanism ........................................................................................................................... 12 2.3 Differences between NMR Diagnostic Platforms .................................................................................. 16 Chapter 3 Experimental Setup .................................................................................................................................. 22 3.1 Longitudinal Recovery (T1) Parameters ................................................................................................. 22 3.2 Transverse Relaxation (T2) Parameters ................................................................................................. 23 3.3 Magnetic Inhomogeniety Factor (T2*) Parameters ............................................................................... 23 3.4 Data Reduction ...................................................................................................................................... 24 3.5 SEM ....................................................................................................................................................... 24 Chapter 4 Test Procedures???????????????????????????????????????????...................25 4.1 Test One: Relaxation Time Dependence on SPIO nanoparticle Concentration???..??????????.25 v 4.2 Test Two: T2* Minimum SPIO Nanoparticle Concentration Detection???????????..???????.28 4.3 Test Three: Validity and Reliability of T2 and T2* .................................................................................. 30 4.4 Test Four: T2 as a Detection Mechanism ............................................................................................... 31 4.5 Test Five: Detection at a Higher SPIO Nanoparticle Concentration ..................................................... 34 4.6 Test Six: Confirmation of T2 Values with Increasing BPM Concentration ............................................. 35 4.7 Test Seven: Determining Particle Settling Using Sagittal Imaging ......................................................... 36 Chapter 5 Results and Discussions ........................................................................................................................... 39 5.1 Relaxation Time and Particle Core Size Comparison ............................................................................. 39 5.2 Reliability of T2 and T2* ......................................................................................................................... 41 5.3 Optimal SPIO Nanoparticle Concentration for T2 Target Detection ...................................................... 47 5.4 Detection of Polystyrene Microspheres (BPM) ..................................................................................... 49 Chapter 6 Conclusion ................................................................................................................................................ 55 Appendix .................................................................................................................................................................. 57 References .............................................................................................................................................................. 64 vi List of Tables Table 2.1 Proton Relaxivity Data for Different SPIO Particle Core Sizes and Coatings ............................................. 12 Table 3.1 T1 Inversion Times .................................................................................................................................... 23 Table 3.2 T2 Echo Times ........................................................................................................................................... 23 Table 3.3 T2* Echo Times ......................................................................................................................................... 24 Table 4.1 MRI Testing Procedures ........................................................................................................................... 25 Table 4.2 Sample 79 and 80 SPIO molar concentration ........................................................................................... 32 Table 5.1 Wet-to-Dry Percentage Change in T2* ..................................................................................................... 44 Table 5.2 Reliability of T2 and T2* ............................................................................................................................ 47 Table 5.3 External Magnetic Fields Effect on T2 for Equal Molar SPIO Samples ...................................................... 51 Table 7.1 Samples used for Test One with their respective SPIO core size and molar concentration .................... 57 Table 7.2 Relaxation rates of each sample in Test One ........................................................................................... 58 Table 7.3 Test Two Samples with their respective SPIO molar concentration. ....................................................... 58 Table 7.4 Test Three Samples with their respective SPIO molar concentration. ..................................................... 59 Table 7.5 Test Four Samples with their respective SPIO and BPM concentration .................................................. 60 Table 7.6 Test Five Samples with their respective SPIO and BPM concentration .................................................... 61 Table 7.7 Test Six Samples with their respective SPIO and BPM concentration ...................................................... 61 Table 7.8 Refitted Test One Sample Concentration Values .................................................................................... 63 vii List of Figures Figure 1.1 Schematic of Proton Spins Aligning to Magnetic Field .............................................................................. 2 Figure 1.2 Schematic of T2 and T2* decay ................................................................................................................... 5 Figure 2.1 Availability of Differing Particle Sizes to Attach to Target Particles ........................................................ 10 Figure 2.2 Dependency of Proton Relaxation Rate on MNP Coatings and Core Sizes .............................................. 11 Figure 2.3 Effects of Dispersed and Clustered MNPs on Surrounding Water Molecules ......................................... 13 Figure 2.4 Change in T2 due to MNP-Target Dissasembly ........................................................................................ 14 Figure 2.5 Selectivity of MNP Sensors ...................................................................................................................... 15 Figure 2.6 Image of an MRI ...................................................................................................................................... 17 Figure 2.7 Images of a NMR Systems Schematic Diagram, Microcoil, and Microfluidic Network ........................... 18 Figure 2.8 Expected Change in T2 with Increasing Target Concentrations .............................................................. 20 Figure 4.1 Image of Test Tube Rack Used for Tests One through Seven ................................................................. 26 Figure 4.2 Test One sample?s SPIO nanoparticles molar quantities ....................................................................... 27 Figure 4.3 SEM Images of the 30nm SPIO particles ................................................................................................. 28 Figure 4.4 Test Two sample?s SPIO nanoparticles molar quantities ....................................................................... 29 Figure 4.5 MR Image of Test Samples in Rack during Testing Procedures .............................................................. 29 Figure 4.6 Test Three sample?s SPIO nanoparticles molar quantities ..................................................................... 30 Figure 4.7 Diagram Indicating Predicated Changes in T2 at SPIO Optimal Concentration ....................................... 32 Figure 4.8 Image of Expected Binding Interaction between SPIO and BPM Molecules .......................................... 33 Figure 4.9 Polystyrene microsphere count with its respective SPIO optimal concentration .................................. 34 Figure 4.10 Polystyrene microsphere samples combined with SPIO optimal concentration .................................. 35 Figure 4.11 Reimaged Test Five sample microsphere quantity to verify the repeatability of their T2 values .......... 36 Figure 4.12 MR Image of Test Seven Sagittal Slice .................................................................................................. 37 viii Figure 4.13 Location of Each Sample in Test Seven Imaging ................................................................................... 38 Figure 5.1a Changes in T1 with SPIO Particle Size and Concentration ..................................................................... 40 Figure 5.1b Changes in T2 with SPIO Particle Size and Concentration ...................................................................... 40 Figure 5.1c Changes in T2* with SPIO Particle Size and Concentration .................................................................... 40 Figure 5.2 Comparison of 30nm SPIO particles Proton Relaxation Data ................................................................. 41 Figure 5.3 Explanation of transverse relaxation ....................................................................................................... 42 Figure 5.4 Calculated Test Two T2* Values ............................................................................................................... 43 Figure 5.5 Consistency between Test Two and Three T2* values .......................................................................... 44 Figure 5.6 Impact of Data Reduction Methodology ................................................................................................ 46 Figure 5.7 Predicted SPIO Optimal Concentration Range ......................................................................................... 48 Figure 5.8 Refitted Test One Sample Concentrations .............................................................................................. 49 Figure 5.9a BP Detection at .0000918nM SPIO Particle Optimal Concentration ..................................................... 50 Figure 5.9b BPM Detection at .0001836nM SPIO Particle Optimal Concentration .................................................. 50 Figure 5.9c BPM Detection at .0002754nM SPIO Particle Optimal Concentration .................................................. 50 Figure 5.10 Comparison of T2 Changes between Top and Bottom of SPIO-BPM Samples ...................................... 52 Figure 5.11 Bar Graph of Figure 5.9 ......................................................................................................................... 53 ix List of Abbreviations MNP Magnetic Nanoparticle SPIO Superparamagnetic Iron Oxide BPM Biotinylated Polystyrene Microsphere MRI Magnetic Resonance Imaging NMR Nuclear Magnetic Resonance IO Iron Oxide SEM Scanning Electron Microscope ROI Region of Interest RF Radio frequency T1 Spin-Lattice Relaxation/Longitudinal Relaxation T2 Spin-Spin Relaxation/Transverse Relaxation T2* Apparent transverse Relaxation Time/Magnetic Inhomogeneity Factor FID Free Induction Decay TE Echo Time TR Repetition Time WB Water Bath 1 CHAPTER 1 INTRODUCTION 1.1 Synopsis Pathogen detection is a necessity in fighting human infection for food product industries and biodefense. Nuclear Magnetic Resonance (NMR) relaxation and Magnetic Resonance Imaging (MRI) have proven to be advantageous in comparison to other detection mechanisms to date such as quantitative Polymerase Chain Reaction (PCR) techniques, Raman spectroscopy [1, 2], fluorescence [1, 3], and colorimetry [1]. The development of new diagnostic platforms that utilize nuclear magnetic resonance (NMR) relaxation has garnered much attention due to their ability to measure bimolecular abundance with high sensitivity [4, 5], rapid disease detection [4, 6], and understanding of pathogen biology at the systems level [7]. It is necessary for these newly developed mechanisms to have high target specificity, almost instantaneous response time, repeatable identification, and capability to detect numerous varying target pathogens. Nuclear magnetic resonance (NMR), is a phenomenon that occurs for nuclei possessing quantum spins that are subjected to a static magnetic field and then exposed to an additional second oscillating resonant magnetic field [8, 9]. The individual protons of the hydrogen atoms are the most commonly used particles in clinical applications that require this spin property [8]. Each proton will possess its own magnetic moment and will produce its own NMR signal when subjected to this phenomenon [8]. Nuclear magnetic resonance relaxation is the study of the interaction of electromagnetic radiation with matter when the specimen is exposed to a static magnetic field. A net magnetization is induced by the applied magnetic field. Normally, without the external magnetic field, the net magnetization is zero. The net magnetization is created by the sum of the spins such that each has their own precision-frequency. Quantum spins are found in protons, electrons and neutrons. Each unpaired electron, proton, and neutron possesses a spin of one half. In NMR, it is unpaired nuclear spins that are of importance. 2 For NMR of water, the spins of the hydrogen nuclei form a net magnetization aligned with an applied static magnetic field (B0). At the resting state, the net magnetization vector of the protons is a combination of their spins aligned longitudinally (parallel to B0) and having no net transverse magnetization components due to spins being out of phase. Figure 1.1a illustrates random spin orientations with no B0 and 1.1b show how the quantum spins can either align with the field or in the opposing direction (anti-parallel). There are more spins aligned with the external magnetic field (low energy) than against it (high energy). As a result, the net magnetization has a longitudinal component aligned with the B0 field while the transverse magnetization component is zero due to the spins rotations being in random phase. Longitudinal = ?nuclear spins Transverse (rotation) = 0 Figure 1.1 (a) Disordered water proton spins (b) Alignment of protons spins to magnetic field (B0) direction Hydrogen (in water) is chosen for clinical applications due to its abundance in biological tissues. While being controlled by the direction of the magnetic field, the protons will also precess about the axis of the field?s direction. The precession frequency or the Larmor frequency of the protons is directly controlled by the applied field strength. More specifically, the precession frequency of each proton is directly controlled by its local magnetic environment and can be described by the Larmor equation found in Eq. 1 below [4]. (a) 3 (1) When placed in a magnetic field of strength Bloc, a particle with a net spin can absorb a photon of frequency ?. The frequency ? depends on the gyromagnetic ratio, ? of the particle [10]. Therefore each particle can only absorb a specific frequency depending on its ?. Their spins do not cancel each other out and a net spin is produced. The protons spin is its own property that gives it a magnetic moment or magnetic vector. When placed in a magnetic field, the protons magnetization vector will align itself with the external magnetic field. To change the spin equilibrium of the protons, a Radio Frequency (RF) pulse is applied to the specimen. The applied RF pulse matches the precision frequency of some of the protons causing a change in the net magnetization vector and energy in the system. In a Magnetic Resonance Imaging (MRI) system, the RF pulse is tuned to the resonance frequency of the hydrogen proton. The application of the RF pulse to the water molecule?s hydrogen proton causes it to align itself in the direction of the RF signal. The directional component of the specimen?s magnetization is shifted from primarily longitudinal to only transverse. After the RF pulse is removed, the transverse magnetization of the protons decays and the longitudinal component recovers until the equilibrium position (static field direction) is reached. The resonance frequency of the proton is the frequency at which the system is able to store and transfer energy with its greatest efficiency. To induce the shifting of directional magnetization components stated above, the electromagnetic RF pulse produced by an MRI must match the frequency of the resonance frequency of the hydrogen protons in order to induce the desired response for imaging. The electromagnetic energy is absorbed momentarily by the atomic nuclei causing a resultant change in spin equilibrium. Once the RF pulse is removed, the nuclei will regress from this new excited state back to its original equilibrium state over a period of time. This period of time is known as the relaxation time of the nuclei and is described as the time required for its transition between the transverse magnetization and longitudinal directions. The NMR signal collected by the MRI is formed by the electromagnetic energy released by the nuclei over its respective relaxation times. 4 The dynamic physical phenomenon of water protons spins transitioning back to their equilibrium positions is known as the relaxation of spins [8]. The system of spins relaxation to equilibrium is broken down into two components: 1. T1 ? recovery of longitudinal magnetization, aligned with the magnetic field 2. T2 ? decay of transverse magnetization, due to spins getting out of phase Longitudinal relaxation (T1) is also known as spin-lattice relaxation and is caused by longitudinal relaxation recovery. The difference in the number of spins in parallel and anti-parallel directions is what contributes to longitudinal magnetization. For T1 relaxation, the proton spins relaxation causes a quantum emission of electromagnetic energy that is exchanged with the surrounding lattice. After RF pulse excitation, the precessing proton spins emit RF energy and revert back from a higher energy state to a lower state. The relaxation time constant T1 can be calculated in Eq.2 by using the measured equilibrium magnetization (A0) and longitudinal magnetization (A). The variables A and A0 can also be referred to as the MR signal amplitude and MR signal amplitude at time (t) = 0 respectively. The time to diminish the difference between the longitudinal magnetization and the equilibrium magnetization is known as the spin-lattice relaxation time (T1) [9]. As the static magnetic field strength is increased the values for T1 also increase. Spin-lattice relaxation is performed by a 180? RF pulse using an inversion-recovery pulse sequence. As the spins rephase, the transverse magnetization grows as the longitudinal magnetization decays. (2) The second relaxation spin parameter that can be measured during NMR relaxation is the transverse relaxation (T2). Transverse magnetization is caused by spins getting into phase coherence. The precession rate of the spins is altered by the interaction of their individual magnetic fields (spin-spin interaction). To find T2 of a specimen a spin-echo procedure is commonly used that utilizes a 90? RF pulse. The transverse magnetization of the spins slowly decays during spin-spin relaxation causing a loss in phase. Spin-spin relaxation (T2) can be calculated with Eq. 3. The T2 relaxation rate is more strongly influenced by the local magnetic environment than T1. For this 5 reason contrast agents are commonly used during MR imaging in order to change the local magnetic environment. By altering the magnetic environment the precessional frequency of the local nuclei is changed according to the Larmor equation. Although T2 values are unrelated to the magnetic field strength, the stronger the field the more rapid the spins precession, thus changing the deviation in spin phases. Typically T1 tends to be longer than T2. This can be seen by observing the width of the received NMR signal. The very quick spin-spin relaxation time creates very broad signals unlike T1. The time limit of the signal or the signal line width in the transverse plane is represented by T2. The line width measured for an NMR signal is used to find T2 by measuring the line width at half-maximum. The line width at half maximum is equal to 1/T2. Although the originally measured line width can determine a T2 value, the value is dependent on the inhomogeneity of the magnetic field. By using the experimental relaxation rate, a different form of T2 known as T2* can be calculated (Eq. 4). Where T ? is the change in relaxation time due to magnetic field inhomogeneities. Figure 1.2 Schematic representation of T2 and T2* decay values derived from the MR signal collected. [25] The inhomogeneity factor, T2* is more important for nuclei of higher precessional frequency. This can be explained by examining the Larmor equation. The resonance frequency for each nucleus is altered as the magnetic field varies through the specimen. Therefore, as the field varies, the frequency and the resultant NMR signal line 6 width also vary. To measure T2*, a gradient-Echo sequence is typically used. The differences between a gradient echo from a spin ?echo procedure is that the flip angle tends to be below 90? and the absence of a 180? RF rephrasing pulse. Without the presence of a magnetic gradient, the NMR signal produced is called Free Induction Decay (FID). T2* is characteristic of the FID signal and will decrease exponentially as the FID signal decays. The relaxation time constant T2* is found to be much faster than T2 would predict. The differences between the two is that T2* takes in to account the tissue specific spin-spin relaxation responsible for pure T2 decay and is more sensitive to magnetic field inhomogeneity?s. It is this second feature that tends to cause inconsistencies in the values for T2*. (3) (4) 1.2 Thesis Structure Chapter 2 reviews all subjects related to this thesis. First, it discusses the uses and characteristics of magnetic nanoparticles. The chapter provides an in depth look at the behavior of MNP?s under an applied magnetic field and their influence on other particles. It describes how MNP?s are used as a detection mechanism and their compatibility with varying environments. Finally, the chapter explains the similarities and differences between the different diagnostic platforms used for NMR relaxation. Chapter 3 contains the experimental setup of this research. It provides the parameters used for all proton relaxivity data acquisition and SEM imaging. A short description of each parameter and its effects on the detection sensitivity, data reduction, and relaxation values is given. The technique utilized for the data reduction of each proton relaxation variable is also addressed. Chapter 4 illustrates the procedure of each test performed during the course of the research. This chapter lists how and why each of the tests was conducted. It contains images of how the samples were positioned during 7 MR imaging; in addition to the values for each of the sample measurements. An explanation for the methodology of each procedure is also stated below in chapter 4. Chapter 5 explains the results of each test performed by the MRI. The impact of the MNP core size on the proton relaxation data is described here, along with the necessary requirements for each relaxation variable to be used in pathogen detection. The reasoning for the selection of T2 instead of T2* as a viable parameter for target recognition is also discussed in this section. The data reduction of the research is further examined in order to compare the necessity for consistent data reduction methods and the reliability of the relaxation time values. The data collected from tests involving target detection provided an interesting discovery. The results of these tests were revealed to be of conflicting values than what was shown in previous reports. Finally, theories that were proposed for this unearthing were evaluated, thus finishing the chapter. Chapter 6 concludes the research by providing a brief summary of the resultant understanding gained during this research. The final statement demonstrates that the detection resulted from the desired binding between the magnetic nanoparticles and the target molecules. The summary conveys the desired optimal MNP concentration for superior target sensitivity and its dependency on target molecule core size for the NMR detection mechanism. An Appendix can be found at the end of the paper for all tables containing sample concentrations listed in Chapter 4. 8 CHAPTER 2 LITERATURE REVIEW 2.1 SPIO Nanoparticles Nanoparticles have received increasing attention over the past few decades due to their unique characteristics, such as large surface-to-volume ratio and size-dependent optical and magnetic properties [4]. Their nontoxic nature along with their nanoscale dimensions hold much promise for further development in nanotechnology [11]. Nanoparticles provide specificity and multivalent affinity due to their ability to conjugate with specific ligands capable of attaching to their target molecules. This interaction allows for highly sensitive and selective diagnostic mechanisms for clinical use. It has been observed that these interactions can be detected by Raman spectroscopy, fluorescence, and colorimetric spectroscopy [4]. The relaxation value of water protons is greatly influenced by the local magnetic field surrounding the particles. SPIO nanoparticles have been reported to alter the magnetic resonance signal of surrounding water molecule protons by changing this local magnetic environment [4].The effective presence caused by magnetic particles on the relaxation time of water molecules operates at a significant distance from the particle [10]. ?Evidence of this is found in the observation that MR images of livers containing the particles show reduced transverse relaxation time T2 across the entire organ even though the phagocytic Kupffer cells containing the particles comprise only 2% of the mass of the liver? [10]. The enhancement of the relaxation time of water protons is due to the diffusion of water molecules between different magnetic environments created by the SPIO nanoparticles. This relationship between the spin- spin relaxation of hydrogen protons and the diffusion of the water molecules into areas of differing magnetic environments is represented by Eq. 5, where D, G, and are the spin diffusion constant, linear gradient, and the time between pulses respectively. The coherence of the phases of the precessing protons diminishes at a rate according to the local magnetic environment. As the water molecules diffuse into different areas, their alteration 9 in phases is controlled by the variation of the Larmor frequencies between each environment. These fluctuations in differing magnetic environments are the direct cause that modifies the resulting magnitude of the NMR signal received, hence to a varying T2 [10]. )2 (5) 2.1a Comparison Between SPIO Nanoparticles Core Size Research supports that different particle sizes can have a contrasting effect when influencing the local magnetic environment of surrounding water molecules. Contemporary research focuses on the use of nanoparticles by replacing the previously used micron sized particles. Unlike nanoparticles, microparticles have lower efficiency and poorer binding ability due to their comparatively lower surface to volume ratio [12]. Nanoscale dimensions offer a wide array of advantages over micron size particles. First, microbeads typically take longer to interact with the target molecules in suspension [12]. Second, micron size particles have proven not to be completely stable in suspension and can cause unwanted precipitation leading to inefficient separation and detection [12, 11]. Third, by reducing the particle dimensions and increasing the surface to volume ratio, the available adsorptive areas increase drastically (Figure 2.1) [12, 13]. Fourth, nanoparticle?s smaller size allows for easier diffusion to areas of denser target molecule concentration thereby increasing their interaction [12]. Finally, under an applied magnetic field, larger particles tend to aggregate causing loss in signal detection [4, 14]. 10 Figure 2.1 Diagram comparing the availability of differing particle sizes to be attached to an immunomagnetically labeled cell. The particles with a smaller sized diameter are more capable of attaching on the cell surface than larger (micron) sized particles because of less steric exclusion. [12] 2.1b SPIO Nanoparticle Stability and Compatibility The most problematic issue regarding nanomaterials used for biological applications is their compatibly with different biological environments. Research has shown it is imperative to design these materials against extreme conditions to prevent potential signal loss [12]. Without proper magnetic nanoparticles (MNP) preparation, the nanoparticles can potentially oxidize, hence causing the formation of antiferromagnetic oxides resulting in a significant loss in magnetic moment [15]. Increasing concern for the synthesis of air-stable nanoparticles with high saturation magnetization capable of withstanding extreme conditions has thus developed for the optimization of contrast agents [15, 16]. Efforts have been made to prevent any loss in magnetic moment by improving the coating layer surrounding the nanoparticle metal core [15, 16]. Examples of these preventive measures consist of inserting the metallic particles into polymer matrixes or passivating them with an oxide shell, gold, or carbon [15]. Iron oxide nanocrystals are commonly coated with hydrophobic surface ligands such as oleic acid [11]. The hydrophobic oleic acid layer provides surface iron lattices analogous to bulk structures and strong binding of the layer to the iron sites [11]. As a MRI contrast agent, the layer protects the particle surface from water proton interaction, causing a decrease in the average proton relaxivity [11]. The hydrophilic polymer coatings help prevent magnetic field assisted particle/particle aggregation of MNPs [14]. The measurable proton relaxation rates are 11 reliant on hydrophilic characteristic of the coating layer [11]. Figure 2.2 illustrates that the iron oxide nanoparticles coated with PEI exhibit the highest proton relaxivity R2. Proton relaxation rate (R2) is the reciprocal value of the spin-spin relaxation time variable T2. Nanoparticles coated with the oleic acid proved to have a much smaller R2 value than that of the PEI coating [11]. Figure 2.2 Dependency of the proton relaxation rate on the iron oxide contrast agent?s coating layer hydrophilicity and particle size. [11] Different coatings on MNPs can alter the hydrodynamic size of the particle. For example, an iron oxide (IO) particle with a 30 nm core was found to have a hydrodynamic size of 56.6nm when coated with PEG-g-PEI, but smaller hydrodynamic sizes of 44-45 nm when coated with PEI or PMO [11]. Furthermore, it was reported that proton relaxivity is greatly influenced by the size of the particles. Smaller particles with a core size of 10 nm were shown to have smaller relaxation rates than larger 30nm core size particles. Even with different particle coatings these results remained consistent. It was suggested that the results were due to differences in surface spin anisotropy, which tends to be more effective for smaller core size particles because of their larger surface area to volume ratios [11]. 12 Table 2.1 Proton Relaxivity Data comparing its impact on SPIO Nanocrystal Core Size and Polymer Coating [11] PMO coated nanoparticles PEG-g-PEI coated nanoparticles PEI coated nanoparticles relaxivity (s-1mM-1) Deionized H2O 10 nm 30 nm 10 nm 30 nm 10 nm 30 nm r1 0.01 7.8 10.9 11.2 14.1 21.5 29.0 r2 0.54 27.2 70.5 39.8 93.3 75.2 107.3 r2* 0.51 28.2 72.2 34.1 87.5 71.3 98.7 r1 = longitudinal relaxation rate; r2 = traverse relaxation rate; and r2* = effective traverse relaxation rate. As shown in Table 2.1, Magnetic nanoparticles influence on proton relaxivity appear to be highly dependent on their surface coating, particle size, and the hydrophilicity of inner capping ligands. These three factors are imperative when selecting nanoparticles for diagnostic platforms involving biological media. The SPIO particles in this study where chosen to have nanoscale dimensions and a protective polymer coating to prevent oxidation and increase detection sensitivity. 2.2 Detection Mechanism NMR detection is based on aggregate formation (Figure 2.3) or dissociation (Figure 2.4) when MNPs bind to the desired target molecules [4]. The streptavidin coated SPIO nanoparticles readily bind with the target biotinylated polystyrene microspheres (BPM). The biotin-streptavidin conjugation was used in this study to mimic the binding interactions that would occur between an antibody and an antigen. The spin-spin relaxation time (T2) is altered due to the formation of these clusters which cause the dephasing rate of nearby proton spins to increase [4]. The precessional frequency of a protons spin is strongly controlled by the local magnetic environment surrounding the water molecule?s hydrogen proton according to the Lamor equation [4]. SPIO nanoparticles will cause changes in the local magnetic fields homogeneity thereby altering the dephasing of nearby water protons. The SPIO nanoparticles can enhance the return of the water protons magnetization directional components from 13 the excited state back to the equilibrium state (static field direction), thus shortening their relaxation time values [4]. The aggregation of SPIO nanoparticles by SPIO-BPM interactions produces stronger local magnetic fields due to the coupling of SPIO nanoparticles magnetic moments. This alteration of the local magnetic environment is the basis of NMR detection. The mechanism aims at determining the presence of the specific targets by measuring changes in spin-spin relaxation (T2) after the formation or dispersion of the SPIO-BPM aggregates [4, 6, 14, 17]. Figure 2.3 Illustration describing the effects of superparamagnetic contrast agents on surrounding water molecule protons. (A) The alignment of the MNP?s magnetic moment to the direction of the applied field, which creates local magnetic fields causing an increase in spin-spin relaxation of the local protons. (B) Formation of nanoparticle clusters creates larger local magnetic field inhomogeneities thereby further increasing T2 relaxation of nearby protons. (C) Nanoparticles are aligned in a linear array along a nanowire. The clustering of the nanoparticles in this format further improves the coupling of the nanoparticles magnetic moments thus improving proton relaxation. [18] 14 Figure 2.4 The disassembly of MNP-Lys aptamer clusters after the addition of the lysozyme protein. Detection was proven from the resulting increase in T2 relaxation time with increasing lysozyme concentration additions. (a) Determination of the lysozyme detection range. (b) T2 weighted MR images of increasing lysozyme concentration. [4] The utilization of these NMR techniques allows for electronic data acquisition instead of previously used bulky optical components. The signal detection is much faster than with surface structure-based devices because of their ability to detect in the specified area instead of relying on the diffusion of the targets to the sensing elements [4, 5, 6]. This technology has the potential for easier, quicker, and higher-throughput sensing operations than its predecessors. Assays that employ these NMR techniques for signal detection measurements also have the capability to be executed in turbid samples with few or no preparation steps [5, 17]. A consequence of this mechanism is the precipitation of the molecules after the formation of the SPIO- BPM clusters. The settling of the SPIO-BPM clusters could reduce the average number of water hydrogen protons that could diffuse between the SPIO nanoparticles magnetic domains. Thus the precipitation of the SPIO-BPM clusters could potentially alter the acquired relaxation values thereby reducing the validity of the BPM detection. By combining the SPIO nanoparticle sensor with lower concentrations of the target molecule (BPM), the precipitation of the clusters can be avoided [4, 19]. Although this method was shown to prevent signal loss due to 15 unwanted settling, the measurements performed were limited to only target molecules of very low molecular weights. Another problem to avoid in SPIO nanosensors is the assembly or disassembly of SPIO clusters due to random binding sequences other than that of the target sequence (streptavidin-biotin) [4, 17, 19]. SPIO nanoparticles are capable of binding with various ligands that have specific binding characteristics to provide specificity and multivalent affinity [4, 5, 17, 20]. The selectivity of the magnetic nanosensors is an imperative parameter to be considered. The example shown in Figure 2.5, displays previous research that has proven that these complications (non-specific binding) can be reduced by improving the selectivity of the target molecules which can provide only slight alterations in T2 [4]. The interference with detection of the target molecules (BPM) in biological media is highly dependent on the selectivity of the SPIO nanoparticles. Figure 2.5 The change of T2 after the addition of the lysozyme target molecule. The chart was to prove the selectivity of the magnetic nanosensors from binding from other undesired interactions that could interfere with the lysozyme detection. [4] SPIO nanoparticles have shown to remain highly stable even when exposed to large temperature fluctuations and different ionic media [5, 6, 21]. Small differences in target conjugations could be detected by T2 measurements within a multitude of different buffer conditions. Research has demonstrated that diagnostic 16 platforms utilizing NMR relaxation techniques are capable of performing selective measurements reliably and at various temperatures without the interference of T2 measurements from target mismatches [5, 6]. The SPIO nanosensor can detect the target BPM in solution by measuring the changes in spin-spin relaxation (T2) upon SPIO-BPM conjugation. The interaction of the SPIO-BPM molecules will create larger clusters that can modify the original (SPIO nanoparticles only) T2 values. It is vital that the resultant T2 changes occur primarily from specific binding (streptavidin-biotin) between the SPIO nanoparticles and the target pathogen (BPM) rather than by nonspecific effects. 2.3 Differences between NMR Diagnostic Platforms NMR relaxometer design has the potential to increase detection sensitivity, lower sample volume, provide multiplexity, and increase signal-to-noise. To date, Benchtop NMR and clinical systems have been the most researched as NMR detection assays [4, 5]. Clinical systems such as MRI?s allow for parallel measurements at higher throughput with a minimum sample volume of 50?l [6, 22]. The average cost of running an MRI test usually costs around $500 per hour. The sensitivity and throughput of an MRI assay can be enhanced through increasing the number of well plates or by decreasing the total sample volume [4]. An increase in the number of well plates has the potential to improve the quantity, quality, and swiftness of data collected [4]. Typically MRI?s have a magnetic field strength of 1.5 to 3.0 Tesla, but there are more powerful MRIs available with field strengths exceeding 11.4 Tesla. At their lowest strength of 1.5 Tesla, SPIO nanoparticles are almost entirely saturated and create the desired, local magnetic inhomogeneities [13]. 17 Figure 2.6 Image of a closed (traditional) MRI scanner [23] Benchtop relaxometers can have magnetic field ranges from .17 to 1.41 Tesla with frequency ranges from 7.5-60 MHz. The proton NMR frequency and magnetic field strength influence both the detection sensitivity and the overall testing time. Another benefit found with most benchtop relaxometers is their capability of controlling the sample temperatures during testing. Heating elements with temperature sensors can be used for better temperature control [5]. These heating elements can also be added to the microfluidic systems of chip sized NMR sensors [5]. By miniaturizing an entire NMR system even further, the mechanism could potentially perform measurements on even smaller sample volumes and in a multiplexed fashion [5, 17]. Miniaturized systems fitted with microfluidics allow for easier control and manipulation of small volumes of liquid. In addition to this, it can reduce the signal to noise ratio, lessen the need for spatial homogeneity of the magnetic field. Finally, it offers additional magnetic separation of target molecule concentrations from a parent specimen [5]. Multiple planar microcoils are typically arranged in an array to perform multichanneled detection instead of the common field gradient coil design [5]. This design allows for easier device miniaturization and portability. Miniaturized sensors are designed to be self-contained and portable devices and have many other potential methods for further size reduction [17]. 18 Figure 2.7 (a) The binding of MNP?s to their targets forming clusters. The formation of these clusters induces larger magnetic inhomogeneities, thereby increasing the dephasing rate of the nearby water proton spins and reducing the spin-spin relaxation time (T2). (b) A NMR systems schematic diagram. (c) An image of a microcoil used for the NMR system. (d) An image of a microfluidic network. (e) Diagram of the NMR electronics used in the NMR system to perform T1 and T2 measurements. [5] In one report researchers developed a chip-NMR biosensor capable of perform eight-multiplexed measurements in 5-10?l sample volumes [5]. A schematic of the chip-NMR system and its microfluidic system is shown above in Figure 2.7c and Figure 2.7e. The NMR system was fitted with microfluidic networks that provided chaotic advection through the meandering channels, allowing for easier manipulation of MNP and target molecule mixing. The chip-NMR biosensor permitted more homogeneous radio-frequency magnetic fields with less electrical 19 resistance. It could also detect multiple targets in mass-limited samples and reduced the usage of expensive reagents. The chip-based DMR (diagnostic magnetic resonance) system was reported to have a sensitivity of as low as 1 x 10-12M. It was capable of performing sensitive detection and multiplexed identification of different target molecules. Their design was compared to a benchtop system with a similar magnetic field strength. Both systems measured the T1 and T2 relaxivities of water containing MNPs bound with their target molecules using inversion- recovery pulse and CPMG sequences respectively. The results of both tests showed excellent agreement (Figure 2.8). Standard benchtop NMR systems require sample volumes greater than 300?l and can only process one sample at a time [5]. Unlike the benchtop system, the chip-NMR sensor from the design above can have a sample volume as low as 5?l per coil. Multiplexed measurements can be performed by each coil independently and is theoretically only limited by the number of coils fitted on each chip [17]. Chip-NMR sensor?s permanent magnets are portable and are capable of producing polarizing magnetic fields, Bo =0.1 -0.5T. The chip-NMR system greatly outperformed the benchtop relaxometer mass sensitivity. It achieved a minimum detection of 1.0ng of avidin in a 5?l sample. The improved mass sensitivity can be attributed to the device miniaturization of the microcoil, which enabled stronger radio frequency field generation with increased filling factors in small sample volumes [5, 17]. 20 Figure 2.8 (a,b) Graphs indicate a comparison between a commercial benchtop relaxometers and the newly developed DMR systems measurements of T1 (a) and T2 (b). (c) The resultant change in T2 with increasing additions of avidin concentration to biotinylated mangetic nanoparticles. (d) A closer inspection of the minimum mass-sensitivity capable by the DMR system for detection of the target molecule. [5] Miniaturizing the source of the external magnetic field produces complications. One problem is a low NMR signal level intensity is collected due to the small sample volume and low magnetic field strength [5]. Microfluidic systems onto chip-NMR systems have shown to overcome these obstacles. Low signal-to-noise ratio has commonly been attributed to the design of the planar coils because of their high series resistance [24]. NMR probes consisting of electroplated planar microcoils integrated on a glass substrate with etched microfluidic channels are a common design for chip sized relaxometers. Reducing the diameter of the microcoils has indicated potential to improve the mass-sensitivity of the spectrometer [24]. An example of an actual microcoil used for NMR detection can be found in the image above in Figure 2.7c. Radio frequency magnetic fields are produced by the microcoil in order to induce and an excitation of the samples and receive their resulting NMR signal. The material selected for the metal lines electroplated on the microcoil is imperative for minimizing the electrical resistance and hence thermal noise of the microcoil [5]. The reduction of the thermal noise is vital for preventing its interference with the NMR signal collected by the microcoil. 21 Currently, the largest obstacle inhibiting sensitivity is designing the spectrometer to prevent probe- induced static magnetic field distortions. By minimizing the size of NMR detection coils, the mass-sensitivity has proven to increase drastically for mass-limited samples [24]. The majority of micro-scale NMR diagnostic platforms focus on solenoid-type microcoils wound around a capillary. To improve this design the planar microcoils can be built using processes based on photolithography techniques [24]. The utilization of micromachined coils has numerous advantages, such as the ability to fabricate micron sized planar coils and the availability of chip-based microfluidic system integration. Further developments in microfabrication can lead to possible improvements for NMR spectrometer design by enhancing sensitivity, lowering cost, and increased portability. 22 CHAPTER 3 EXPERIMENTAL SETUP The relaxation times of each the samples were measured using a human knee Xmit/Rcv RF coil in a clinical Siemens Verio 3 Tesla MRI Scanner. Proton density, Repetition time (TR) and Echo Time (TE) are some of the conditions and sequence parameters that affect the overall acquisition time and the imaging contrast. The Echo Time (TE) is the time between the application of the RF excitation pulse and MR echo signal collection. The MR echo signal is collected at time TE after excitation. The time between RF excitation pulses is the repetition time (TR).The repetition times were selected in order to guarantee the recovery of the transverse magnetization back to the longitudinal equilibrium condition. A long TR time allowed for maximum signal available for the next TR. This gave less T1 weighting, and more T2 and proton density weighting. The tradeoff for improving the results due to a long TR, is the lengthening of the overall scan time. The sequences performed applied a constant TR and varied the time of inversion (TI) for T1 scans or TE for T2 and T2* scans listed below under each variables sub-section. 3.1 T1 Inversion Recovery (T1) Parameters: The sequence type chosen for T1 measurements for all MR testing was a non-selective adiabatic inversion with slice-selective spoiled gradient-echo acquisition. A standard coronal slice of approximately 5mm was used to create the slice images of each of the samples. Each sample image was fitted within a 128 x 128 pixel matrix to encompass the entire sample area. A total of 54 time of inversion (TI) echo times ranging from 5.6 to 3000 ms were used for each sample with a TR time of 5 seconds between each RF inversion pulse. The total acquisition time was 10.6 hours for the entire procedure. The readout flip angle was performed at 60 degrees and a readout bandwidth of 289 Hz/Pix. The inversion times (TI) used for T1 imaging can be found in Table3.1 below. 23 Table 3.1 Inversion Times (TI) used for imaging of T1 (ms) 5.6 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0 100.0 110.0 120.0 130.0 140.0 150.0 160.0 170.0 180.0 190.0 200.0 220.0 240.0 260.0 280.0 300.0 320.0 340.0 360.0 380.0 400.0 425.0 450.0 475.0 500.0 550.0 600.0 650.0 700.0 750.0 800.0 850.0 900.0 950.0 1000.0 1200.0 1400.0 1600.0 1800.0 2000.0 2250.0 2500.0 2750.0 3000.0 3.2 Transverse Relaxation (T2) Parameters: To calculate the T2 values of each sample a slice-selective spin-echo MRI sequence type was employed. Similar to the T1 measurements, a 5 mm coronal slice with an image matrix size of 128 x 128 pixels was used to create each sample image. The acquisition time of the entire procedure was roughly 9.25 hours for the 52 spin- echo times (TE) collected and a TR time of 5 seconds between each RF excitation pulse. The excitation flip angle was 90 degrees, refocus pulse was 180 degrees, and a readout bandwidth of 355 Hz/Pix. The echo times (TE) used for T2 imaging can be found in Table 3.2 below. Table 3.2 Echo Times (TE) used for imaging of T2 (ms) 7.0 7.5 8.0 8.5 9.0 9.5 10.0 11.0 12.0 13.0 14.0 16.0 18.0 20.0 25.0 30.0 35.0 40.0 45.0 50.0 60.0 70.0 80.0 90.0 100.0 125.0 150.0 175.0 200.0 225.0 250.0 275.0 300.0 350.0 400.0 450.0 500.0 550.0 600.0 650.0 700.0 750.0 800.0 850.0 900.0 950.0 1000.0 1100.0 1200.0 1300.0 1400.0 1500.0 3.3 Magnetic Inhomogeneity Factor (T2*) Parameters: A slice-selective spoiled gradient-echo MRI sequence type was chosen for all T2* measurements. Each image slice was 4 mm thick and created an image matrix size of 128 x 128 pixels. More echo times (TE) were used for T2* measurements than that of T1 and T2 to ensure sufficient data collection. The procedure was completed 24 within 4.25 hours to cover the 61 TE chosen and a TR time of 1.0 sec between each RF excitation pulse. The excitation flip angle was 90 degrees and a readout bandwidth of 488 Hz/Pix. The TE used for T2* imaging can be found in Table 3.3 below. Table 3.3 Echo Times (TE) used for imaging of T2* (ms) 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0 8.5 9.0 9.5 10.0 11.0 12.0 13.0 14.0 15.0 16.0 17.0 18.0 19.0 20.0 22.0 24.0 26.0 28.0 30.0 32.0 34.0 36.0 38.0 40.0 45.0 50.0 55.0 60.0 65.0 70.0 75.0 80.0 85.0 90.0 95.0 100.0 110.0 120.0 130.0 140.0 150.0 160.0 170.0 180.0 190.0 200.0 225.0 250.0 275.0 300.0 3.4 Data Reduction Data reduction of the acquired MRI signal intensities was performed for every test procedure and relaxation time. The T2 and T2* values were calculated by plotting the MRI signal intensities versus echo times according to Eq. 3 and 4 respectively. A decay curve was fitted for each samples data series plot. The resultant equation from the fitted trend line was used to extrapolate the values for T2 and T2*. A nonlinear curve fitting method based on Eq.2 was used to determine the magnitude of the MRI signals at different inversion time points, and to calculate the T1 relaxation time of each sample. 3.5 SEM SEM images were taken by a JEOL JSM 7000F SEM instrument. The samples were deposited onto a copper 400 mesh grid followed by drying. Afterwards, the samples imaged under a magnification of 70,000-X at an accelerating voltage of 20 kV. No sputter coating was required since the particles were already conductive. 25 CHAPTER 4 TEST PROCEDURES Table 4.1 Test Procedures Test Number MRI Scanning Procedures Test One MRI scan of different SPIO nanoparticle core sizes and relaxation rates Test Two MRI scan to determine T2* dependence on SPIO nanoparticle concentration Test Three MRI scan to test the validity and reliability of T2 and T2* Test Four MRI scan to detect polystyrene microspheres (BPM) using T2 Test Five MRI scan to verify the optimal concentration Test Six MRI scan to confirm the T2 values and the trend to increase with BPM concentration Test Seven MRI scan to validate particle settling during imaging using T2 4.1 Test One: Relaxation Time Dependence on SPIO Nanoparticle Concentration Test One samples included 38 samples, 4 of which were pure water samples, 16 included only the 10nm MNP, and the remaining 18 contained only the 30nm MNP. The magnetic nanoparticles were superparamagnetic iron oxide (SPIO) nanoparticles with an amphiphilic polymer coating conjugated with streptavidin. These were acquired from the company Oceannanotech. SPIO nanoparticles tested were of two different inorganic core sizes, 10nm and 30nm. Each nanoparticles hydrodynamic sizes was about 8-10nm larger than its inorganic core size. All samples used were sonicated prior to MRI testing. 26 Figure 4.1 Image of the test tube rack and test tube samples used for Tests One through Seven. The first test was aimed at determining the changes of the T1, T2, and T2* relaxation times with increasing concentrations of the SPIO nanoparticles. Measured concentrations of the SPIO nanoparticles were first dispersed into test tubes containing deionized water. Each test tube sample for all tests contained 1ml of deionized water. The samples containing the SPIO nanoparticles increased over a range of 18 SPIO nanoparticle concentrations from low to highly concentrated. The range of SPIO nanoparticle concentration for the 10nm and the 30nm were of similar molar values (Figure 4.2). 27 Figure 4.2 Test One sample?s SPIO nanoparticles molar quantities Two different testing procedures were used in the experiment. These procedures measured the respective T1, T2, and T2* relaxation values of each of the SPIO nanoparticles. Multiple properties of the SPIO nanoparticles were calculated and measured in test one. The first property tested measured the differences between the particle sizes and their individual relaxation values. The second targeted advantages and disadvantages between T1, T2, and T2*. It was necessary to compare their accuracy, consistency, and scan rate of each of the relaxation times in order to determine the most optimal variable for use as a detection mechanism. The final component of Test One was to calculate the contour of the curve for each relaxation variable versus particle concentration. The 38 samples were held in a test tube rack and were tested simultaneously under 3.0 Tesla MRI for their T1, T2, and T2* values. The test tube rack and samples were initially submersed in a water bath to create a more consistent magnetic medium, potentially increasing the accuracy of the results due to improved signal homogeneity. The tests were repeated without the water bath for comparison of the results and to determine the potential impact of the medium. The background image of the water bath was also calculated for reference to the sample data for all of the proceeding tests. 0 0.0002 0.0004 0.0006 0.0008 0.001 0.0012 0.0014 0.0016 0.0018 1 3 5 7 9 11 13 15 17 19 SPI O M olar Q uan tity (n M) Sample Number 30nm MNP Molar Quantity (nM) 0 0.0002 0.0004 0.0006 0.0008 0.001 0.0012 0.0014 0.0016 21 23 25 27 29 31 33 35 37 Sample Number 10nm MNP Molar Quantity (nM) 28 To prove the predicted SPIO nanoparticle concentration and particle size, 4 of the MNP samples from Test One were imaged under a Scanning Electron Microscope. The images were magnified to 70,000X. A slight measurable offset for the expected SPIO nanoparticle concentrations of Test One samples had occurred due to the imprecision of the original pipette used. The SPIO nanoparticle concentration was calculated from the SEM images and the concentrations were re-fitted according to their expected relaxation values. The re-fitted concentrations were used for later comparisons to the T2 results found in Test Three. Figure 4.3 SEM images of the 30 nm SPIO nanoparticles used in Test One. 4.2 Test Two: T2* Minimum SPIO Nanoparticle Concentration Detection Test Two included a total of 30 samples. Of these, only two were pure water samples. Based upon information gathered from Test One, it was decided that further testing would only include the 30nm inorganic core size SPIO nanoparticles. Also, T2* was predicted to be the most applicable parameter from Test One and was therefore chosen as the only relaxation variable used for this test. A gradient echo procedure was used to measure T2* in Test Two. Test Two focused on determining the curve contour for a lower concentration range of SPIO nanoparticles for T2*. To obtain the lower concentration range, samples 17 and 18 from Test One were diluted to form the 28 new samples used in Test Two (Figure 4.4). Of the 28 samples containing SPIO nanoparticles, 2 of them contained the same SPIO nanoparticle concentration as samples 1 and 2 from Test One. This was to assess the reliability and validity of the results between Test One and Two. The purpose of Test Two was to find the optimal concentration values to be used as a detection mechanism and to predict the expected T2* values over a wide range of SPIO nanoparticle concentrations. 29 Figure 4.4 Test Two sample?s magnetic nanoparticles molar quantities Figure 4.5 MR image of the regions targeted for each sample. The numbers listed are the respective samples numbers from Test Two. The arrow indicates that the concentration of SPIO nanoparticles decreases down each row. Areas circled outside of the test tubes (green and yellow) were measured for their relaxation values in order to be used as a referecne. 0 0.00002 0.00004 0.00006 0.00008 0.0001 0.00012 0.00014 0.00016 0.00018 0.0002 39 41 43 45 47 49 51 53 55 57 59 61 63 65 67 SP IO M ola r Qu ant ity (n M) Sample Number 30nm MNP Molar Quantity (nM) 30 4.3 Test Three: Validity and Reliability of T2 and T2* Only the 30nm inorganic particle size was selected for use in all proceeding tests due to the information collected from Test One and Two. Based on results from Test Two, T2* was not considered a reliable predictor of concentration values due to lack of consistent findings between test repetitions. As a result, both T2 and T2* were calculated in Test Three. A total of 29 samples were used in Test Three, one of which was a pure water sample. Of the 29, only 24 were newly created samples (Figure 4.6). The remaining 4 were existing samples from Test One, selected to prove if there were any loss in magnetization of the SPIO nanoparticles over a long period of time. Test tube samples were randomly distributed to measure potential effects of sample interaction and their relaxation time changes based on their test tube rack position. Test Three SPIO nanoparticle concentration range spanned over the entire concentration range utilized in both Test One and Test Two. Each of the samples prepared for Test Three were measured to contain similar SPIO nanoparticle quantities as samples used in the previous two tests. Figure 4.6 Test Three sample?s SPIO nanoparticles molar quantities Using the SPIO nanoparticle concentration versus T2 data, the Test One SPIO nanoparticle concentrations were re-fitted to their expected values with respect to their individual T2 times. A trend line was fitted to the Test Three T2 versus SPIO nanoparticle particle concentration data. The T2 values of each sample for the Test One data 0 0.0002 0.0004 0.0006 0.0008 0.001 0.0012 0.0014 0.0016 0.0018 69 71 73 75 77 79 81 83 85 87 89 91 93 95 SPI O Mo lar Q uan tity (n M) Sample Number Test Three Sample Concentrations 31 were plotted along this curve to determine their predicted SPIO nanoparticle concentrations. The newly calculated SPIO nanoparticle concentrations for Test One more accurately fit than what was previously measured. 4.4 Test Four: T2 as a Detection Mechanism The primary purpose of Test Four was to prove that T2 could be used to detect the presence of biotinylated polystyrene (BPM) particles. The biotin coated polystyrene (BPM) particles were ordered from Spherotech. These particles were 3-3.4?m in size with 4.96*105 Streptavidin-FITC binding sites available per polystyrene particle. There were 42 samples total in Test Four, with one being a pure water sample. From the T2 versus SPIO nanoparticle concentration found in Figure 4.7, two sample concentrations were selected to be used for the detection mechanism. These two sample concentrations were selected based on their position on the T2 versus concentration curve where slight changes in magnetic environment could greatly alter the T2 values. It was predicted that these SPIO nanoparticle concentrations would show the most noticeable change in their T2 values after the addition of the BPM particles. 32 Figure 4.7 Diagram indicating the predicted changes in T2 values after the addition of the target molecules. The resultant change in T2 was predicted to show a larger deviation for IO concentrations along the upward portion of the curve shown above. The sample SPIO nanoparticle concentrations chosen were from sample numbers 79 and 80. Of the 41 SPIO nanoparticle samples, 2 contained only SPIO nanoparticle concentrations samples number 79 and 80, respectively. The remaining 39 samples were divided into 20 containing only sample number 79?s concentration and 19 with only sample number 80?s concentration. Later, the biotinylated polystyrene particles were added in specific concentrations to these 39 samples. After the addition of the biotinylated polystyrene (BPM) particles, the samples were sonicated to support biotin-streptavidin interaction. Table 4.2 Sample 79 and 80 SPIO nanoparticle molar concentration Sample SPIO Molar concentration 79 .0000918 nM 80 .0001836 nM 33 Figure 4.8 Diagram representing the expected binding between the SPIO nanoparticles coated with streptavidin and the biotinylated polystyrene microspheres. (1) The samples are originally formed with only dionized water and the desired SPIO nanoparticle concentration. (2) The samples are later combined with a specific concentration of the biotinylated polystyrene microsphere (BPM). The resultant change in T2 will be calculated in order to provide proof of the method as a detection mechanism. (3) Non-specific binding of the SPIO nanoparticles could potentially influence the accuracy of the relaxation time measurements. The addition of the biotinylated polystyrene microspheres to the samples containing the SPIO nanoparticles was expected to bind together forming SPIO-BPM clusters. The formation of these clusters was predicted to induce larger local magnetic field inhomogeneities, thereby shortening the spin-spin relaxation rate (smaller T2) of surrounding water proton spins. The resultant change in spin-spin relaxation (T2) due to the addition of these microspheres would verify this method as a detection mechanism (Figure 4.8). The biotinylated polystyrene particles added to the 39 samples were over a range of 1?l to 100?l (253000 microspheres/ ?l) to determine the minimum concentration for detection. Each BPM concentration was repeated twice to assure its test reliability. The testing was divided into two parts (Figure 4.9), samples containing only SPIO nanoparticle concentrations used in sample number 79 and those with only concentrations from sample number 80. One sample from each test part contained pure water and the specific SPIO nanoparticle concentration of sample numbers 79 and 80. The remaining samples for both Test Four parts were combined with increasing amounts of BPM particles listed in Table 4.5. The samples were all sonicated for 20 minutes and stored overnight. Finally, after 24 hours, the Test Four samples were tested for their T2 values. 34 Figure 4.9 Displays the polystyrene microsphere (BPM) count within each sample number with its respective SPIO nanoparticle optimal concentration After the MRI scan and data reduction, the T2 values for samples containing biotinylated polystyrene particles were compared to the T2 value of their respective SPIO nanoparticle concentration without the polystyrene additions. This value would provide the ?T2 value for each of the biotinylated polystyrene samples. The ?T2 value was predicted to show proof of the existing BPM particle concentrations in each of the samples [4, 5, 6]. ?T2 was calculated as follows: ?T2 = T2(original) ? T2(detection) T2(original) = MNP concentration without polystyrene bead T2(detection) = MNP concentration with polystyrene bead 4.5 Test Five: Detection at a Higher SPIO Nanoparticle Concentration After discussing the results of the previous test it was concluded that the detection needed to perform at a higher SPIO nanoparticle concentration to assure the unexpected T2 values from Test Four were due to limited SPIO-BPM interaction. There were 12 new samples created for Test Five (Figure 4.10). Of these 12 new samples, 1 was a pure water sample and 2 were samples containing only the SPIO nanoparticle concentration selected. The SPIO nanoparticle concentration selected for Test Five was of the same molar concentration (0.0002754nM) as 0 5000000 10000000 15000000 20000000 25000000 30000000 97 99 101 103 105 107 109 111 113 115 117 Mi cro sp he re Qu an tity (B PM ) Sample Number Optimal Concentration (.0000918nM) 0 5000000 10000000 15000000 20000000 25000000 30000000 118 120 122 124 126 128 130 132 134 136 Mi cro sp he re Qu an tity (B PM ) Sample Number Optimal Concentration (.0001836nM) 35 sample number 81 from Test Three. This was to compare the consistency the T2 values between Tests Three and Five. The molar concentration was also chosen to be of a higher value than that use in Test Four. All but the pure water sample, were filled with the same SPIO nanoparticle molar concentration of 0.0002754nM. After the addition of the SPIO nanoparticles, increasing concentrations of the polystyrene microspheres (BPM) were added to 9 of the 12 new samples. The concentration of the BPM added were of the same values as what was added to each test tube in Test Four for data comparison. Unlike Test Four, the samples from Test Five were sonicated 30 minutes prior MR imaging. Figure 4.10 Polystyrene microsphere (BPM) count within each sample number with its respective SPIO nanoparticle optimal concentration After the data reduction of the Test Five data, the ?T2 of the new samples were calculated. The new ?T2 data was then compared to the ?T2 values from the previous test to determine if there was any trend with the T2 values and the increasing additions of polystyrene microspheres (BPM). 4.6 Test Six: Confirmation of the T2 Values with Increasing BPM Concentration To verify the T2 values measured from Test Five, an additional test was performed. Test Six included 30 new samples and the 12 samples from Test Five (Figure 4.11). The 30 new samples each contained the same SPIO nanoparticle molar concentration that was used from Test Five (0.0002754nM). After the addition of the SPIO nanoparticle to each sample, the polystyrene microspheres (BPM) were then combined with the new samples in 0 5000000 10000000 15000000 20000000 25000000 30000000 139 140 141 142 143 144 145 146 147 148 149 Mi cro sp he re Qu an tity (B PM ) Sample Number Optimal Concentration (.0002754nM) 36 the same increasing concentration as what was used in Test Five. Each BPM addition was repeated 3 times in order to insure accurate data. Figure 4.11 Polystyrene microsphere (BPM) count within each sample number with its respective SPIO nanoparticle optimal concentration. The samples used in Test Five were retested to verify the repeatability of their T2 values. All samples were sonicated 24 hours prior to MR imaging and were randomly distributed along the test tube rack. Finally, after data reduction, the ?T2 was calculated for each sample. 4.7 Test Seven: Determining Particle Settling Using Sagittal Imaging In the preceding test it was verified that a visible precipitate formed over the course of the MRI scanning sequence. Due to the larger micron size of the polystyrene spheres, it was believed that during the formation of the SPIO-BPM clusters, the aggregates would become too dense and settle to the bottom of samples. To confirm whether this phenomenon was occurring during the scanning procedure, two sagittal slices were taken of each of the test tubes for Test Seven. 0 5000000 10000000 15000000 20000000 25000000 30000000 139 141 143 145 147 152 154 156 158 160 162 164 166 168 170 172 174 176 178 180 182 M icr osp he re Qu an tity (B PM ) Sample Number Optimal Concentration (.0002754nM) 37 Figure 4.12 MRI of the sagittal slices taken of the first column of test tubes within the rack prior to the scanning sequence. In order to compare the consistency of the T2 values for Test Six and Seven, the samples from Test Six (Figure 4.11) were reused. All samples were sonicated 5 minutes prior to the start of the MRI sequence to ensure as little settling as possible. As shown in Figure 4.12, a sagittal slice of the top and bottom of the samples were taken throughout the imaging over a period of 48 hours. The samples remained within the MRI?s 3.0 Tesla field during the 48 hour imaging procedure. Each of the six columns shown below in Figure 4.13 was tested over a total acquisition time of 8 hours. Since it was unknown when the particle settling would occur, each of the columns from Figure 4.13 were tested individually starting from the far left column then proceeding with next adjacent column to the right. The far left column was tested during the start of the scanning sequence until 8 hours later after its acquisition time was completed while the far right column was tested after 40 hours under the magnetic field until 8 hours later its own acquisition time was completed. Backgrou nd 1 2 4 6 8 10 12 14 3 5 7 9 11 13 15 38 Figure 4.13 Locations of each of the samples in the test rack from an aerial view. The colored circled regions are the areas targeted for their T2 signal during their acquisition time. Numbers 1-10 from the image above correspond to the sample numbers 139-148 while numbers 11- 14 are samples 151-154. Finally the above numbers 15-42 parallel the samples 162-182. B a c k g r o u n d 1 2 3 4 5 6 7 B l a c k D o t e n d 8 9 10 11 12 13 14 15 16 17 18 1 9 43 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 39 CHAPTER 5 RESULTS and DISCUSSION 5.1 Relaxation Time and Particle Size Comparison It has been reported previously that by differing the core size of the SPIO nanoparticles, noticeable changes in the relaxation rates of the surrounding water molecules would occur [4, 5, 6, 17]. In the first test, two different IO core sizes were used to monitor their respective influences on the T1, T2, and T2* of the water molecules. The findings of Test One showed that the 30 nm SPIO nanoparticle tested with a greater effect on the changes of T1, T2, and T2* than the smaller 10 nm SPIO nanoparticle. It was theorized that although the reduction of the core would decrease the magnetic domain size of the particle, the magnetic field surrounding the particle would also be condensed to a point where the influence would not be great enough until higher concentrations of the particle were used. In Figure 5.1, a comparison of the different SPIO particle core sizes and their impact on T1, T2, and T2* is shown below. 40 0 500 1000 1500 2000 2500 0 0.0002 0.0004 0.0006 0.0008 0.001 0.0012 0.0014 0.0016 0.0018 T1 (m s) SPIO Nanoparticle Concentration (nM) Particle size vs. T1 10nm (T1) 30nm (T1) 0 50 100 150 200 250 300 350 0 0.0002 0.0004 0.0006 0.0008 0.001 0.0012 0.0014 0.0016 0.0018 T2 (m s) SPIO Nanoparticle Concentration (nM) Particle size vs. T2 10nm (T2) 30nm (T2) 0 5 10 15 20 25 30 35 40 45 0 0.0002 0.0004 0.0006 0.0008 0.001 0.0012 0.0014 0.0016 0.0018 T2* (m s) SPIO Nanoparticle Concentration (nM) Particle Size vs. T2* 10nm (T2*) 30nm (T2*) 41 Figure 5.1 Test One: top (a), middle (b), and bottom (c) graphs display the varying changes in T1, T2, T2* associated with particle size and concentration It was also verified from Test One that the presence of the SPIO nanoparticles in the solution had little to no effect on the values for T1. Any of the particle concentration dependencies of the longitudinal magnetization fluctuations seemed to follow no trend and were of similar values as the pure water samples. In Figure 5.2 it can be seen that the values for T2 and T2* were of a much smaller magnitude than that of T1 for all sample concentrations and particle sizes. This finding followed other research previously mentioned and was decided that T1 would not be the essential variable for further development of the detection mechanism. Figure 5.2 Test One: comparison of the 30nm SPIO nanoparticles proton relaxation data Following this judgment, further testing would include an expanded range of the SPIO nanoparticle concentrations to determine if any related T2 and T2* changes would follow a specific trend. This trend would potentially allow for discovery of the optimal SPIO nanoparticle concentration for target detection. 5.2 Reliability of T2 and T2* As can be seen from Figure 5.4 the Test Two measured values for T2* of the 30nm SPIO particle concentrations did not follow the expected trend theorized for the lower concentrations of the SPIO nanoparticles. The values were found to be highly disordered and were greatly influenced by any external interactions or particle aggregation. T2* was severely affected by any disturbances in the magnetic field. Any inhomogeneities in the 1 10 100 1000 0 0.0002 0.0004 0.0006 0.0008 0.001 0.0012 0.0014 0.0016 0.0018 Tim e (m s) SPIO Nanoparticle Concentration (nM) 30nm (T1) 30nm (T2) 30nm (T2*) LOG 42 magnetic field would alter the expected trend in the T2* vs. concentration. As shown in Figure 5.3 (a), the equilibrium position of the transverse directional component is aligned in the direction of the spectrometers static magnetic field (B0). Once the RF excitation pulse is applied, the transverse component will transition to the direction of the applied pulse (B1). The relaxation period (T2) for the decay of this directional component from the RF excitation pulse induced excited state to the equilibrium position, Figure 5.3 (b) and (c), is dependent on the strength of the magnetic field (B0) [8]. Unlike T2, the local magnetic environment (T?) greatly manipulates T2*. This can be seen from Eq.4 where the variable T? is strongly dependent on any surrounding magnetic disturbance. The inhomogeneity factor (T?) considers the influence of the magnetic moments of surrounding molecules on the water hydrogen proton?s transverse magnetization directional component (T2) [8, 9]. The local magnetic dipoles enhance the decay of the hydrogen proton?s transverse magnetization along with the external field thereby shortening the relaxation period [8]. Figure 5.3 Left (a), water hydrogen proton?s dipole moment is aligned with spectrometers static magnetic field (B0). The directional component is changed to B1 after RF excitation pulse. Middle (b), decay of water hydrogen protons transverse magnetization directional component. Right (c), NMR signal acquired by spectrometer over transverse relaxation period [25]. The observed frequency collected is highly dependent on the variation of the magnetic field as it transitions through the NMR tube. Any fluctuations in the magnetic field would effectively alter this frequency thereby broadening the NMR line (Figure 5.3 c). As demonstrated in Eq.4 this broadening effect is directly proportional to the frequency and the inhomogeneity of the magnetic field. Considering that the SPIO particle dispersion in the targeted area for imaging would be of differing concentrations for each sample, it was predicted that T2* could not be used as a detection mechanism. Although the data suggests that at higher concentrations of 43 SPIO nanoparticles, the average values of T2* appear to remain consistent, it may still not remain dependable for larger volume samples. Figure 5.4 Test Two: calculated T2* values for the 30 nm SPIO nanoparticles. Due to the inconsistencies found from the T2* data collected from Test One and Two, it was considered necessary to determine if any exterior effects were disturbing the acquired NMR signal. Multiple modifications to the setup were made for Test Three in order to demonstrate T2*?s variations. First, it was believed that the magnetic fields within each sample could feasibly alter the received signal from any adjacent sample. This was tested by randomizing the particular order of the samples in the rack and examining any changes in relaxation rates from samples that were in close proximity of higher SPIO nanoparticle concentrations. The theory was found to be disproven due to the lack of any particular patterns that would suggest these influences. Second, the experimental setup was examined for any possible contaminations that would inhibit the NMR signal. The test tubes, pipettes, and test tube rack were all inspected prior to Test Three and were all confirmed to be clean. The T2* data for the lower concentrations in Test Three, shown in Figure 5.5, was also found to be disordered. 0 20 40 60 80 100 120 0 0.00005 0.0001 0.00015 0.0002 T2* (m s) SPIO Nanoparticle Concentration (nM) T2* (pure water samples) T2* (30nm MNP) 44 Figure 5.5 Test Three: Lower concentrations of the SPIO nanoparticles were retested for their T2* values in Test Three to determine the consistency between Test Two and Three values. Third, the effect of the NMR signal transitioning between the water bath, test tube, and sample solution was reviewed. It was believed that the route taken by the signal through alternating mediums could potentially hinder its reception and reduce the signal-to-noise ratio. The samples from Test One were retested for their T1, T2, and T2* values when imaged without the water bath during the imaging sequences. A variance in T2* between the tests with and without the water bath can be seen in Table 5.1. However, this variance could not be verified to be influenced by the water bath since there were many other factors that could cause the indicated discrepancies. More importantly the water bath data shows a more consistent result that match predictions while the without water bath has more outliers. Table 5.1 Wet-to-Dry Percentage Change in T2* Sample Number 30nm SPIO Molar Quantity (nM) Wet (With WB) Dry (Without WB) % Change T2* 1 0.0000918 14.16 15.73 11.09 2 0.0001836 12.82 17.75 38.48 3 0.0002754 9.63 12.14 26.08 4 0.0003672 8.02 9.22 14.87 5 0.000459 6.16 7.29 18.28 6 0.0005508 7.91 8.48 7.21 7 0.0006426 6.27 7.64 21.85 8 0.0007344 5.95 6.92 16.32 9 0.0008262 5.60 6.92 23.40 0 10 20 30 40 50 60 70 80 90 100 0 0.00002 0.00004 0.00006 0.00008 0.0001 0.00012 0.00014 0.00016 0.00018 0.0002 T2* (m s) SPIO Nanoparticle Concentration (nM) Test Three (T2*) 45 10 0.000918 5.27 5.91 12.16 11 0.0010098 4.73 5.06 6.84 12 0.0011016 4.46 4.89 9.76 13 0.0011934 3.89 4.02 3.32 14 0.0012852 3.97 4.32 8.89 15 0.001377 4.06 4.62 13.76 16 0.0014688 4.08 4.32 5.83 17 0.0015606 3.90 4.07 4.36 18 0.0016524 3.83 3.89 1.51 19 0 38.95 27.83 -28.55 20 0 20.57 40.07 94.84 Sample Number 10nm MNP Molar Quantity (nM) Wet (With WB) Dry (Without WB) % Change T2* 21 0.000086 17.97 34.69 92.99 22 0.000172 28.08 28.87 2.82 23 0.000258 21.32 29.60 38.80 24 0.000344 20.35 29.18 43.36 25 0.00043 22.49 25.00 11.15 26 0.000516 20.70 23.29 12.51 27 0.000602 20.23 22.42 10.84 28 0.000688 22.76 33.42 46.86 29 0.000774 23.44 34.98 49.24 30 0.00086 39.56 35.27 -10.83 31 0.000946 26.71 18.84 -29.48 32 0.001032 28.65 20.90 -27.05 33 0.001118 21.34 21.32 -0.10 34 0.001204 40.15 17.97 -55.23 35 0.00129 22.07 26.62 20.65 36 0.001376 19.61 39.74 102.58 Fourth, it can be seen in Figure 5.6 how any changes in the methodology for data reduction can greatly alter the calculated relaxation rates. It was also proven that the relaxation values were more susceptible to fluctuations when at lower SPIO nanoparticle molar concentrations. This led to the belief that for any acceptable experimentation a standard must be maintain for all data reduction calculations. Sampled data was restricted to uphold a minimum R2 curve fitting value of .98 and would include all relevant data points until the curve 46 plateaued. Otherwise, the sample data series would be considered to have too small of a signal-to-noise ratio and be rejected as a suitable calculation. Figure 5.6 Displays the impact of the data reductions methodology on T2 results. Both images were created for the calculation of sample 149?s T2 value. The top image contains fewer echo times in comparison to the bottom resulting in a more accurately fit curve and shorter T2 value. Lastly, in Test Three there were 4 reused samples that were imaged previously during Test One. Both the T2 and T2* values of these 4 Test One samples shown in Table 5.2 remained consistent between Tests One and T2 (sec)0.0349 y = 0.2675e-28.67x R? = 0.9979 0 0.05 0.1 0.15 0.2 0.25 0.0000 0.0200 0.0400 0.0600 0.0800 0.1000 0.1200 0.1400 Mag (AU ) TE (sec) ROI 11 Pix Mean ROI 11 Pix Mean Expon. (ROI 11 Pix Mean) T2 (sec)0.0399 y = 0.2423e-25.06x R? = 0.9768 0 0.05 0.1 0.15 0.2 0.25 0.0000 0.0500 0.1000 0.1500 0.2000 Mag (AU ) TE (sec) ROI 11 Pix Mean ROI 11 Pix Mean Expon. (ROI 11 Pix Mean) 47 Three. Another 12 samples were also imaged in Test Three that contained equal SPIO nanoparticle molar concentrations to that of the previous samples used in Test Two. The T2* results from both of these tests (Figure 5.4 and 5.5) were found to be comparatively erratic. Table 5.2 Samples from the previous Test One were repeated in Test Three to confirm the reliability of each test to repeat similar results. Sample Number Test One T2 (ms) Test Three: T2 (ms) Test One: T2* (ms) Test Three: T2* (ms) 4 10.75 11.5 8.77 6 5 8.7 10.2 6.29 7 6 11.11 15.6 8.2 8.2 14 5.08 6.01 4.26 4.8 It was concluded that the reliability of T2* would only remain dependable when at higher concentrations of the SPIO nanoparticles. After comparing the T2 data from Tests One and Three, it was hypothesized that only T2 would act as a sufficient variable for target detection. 5.3 Optimal SPIO Nanoparticle Concentration for T2 Target Detection The lower concentrations of the SPIO nanoparticles were selected for use in both Test Two and Three in order to determine an ?optimal? concentration to be later used for detection of target molecules. The optimal concentration would be the SPIO nanoparticle molar amount at which any modification to the particles dispersion would cause the greatest fluctuation to the resultant change in proton relaxation or ?T2. This theory was discussed previously in Chapter 4.4 which contained an image (Figure 4.4) demonstrating the ideal behavior of the transverse relaxation at this molar quantity. Figure 5.7 displays the nanoparticle concentration range that was predicted to be most sensitive to changes in ?T2 due to particle dispersion. The suggested molar range follows along the T2 curve between the SPIO nanoparticle molar concentrations of .0001-.0005nM. 48 Figure 5.7 Test Three: illustrates the predicted MNP concentration range for detection of the target molecules with greatest sensitivity. It should be noted that the measured concentrations of the SPIO nanoparticles for Test One were off by a small percentage than from what was originally expected. This was due to a slight malfunction from the pipette that was previously used. All tests following the incident were measured with a properly functioning pipette. The original concentrations for Test One were recalculated based on the trend discovered during Test Three. As shown in Figure 5.7 above, the changes in T2 follows a tendency that varies directly with the modifications in SPIO nanoparticle concentrations between each solution. Based on this discovery, the previous Test One samples were replaced with newer SPIO nanoparticle concentration values that followed this curve (Figure 5.8). The newly assigned concentrations for Test One samples can be found in Table 7.8 in the Appendix below. 0 20 40 60 80 100 120 140 0 0.0005 0.001 0.0015 T2 (m s) SPIO Nanoparticle Concentration Test Three: T2 Detection Range 49 Figure 5.8 Shows the expected values for the Test One sample concentrations based on their recorded T2 values and the tendency for each SPIO particle concentration to follow the above trend. 5.4 Detection of Polystyrene Microspheres (BPM) Three molar concentrations within the ideal SPIO nanoparticle molar concentration range were used for SPIO nanoparticle target detection of the polystyrene microspheres (BPM) in Tests Four, Five, Six, and Seven. Originally it was predicted that the most favorable SPIO molar concentrations would be for the lowest quantities in the optimal range. The results obtained from Test Four conveyed a very interesting phenomenon that has not yet been explained in current literature. Unlike in previous reporting, the consequential decrease in transverse relaxation did not occur for all samples upon the addition of the target binding molecule (BPM) with the SPIO nanoparticle. As shown in Figure 5.9a, the data collected from Test Four did not provide the desired indication of T2 reducing due to particle aggregation for all test samples. The same incident also occurred in Tests Five and Six (Figure 5.9b and 5.9c) when it was decided to use a greater SPIO nanoparticle molar amount. After calculating the ?T2 value for each target BPM concentration, the only noticeable pattern was the increase in transverse relaxation with increasing additions of the polystyrene microspheres (BPM). 0 10 20 30 40 50 60 70 80 90 100 0 0.0005 0.001 0.0015 0.002 0.0025 T2 (m s) Particle Concentration (nmole/ml) T2 (test 3) Fitted T2 (test 1) Power (T2 (test 3)) 50 Figure 5.9 The top (a), middle (b), and bottom (c) graphs display the resulting changes in transverse relaxation based on the optimal concentration of the SPIO nanoparticle chosen and the quantity of polystyrene microspheres (BPM) present within the solution. Although the influence of the BPM augmentation did increase the transverse relaxation time, the T2 values were not of similar magnitude of that of the pure water samples suggesting that there were still water molecules being impacted by the local magnetic environment. A couple of theories were formulated in order to 20 70 120 170 220 270 0 5000000 10000000 15000000 20000000 25000000 30000000 T2 (m s) Polystyrene Microspheres (BPM) Optimal Concentration (.0000918 nM) 20 30 40 50 60 70 80 90 100 0 5000000 10000000 15000000 20000000 25000000 30000000 T2 (m s) Polystyrene Microspheres (BPM) Optimal Concentration (.0001836 nM) 20 30 40 50 60 70 80 0 5000000 10000000 15000000 20000000 25000000 30000000 T2 (m s) Polystyrene Microspheres (BPM) Optimal Concentration (.0002754 nM) 51 explain this phenomenon. First, it was believed that the increase in T2 was caused by the diffusion of the SPIO nanoparticles to areas of high concentration along the test tube wall when under the applied magnetic field. This would reduce total number of water molecules being affected by the magnetic moments of the SPIO nanoparticles. Although the occurrence would have been encountered in previous testing, it would not have been noticeable until the addition of the polystyrene microspheres (BPM) that would inhibit this diffusional process. Second, there were visible precipitates found for samples containing higher concentrations of the target BPMs. Upon interaction with the SPIO nanoparticles, the target BPMs would form large clusters that would become too dense to remain suspended in solution. The area where the precipitates would form would not be in the space where the coronal slice was imaged along the test tubes. This would also increase T2 values due to the lessening of water molecules being affected by SPIO nanoparticles. In response to the first theory, it was demonstrated from Tests One through Six that the samples containing only the SPIO nanoparticles would measure for similar T2 values between subsequent tests when exposed to an external magnetic field for short periods of time. Alternatively in Test Seven, samples that contained equal SPIO nanoparticle molar concentrations and were not mixed with the polystyrene microspheres (BPM) revealed differing T2 values when under the MRI?s 3.0 Tesla field for prolonged periods of time. Samples 139 and 182 were imaged in Test Seven and contained identical SPIO nanoparticle concentrations without the addition of the target molecules. In Test Seven, the test tubes were placed within the MRI?s magnetic field for continuous imaging of 48 hours. Sample 139 was imaged at the beginning of the procedure until the acquisition time completed on the 8th hour while Sample 182 was imaged at the start of the 40th hour until the final acquisition time ended at 48 hours from the start of the test. As shown in Table 5.3, the T2 value for the Sample 182 was significantly increased after the 40-48 hour period unlike Sample 139. With the data shown Table 5.3 it was reasonable to assume that a portion of the T2 hindrance discovered in Test Seven was due to the diffusion of the unbound SPIO nanoparticles to highly concentrated areas thus reducing the quantity of water molecules affected by their magnetic dipoles. Table 5.3 A comparison between the transverse relaxation (T2) times of two samples containing equal molar quantities of the SPIO nanoparticles. Each sample was imaged at different periods of time while under a continuously applied external magnetic field. 52 Sample number Top of test tube T2 (ms) Bottom of test Tube T2 (ms) Hours under magnetic field 139 33.8 32.2 0-8hrs 182 67.6 87.3 40-48hrs To verify the formation of the SPIO-BPM precipitates during MR imaging, a sagittal slice of each test tube was taken in Test Seven in order to image both the top and bottom of the test tubes simultaneously. The procedure was conducted to validate any discrepancy between each images value for T2. The results of Test Seven indicated that the majority of the samples experienced a lengthening of T2 at the bottom of samples as a consequence of increasing target concentrations. This effect was most present for samples holding the highest quantities of the BPMs and occurred for each of the columns signal acquisition time periods over the total 48 hours with the exception of the initial 0-8 hour scan. It should also be noted that the T2 versus BPM concentration dependencies did still follow the previous pattern of increasing with additional concentrations. Figures 5.10 and 5.11 below display the divergences between the T2 values of top and bottom of the test tubes in Test Seven. Figure 5.10 Test Seven: displays changes in sample transverse relaxation with differing target BPM concentrations. The top and bottom of each test sample was imaged to determine any variations in T2 due to particle settling and/or aggregation. 0 50 100 150 200 250 300 350 400 0 5000000 10000000 15000000 20000000 25000000 30000000 T2 (m s) Polystyrene Microspheres (BPM) Top of test tube (T2) Bottom of test tube (T2) 53 Figure 5.11 Bar graph of Test Seven data for easier visualization of T2 results. The greater degree of growth in T2 for the bottom images than the top images provided an interesting discovery. It was originally predicted that the formation of the SPIO-BPM clusters would consequentially reduce the transverse relaxation (T2) but was later believed to increase after reviewing the T2 results from Tests Four, Five, and Six. During these previous tests, the imaging was centralized over the test tube samples thus neglecting the resultant dispersion of the SPIO nanoparticles after mixing. The greater extension of the transverse relaxation for the bottom rather than the top of the test tubes potentially suggests that the target BPM particles size was too large for the predicted T2 reduction effects to occur. As the accumulation of the SPIO-BPM clusters progresses, the density of the clusters becomes too great to remain suspended within the solution. The transference of the clusters to the bottom of the test tube is initiated thus altering the even dispersion of the molecules. This development would transpire more rapidly for larger micron sized target molecules rather than smaller nanoparticles. The coupling of the SPIO nanoparticle magnetic moments due to SPIO-BPM clustering has been shown in Test Seven to have been inhibited rather than enhanced. The probability of there being multiple SPIO nanoparticles conjugated close enough to each other on the surface of a target molecule (BPM) for the coupling of their magnetic moments to occur, decreases with increasing target molecule size. In previous research that experienced the decrease in transverse relaxation (T2) due to SPIO-Target clustering typically only targeted nano- sized molecules for detection. This research proposes that there is a maximum target molecule size for NMR 0 50 100 150 200 250 300 350 400 T2 (m s) Polystyrene Microspheres (BPM) Top of test tube (T2) Bottom of test tube (T2) 54 detection when expecting a reduction of T2 upon the arrangement of the SPIO-BPM clusters. For larger micron sized target particles, Test Seven has shown an alternative method for pathogen detection. 55 CHAPTER 6 CONCLUSION This research has demonstrated a potential detection technique that uses superparamagnetic iron oxide (SPIO) nanoparticles for the capture and recognition of target molecules present in biological media. The SPIO nanoparticle core size was shown to have comparatively differing effects when influencing proton relaxation. A broad concentration range of the SPIO nanoparticles was utilized to test their impact on the three relaxation variables T1, T2, and T2* of the surrounding water molecules. The longitudinal relaxation was found to not be affected by the local SPIO nanoparticles while T2 and T2* were shown to be greatly influenced. T2* was later proven to be unreliable between each subsequent test and was therefore determined to not be applied as the primary variable for the detection mechanism. An optimal SPIO nanoparticle concentration range for the determination of the target pathogen (BPM) in solution was developed. Unlike what was previously predicted, the interaction of the SPIO nanoparticles with the target BPM did not drastically reduce the transverse relaxation (T2) times but instead increased upon formation of the SPIO-BPM clusters. Repeated tests confirmed varying increases in T2 along both ends of the BPM concentration range with a noticeable tendency to increase more substantially for the highest BPM quantities. A visible precipitate (SPIO-BPM clusters) was discovered for all samples containing the larger concentrations of the target BPM molecules. Based on this sighting, the previous BPM concentrations were continuously imaged over a 48 hour period at the top and bottom of the test tubes to verify any particle settling during the scanning procedure. The results indicated that the majority of the samples experienced a lengthening of T2 at the bottom of samples as a consequence of increasing target BPM concentrations. Upon further inspection it was shown that all unbound SPIO nanoparticles could potentially slowly diffuse together forming highly concentrated areas and the SPIO-BPM clusters could eventually grow too large forcing them to settle to the 56 bottom of the test tube. Both occurrences would forcibly reduce the average quantity of water hydrogen protons affected by the SPIO nanoparticles magnetic moments thus increasing the transverse relaxation (T2) time of the solution. Finally it was concluded that for NMR detection of larger micron-sized target molecules in solution, a rise in transverse relaxation was to be expected. 57 Appendix Table 7.1 List of Samples used for Test One with their respective SPIO nanoparticle core size and molar concentration Sample Number SPIO Nanoparticle Core Size 30nm SPIO Molar Quantity (nM) 10nm SPIO Molar Quantity (nM) 1 30nm 0.0000918 0 2 30nm 0.0001836 0 3 30nm 0.0002754 0 4 30nm 0.0003672 0 5 30nm 0.000459 0 6 30nm 0.0005508 0 7 30nm 0.0006426 0 8 30nm 0.0007344 0 9 30nm 0.0008262 0 10 30nm 0.000918 0 11 30nm 0.0010098 0 12 30nm 0.0011016 0 13 30nm 0.0011934 0 14 30nm 0.0012852 0 15 30nm 0.001377 0 16 30nm 0.0014688 0 17 30nm 0.0015606 0 18 30nm 0.0016524 0 19 Water 0 0 20 Water 0 0 21 10nm 0 0.000086 22 10nm 0 0.000172 23 10nm 0 0.000258 24 10nm 0 0.000344 25 10nm 0 0.00043 26 10nm 0 0.000516 27 10nm 0 0.000602 28 10nm 0 0.000688 29 10nm 0 0.000774 30 10nm 0 0.00086 31 10nm 0 0.000946 32 10nm 0 0.001032 33 10nm 0 0.001118 34 10nm 0 0.001204 35 10nm 0 0.00129 36 10nm 0 0.001376 37 Water 0 0 38 Water 0 0 58 Table 7.2 Relaxation rates of each sample in Test One Sample Number SPIO core size T1(ms) T2(ms) T2*(ms) 1 30nm 1954 45.45 15.63 2 30nm 1822 25.64 12.5 3 30nm 1598 14.49 10.64 4 30nm 1452 10.75 8.77 5 30nm 1340 8.7 6.29 6 30nm 1488 11.11 8.2 7 30nm 1371 9.43 7.09 8 30nm 1320 8.55 6.21 9 30nm 1250 7.94 5.62 10 30nm 1147 7.3 5.43 11 30nm 1072 6.23 4.9 12 30nm 1042 5.52 4.72 13 30nm 951 5.52 4.22 14 30nm 985 5.08 4.26 15 30nm 1028 5.99 4.37 16 30nm 940 4.9 4.15 17 30nm 902 5.35 4.12 18 30nm 877 5.92 4.12 19 Water 2105 1250 41.67 20 Water 2102 1428.57 19.61 21 10nm 1898 333.33 20 22 10nm 1753 166.67 30.3 23 10nm 1433 100 20.41 24 10nm 1289 90.91 18.18 25 10nm 1375 100 20.41 26 10nm 1863 250 19.61 27 10nm 1876 250 19.61 28 10nm 1859 250 25 29 10nm 1950 333.33 25 30 10nm 1910 333.33 41.67 31 10nm 1566 125 24.39 32 10nm 1863 200 28.57 33 10nm 1842 200 18.87 34 10nm 1841 200 40 35 10nm 1845 200 23.26 36 10nm 1827 200 20 37 Water 2114 1428.57 32.26 38 Water 2111 1428.57 28.57 Table 7.3 List of Test Two Samples with their respective SPIO molar concentration. Sample Number 30nm 30nm SPIO Molar Quantity (nM) T2* (ms) 39 0 17.53345659 40 0.000016524 24.17494292 41 0.000024786 35.79919487 42 0.000033048 48.74822399 59 43 0.00004131 46.61648252 44 0.000049572 45.3406074 45 0.000057834 40.6569547 46 0.000066096 16.67018219 47 0.000074358 21.03710934 48 0.00008262 29.23709996 49 0.0000918 34.54834725 50 0.000099144 26.75877179 51 0.000107406 33.15266425 52 0.000115668 35.78062851 53 0.00012393 12.52498074 54 0.000132192 20.9320007 55 0.000140454 22.50352592 56 0.000148716 22.5281837 57 0.0001836 23.40453537 58 0 35.11420547 59 1.5606E-06 12.14952518 60 3.1212E-06 40.05383465 61 4.6818E-06 36.20958199 62 6.2424E-06 36.96113631 63 0.000007803 90.48134242 64 9.3636E-06 62.4342734 65 1.09242E-05 97.57548901 66 1.24848E-05 26.60273035 67 1.40454E-05 37.5276156 68 0.000015606 30.75677022 Table 7.4 List of Test Three Samples with their respective SPIO molar concentration. Sample Number 30nm SPIO Molar Quantity (nM) T2*(ms) T2(ms) 69 0 59.6 2212.4 70 0.000016524 51.7 512.6 71 0.000024786 44.3 430.1 72 0.000033048 56.7 252.3 73 0.00004131 40.5 287.4 74 0.000049572 89 175.2 75 0.000057834 42.6 162 76 0.000066096 43 181.6 77 0.000074358 60.4 130.4 78 0.00008262 40 120.8 60 79 0.0000918 49.7 126.2 80 0.0001836 23.7 52.7 81 0.0002754 12.8 29.2 82 0.0003672 15.9 28.9 83 0.000459 10.7 20.5 84 0.0005508 12.9 15.5 85 0.0006426 10.7 15.4 86 0.0007344 10.1 13.1 87 0.0008262 8.5 10.5 88 0.000918 7.2 11.1 89 0.0010098 5.8 9.5 90 0.0011016 5.1 9.3 91 0.0011934 6.2 9 92 0.0012852 6.4 9 93 0.001377 5.7 7.4 Table 7.5 List of Test Four Samples with their respective MNP and Polystyrene-biotin microsphere molar concentration. Sample SPIO molar concentration (nM) BPM (?l) BPM quantity T2(ms) 97 0.0000918 0 0 122.6 98 0.0000918 1 253000 98.8 99 0.0000918 1 253000 121.4 100 0.0000918 2.7 683100 110 101 0.0000918 2.7 683100 100.7 102 0.0000918 5.4 1366200 145.6 103 0.0000918 5.4 1366200 103.9 104 0.0000918 10 2530000 122.5 105 0.0000918 10 2530000 113.5 106 0.0000918 20 5060000 124.9 107 0.0000918 20 5060000 104.6 108 0.0000918 30 7590000 99.1 109 0.0000918 30 7590000 96 110 0.0000918 40 10120000 118.3 111 0.0000918 40 10120000 118.3 112 0.0000918 50 12650000 133.9 113 0.0000918 50 12650000 117.9 114 0.0000918 8.1 2049300 95 115 0.0000918 8.1 2049300 103.9 116 0.0000918 100 25300000 217.8 117 0.0000918 100 25300000 210.8 61 118 0.0001836 0 0 53.3 119 0.0001836 1 253000 49.8 120 0.0001836 1 253000 51.2 121 0.0001836 2.7 683100 51.6 122 0.0001836 2.7 683100 53.2 123 0.0001836 5.4 1366200 57.2 124 0.0001836 5.4 1366200 54.3 125 0.0001836 10 2530000 55.6 126 0.0001836 10 2530000 62.4 127 0.0001836 20 5060000 55.6 128 0.0001836 20 5060000 52 129 0.0001836 30 7590000 55.4 130 0.0001836 30 7590000 57.9 131 0.0001836 40 10120000 55.1 132 0.0001836 40 10120000 55.2 133 0.0001836 50 12650000 57.6 134 0.0001836 50 12650000 56.1 135 0.0001836 8.1 2049300 54.5 136 0.0001836 8.1 2049300 52.3 137 0.0001836 100 25300000 88.7 138 0 0 0 822.4 Table 7.6 List of Test Five Samples with their respective SPIO nanoparticle and BPM molar concentration. Sample SPIO molar concentration (nM) BPM (?l) BPM particle quantity T2 (ms) 139 0.0002754 0 0 36.1 140 0.0002754 5.4 1366200 37 141 0.0002754 10 2530000 36.4 142 0.0002754 15 3795000 63.4 143 0.0002754 20 5060000 39 144 0.0002754 30 7590000 36.6 145 0.0002754 40 10120000 68.5 146 0.0002754 50 12650000 43.8 147 0.0002754 60 15180000 38.4 148 0.0002754 100 25300000 69.3 149 0.0002754 0 0 34.9 150 0 0 0 1069.5 Table 7.7 List of Test Six Samples with their respective SPIO nanoparticle and BPM molar concentration. Sample SPIO molar concentration (nM) BPM (?l) BPM particle quantity T2(ms) 62 139 0.000275 0 0 35.5 140 0.000275 5.4 1366200 34.8 141 0.000275 10 2530000 43.5 142 0.000275 15 3795000 70.8 143 0.000275 20 5060000 40.9 144 0.000275 30 7590000 43.1 145 0.000275 40 10120000 97.4 146 0.000275 50 12650000 73.1 147 0.000275 60 15180000 76.3 148 0.000275 100 25300000 320.9 151 0 0 0 1416.4 152 0.000275 5.4 1366200 49.4 153 0.000275 5.4 1366200 47 154 0.000275 5.4 1366200 55.8 155 0.000275 8.1 10 75.6 156 0.000275 8.1 10 55.9 157 0.000275 8.1 10 40.3 158 0.000275 8.1 15 88.6 159 0.000275 8.1 15 171.1 160 0.000275 8.1 15 31.1 161 0.000275 8.1 20 58.3 162 0.000275 20 5060000 58.2 163 0.000275 20 5060000 38.3 164 0.000275 30 7590000 36.8 165 0.000275 30 7590000 45 166 0.000275 30 7590000 56.3 167 0.000275 40 10120000 119 168 0.000275 40 10120000 30.8 169 0.000275 40 10120000 47.7 170 0.000275 50 12650000 51.3 171 0.000275 50 12650000 59.9 172 0.000275 50 12650000 56.1 173 0.000275 60 15180000 87.3 174 0.000275 60 15180000 82.6 175 0.000275 60 15180000 63.7 176 0.000275 80 20240000 89 177 0.000275 80 20240000 75.9 178 0.000275 80 20240000 96 179 0.000275 100 25300000 91.1 180 0.000275 100 25300000 101.3 181 0.000275 100 25300000 88.8 63 182 0.000275 0 0 39.3 Table 7.8 Test One Fitted Concentrations Sample Number T2 (ms) Fitted Concentration (nM) 1 45.45 0.000215428 2 25.64 0.000383195 3 14.49 0.000680402 4 10.75 0.000918773 5 8.7 0.001136716 6 11.11 0.000888825 7 9.43 0.00104821 8 8.55 0.00115678 9 7.94 0.001246207 10 7.3 0.001356152 11 6.23 0.001590591 12 5.52 0.00179649 13 5.52 0.00179649 14 5.08 0.00195307 15 5.99 0.001654713 64 References [1] J. Liu, Z. Cao and Y. Lu, "Functional Nucleic Acid Sensors," Chemical Reviews, vol. 109, no. 5, pp. 1948-1998, 2009. [2] P. K. Jain, X. Huang, I. H. El-Sayed and M. A. El-Sayed, "Noble Metals on the Nanoscale: Optical and Photothermal Properties and Some Applications in Imaging, Sensing, Biology, and Medicine," Accounts of Chemical Research, vol. 41, no. 12, pp. 1578-1586, 2008. [3] L. Wang, K. Wang, S. Santra, X. Zhao, L. R. Hilliard, J. E. Smith, Y. Wu and W. Tan, "Watching Silica Nanoparticles Glow in the Biological World," Analytical Chemistry, vol. 78, no. 3, pp. 646-654, 2006. [4] S. Bamrungsap, M. I. Shukoor, T. Chen, K. Sefah and W. Tan, "Detection of Lysozyme Magnetic Relaxation Switches Based on Aptamer-Functionalized Superparamagnetic Nanoparticles," Analytical Chemistry, vol. 83, no. 20, pp. 7795-7799, 2011. [5] H. Lee, E. Sun, D. Ham and R. Weissleder, "Chip-NMR biosensor for detection and molecular analysis of cells," Nat Med, vol. 14, pp. 869-874, 2008. [6] J. M. Perez, L. Josephson, T. O'Loughlin, D. Hogemann and R. Weissleder, "Magnetic relaxation switches capable of sensing molecular interactions," Nat Biotech, vol. 20, pp. 816-820, 2002. [7] L. Anderson, S. Holden, B. Davis, E. Prescott, C. Charrier, N. Bunce, D. Firmin, B. Wonke, J. Porter, J. Walker and D. Pennell, "Cardiovascular T2-star (T2*) magnetic resonance for the early diagnosis of myocardial iron overload," European Heart Journal, vol. 22, no. 23, pp. 2171-2179, 2001. [8] D. Hoa, Nuclear Magnetic Resonance, 2009. [9] J. P. Hornak, The Basics of NMR, 1997. [10] P. A. Hardy and R. M. Henkelman, "Transverse relaxation rate enhancement caused by magnetic particulates.," Magnetic Resonance Imaging, vol. 7, no. 3, pp. 265-275, 1989. [11] H. Duan, M. Kuang, X. Wang, Y. A. Wang, H. Mao and S. Nie, "Reexamining the Effects of Particle Size and Surface Chemistry on the Magnetic Properties of Iron Oxide Nanocrystals: New Insights into Spin Disorder and Proton Relaxivity," The Journal of Physical Chemistry C, vol. 112, no. 22, pp. 8127-8131, 2008. [12] H. Xu, Z. P. Aguilar, L. Yang, M. Kuang, H. Duan, Y. Xiong, H. Wei and A. Wang, "Antibody conjugated magnetic iron oxide nanoparticles for cancer cell separation in fresh whole blood," Biomaterials, vol. 32, no. 36, pp. 9758-9765, 2011. [13] Y. Chu, P. Zhang, J. Hu, W. Yang and C. Wang, "Synthesis of Monodispersed Co(Fe)/Carbon Nanocomposite Microspheres with Very High Saturation Magnetization," The Journal of Physical Chemistry C, vol. 113, no. 10, pp. 4047-4052, 2009. 65 [14] I. Koh and L. Josephson, "Magnetic Nanoparticle Sensors," Sensors, vol. 9, no. 10, pp. 8130-8145, 2009. [15] D. D. Stark, "Hepatic iron overload: paramagnetic pathology.," Radiology, vol. 179, no. 2, pp. 333-335, 1991. [16] A. Senpan, S. D. Caruthers, I. Rhee, N. A. Mauro, D. Pan, G. Hu, M. J. Scott, R. W. Fuhrhop, P. J. Gaffney, S. A. Wickline and G. M. Lanza, "Conquering the Dark Side: Colloidal Iron Oxide Nanoparticles," ACS Nano, vol. 3, no. 12, pp. 3917-3926, 2009. [17] H. Lee, T.-J. Yoon and R. Weissleder, "Ultrasensitive Detection of Bacteria Using Core???Shell Nanoparticles and an NMR-Filter System," Angewandte Chemie International Edition, vol. 48, no. 31, pp. 5657-5660, 2009. [18] H. G. R. L. I. A. Jaganathan, "Characterizing proton relaxation times for metallic and magnetic layer-by-layer- coated, DNA-templated nanoparticle chains," Nanotechnology, vol. 21, pp. 245103-245109, 2010. [19] I. Koh, R. Hong, R. Weissleder and L. Josephson, "Sensitive NMR Sensors Detect Antibodies to Influenza," Angewandte Chemie International Edition, vol. 47, no. 22, pp. 4119-4121, 2008. [20] T. J. Lowery, R. Palazzolo, S. M. Wong, P. J. Prado and S. Taktak, "Single-Coil, Multisample, Proton Relaxation Method for Magnetic Relaxation Switch Assays," Analytical Chemistry, vol. 80, no. 4, pp. 1118-1123, 2008. [21] L. Josephson, J. M. Perez and R. Weissleder, "Magnetic Nanosensors for the Detection of Oligonucleotide Sequences," Angewandte Chemie, vol. 113, no. 17, pp. 3304-3306, 2001. [22] D. Kim, M. K. Yu, T. S. Lee, J. J. Park, Y. Y. Jeong and S. Jon, "Amphiphilic polymer-coated hybrid nanoparticles as CT/MRI dual contrast agents," Nanotechnology, vol. 22, no. 15, p. 155101, 2011. [23] D. M. Ghesani and D. C. R. Goldfarb, General MRI. [24] C. Massin, F. Vincent, A. Homsy, K. Ehrmann, G. Boero, P.-A. Besse, A. Daridon, E. Verpoorte, N. de Rooij and R. Popovic, "Planar microcoil-based microfluidic NMR probes," Journal of Magnetic Resonance, vol. 164, no. 2, pp. 242-255, 2003. [25] D. M. Puddephat, Principles of magnetic resonance imaging, 2010.