by
Guangyi Cao
B.Sc., Shandong University, 2011
A Thesis Submitted in Partial Fulfillment of the Requirements for the Degree of
MASTER OF APPLIED SCIENCE
in the Department of Electrical and Computer Engineering
c Guangyi Cao, 2014 University of Victoria
All rights reserved. This dissertation may not be reproduced in whole or in part, by photocopying or other means, without the permission of the author.
Quantification of a Lung Cancer Biomarker Using Surface Enhanced Raman Spectroscopy by Guangyi Cao B.Sc., Shandong University, 2011 Supervisory Committee
Dr. Reuven Gordon, Supervisor
(Department of Electrical and Computer Engineering)
Dr. Jens Bornemann, Departmental Member
Supervisory Committee
Dr. Reuven Gordon, Supervisor
(Department of Electrical and Computer Engineering)
Dr. Jens Bornemann, Departmental Member
(Department of Electrical and Computer Engineering)
ABSTRACT
Detecting lung cancer is difficult as it is hidden in the body, and current clin-ical methods are not e↵ective at an early stage; the one-year survival rate after diagnosis in the World is just 29-33%. Acetyl amantadine (AcAm) is recognised as an exogeneous cancer biomarker because it is the product of a metabolic process known to be significantly up-regulated in cancerous cells. After ingestion, the an-tiparkinson and antiviral drug amantadine is acetylated in the body by the enzyme spermidine/spermine N1 acetyltransferase to give AcAm, which can be detected in patient’s urine. However, techniques previously used to quantify AcAm in urine, such as liquid chromatography-mass spectrometry (LC-MS), are undesirable for clin-ical adoption due to high costs and long run times. Further costs and delays result from the requirement for solid phase extraction (SPE). Therefore, it is highly de-sired to lower the costs and delays in processing by exploring di↵erent quantification approaches, ideally without the need for SPE processing.
In this thesis, I investigate the use of surface enhanced Raman spectroscopy (SERS) to quantify AcAm in urinalysis. I prepare two kinds of Raman substrates with hydrophobic pocket surface capture agents -cyclodextrin ( -CD) that work to extract the AcAm from the urine, followed by the surface enhanced Raman measurement using two kinds of Raman systems. The detection strategy is more economical than the currently used LC-MS approach, and enables development of an easy-to-use point-of-care tool that should provide a more rapid turnaround to
the health care provider. The next step will be to use real samples. If it is achieved, it will be a promising step in early cancer diagnostics.
Contents
Supervisory Committee ii
Abstract iii
Table of Contents v
List of Figures viii
List of Acronym xii
Acknowledgements xiv
Dedication xv
1 Introduction 1
1.1 Outline of the thesis . . . 3
2 Background 5 2.1 Lung cancer . . . 5
2.2 Early Lung Cancer Detection . . . 6
2.2.1 Amantadine (Am) . . . 7
2.2.2 Spermidine/Spermine Acetyltransferase (SSAT) . . . 8
2.3 Surface Enhanced Raman Spectroscopy . . . 9
2.3.1 Raman scattering . . . 9
2.3.2 Surface enhanced Raman scattering . . . 11
2.3.3 SERS substrates . . . 15
2.3.4 Applications to Analytical Chemistry . . . 17
2.3.5 SERS in Cancer Diagnostics and Imaging . . . 18
2.4 Supramolecule . . . 19
2.4.2 Cyclodextrins (CDs) . . . 20
2.4.3 Cucurbit[n]urils (CB[n]) . . . 22
2.5 Spectra Processing . . . 22
2.5.1 Overview of spectra processing . . . 22
2.5.2 Pre-processing . . . 23
2.5.3 Spectral analysis . . . 25
3 Quantification of a Cancer Biomarker Using -CDs Functionalized Gold Nanorods 28 3.1 Renishaw Raman Microspectroscopy System . . . 29
3.2 Methods . . . 29
3.2.1 Synthesis of gold nanorods . . . 29
3.2.2 Distinction between Am and AcAm . . . 30
3.2.3 Solid Phase Extraction (SPE) . . . 32
3.3 SERS analysis . . . 33
3.4 Results . . . 34
3.5 Conclusion and Discussion . . . 39
4 Quantification of an Exogenous Cancer Biomarker in Urinalysis by Raman Spectroscopy 42 4.1 Introduction . . . 42 4.2 Setup . . . 43 4.3 Materials . . . 44 4.4 Method . . . 44 4.5 Data analysis . . . 47 4.5.1 Summed peak . . . 48
4.5.2 Standard error of the mean . . . 48
4.5.3 Limit of detection (LOD) . . . 50
4.6 Results . . . 50
4.6.1 Reproducibility experiment . . . 53
4.7 Discussion . . . 54
5 Summary, Conclusion and Future work 58 5.1 Summary and Conclusion . . . 58
A Appendix A 61
List of Figures
Figure 2.1 (a):Non-small-cell lung caricinoma; (b): Small-cell lung carici-noma (Reprinted from Wikipedia). . . 6 Figure 2.2 Structure of Amantadine (Reprinted from Wikipedia). . . 7 Figure 2.3 Amantadine is acetylated by the high levels of SSAT to produce
AcAm. . . 8 Figure 2.4 Energy diagram illustrating Stokes and anti-Stokes Raman
scat-tering. The concept of virtual state has no physical meaning but it serves as a mathematical construction of perturbation theory. (This image is from the Dr. Aftab Amed’s thesis.) . . . 10 Figure 2.5 Schematic diagram illustrating a localized surface plasmon. . . 12 Figure 2.6 A dielectric metal sphere in a uniform electric field. . . 13 Figure 2.7 (a):TEM image of nano rods [45]. (b) Klarite Raman substrate. 17 Figure 2.8 Chemical structure of the three main types of cyclodextrins
(The image is from Wikipedia). In this thesis, I mainly used the -cyclodextrins (Reprinted from Wikipedia). . . 20 Figure 2.9 The schematic of how the mono-thiolated -cyclodextrins
en-capsulate on the gold surface. In this thesis, I mainly used the -cyclodextrins. . . 21 Figure 2.10Illustration of the concept of detection limit and quantitation
limit by showing the theoretical normal distributions associ-ated with blank, detection limit, and quantification limit level samples (Reprinted from Wikipedia). . . 27 Figure 3.1 Renishw InVia micro-Raman spectrometer. . . 30 Figure 3.2 Normalized spectra of AcAm and Am. Extra peaks in AcAm
at 675, from 950 to 1000, 1250 and around 1625 to 1650. Shift in peak locations near 1100, 1200 and 720. . . 31
Figure 3.3 SPE procedures. (a): Prime; (b):Load; (c): Wash and Eluent; (d):Dry with the nitrogen gas. . . 33 Figure 3.4 Raman mapping of the 10 mesh points. The dash lines below
each spectra are the baseline calculated by a fifth-order poly-nomial in a least-squares manner. . . 35 Figure 3.5 Spectra after the baseline removal. There are some overlaps
among the spectrum. For examples, the C-N binding mode at around 1450 cm 1, and the C=O binding mode at 1600 cm 1. 36
Figure 3.6 Sum of all of the 10 spectra of the drop. There are some new feature peaks in the spectrum after the sum of the 10 spectra. It is more appropriate to take the 10 spectra from the di↵erent locations of the drop than to take just 1 spectrum from the drop. 36 Figure 3.7 Summed spectra were preprocessed by Savitsky-Golay
smooth-ing. After the smooth of the spectra, the sharp peaks due to background noise in the spectra disappeared and the signal to noise ratio of the spectrum improved. . . 37 Figure 3.8 Summed spectra of the control experiments. The red line shows
the -CD functionalized gold nanorods with mock urine in-cluding AcAm. The black dash line shows the functionalized nanorods with mock urine which didn’t include AcAm. The blue dash line shows the pure nanorods in mock urine without
-CD encapsulation. . . 38 Figure 3.9 The calibration curve of di↵erent concentrations of AcAm in
mock urine samples. It shows the linear fit at the concentrations between 100 ng/mL and 600 ng/mL. . . 39 Figure 3.10Bright field images on the substrates. (a) The image taken by
20⇥ objective. (b) The image taken from the red square area in the left figure using 100⇥ objective. . . 41 Figure 4.1 Setup of low-cost Raman system. (1), Laser source, (2),
Cur-rent Driver, (3), Temperature Controller, (4), Emitting/Collecting Fiber probe, (5), Spectrometer. . . 43
Figure 4.2 Substrate preparation. Note that the Klarite sample needs to be rinsed with DI water and dried before it is ready to be im-mersed to the urine samples. After 4 h’s immersion in the urine sample, Klarite needs to be rinsed and dried again. . . 44 Figure 4.3 SEM images of Klarite Raman substrate with and without
-CD encapsulation. (a) SEM image of bare Klarite. (b) SEM images of Klarite with -CD functionalization. . . 45 Figure 4.4 AFM images of Klarite Raman substrate (a) before and (b)
after -CD functionalization. . . 45 Figure 4.5 Summed intensity of AcAm versus incubation time. There is a
plateau in the figure, which indicates that 4 h incubation time is enough for -CD to saturate the surface. . . 46 Figure 4.6 Histograms of the AcAm Raman intensity for each AcAm
con-centration. The distribution is close to the Gaussian distri-bution (indicated by the blue curve). (a) 1000 ng/mL. The horizontal arrow indicates the error which the vertical indicate the mean values. . . 49 Figure 4.7 Powdered AcAm of Raman spectrum and the regions of interest
shown in grey. . . 51 Figure 4.8 Spectra for 100 ng/mL of AcAm (red) with -CD and the
100 ng/mL of AcAm without using -CD (black). The regions of interests are shown in grey. . . 52 Figure 4.9 Summed intensity versus AcAm concentration. The background
level is shown in orange and limit of detection is shown in light blue, which corresponds to 1 ng/mL. The dynamic range is shown in light green, which is from 1 ng/mL to 300 ng/mL. The dash line shows the calculated signal of AcAm, which equals background level plus three times standard error of blank sample. 53 Figure 4.10Raman spectrum of corticosterone powder. . . 54 Figure 4.11Spectra for 100 ng/mL AcAm in artificial urine with and
with-out corticosterone. The regions of interest are shown in grey (AcAm) and blue (corticosterone) . . . 55
Figure 4.12Summed intensity versus AcAm concentration. The background level is shown in orange and LOD level is shown in light pur-ple, which corresponds to 30 ng/mL. The dash line shows the calculated signal of AcAm. . . 56 Figure 4.13AcAm summed intensity with di↵erent concentrations of
corti-costerone. The concentration of AcAm was 300 ng/mL. . . 56 Figure 4.14Summed intensity of AcAm versus concentration. The red dots
are from the data and the black dots with error bars are from the reproducibility experiments. . . 57 Figure 5.1 Feasibility studies on CB7. Highlight regions are the regions of
interest of AcAm. The figure shows that it is possible to use CB7 to replace the CDs. It also shows that CB7 has a quite large Raman cross section so it is good for us to quantify the CB7 as well. . . 59 Figure 5.2 Directivity Enhanced Raman Spectroscopy Using
List of Acronyms
AcAm Acetyl Amantadine
Am Amantadine
NSCLC non-small cell lung cancer
SCLC small cell lung cancer
CEA Carcinoembryonic antigen
SCC Squamous cell carcinoma antigen
SSAT spermidine / spermint N1 Acetyltransferase
SERS surface enhanced Raman scattering
CDs Cyclodextrins
S/N signal to noise
LOD limit of detection
LC-MS Liquid chromatography mass spectrometry
CTAB hexadecyltrimethylammoniumbromide
SPE Solid Phase Extraction
DMSO dimethyl sulfoxide
AFM atomic force microscope
SEM scanning electron mictroscope
AU Artificial urine
ACKNOWLEDGEMENTS I would like to thank:
My parents, badminton, and the weather, for supporting me in the low mo-ments.
Dr. Reuven Gordon for accepting me to his research group and allowing me to work under his supervision. His perpetual energy and enthusiasm in research has always motivated me to achieve better. I am also grateful to Dr. Reuven Gordon for his support and recommendation on my PhD application.
Dr. Tao Lu for the support on my studies and recommendation on my PhD ap-plication.
My friends and colleagues at UVic for the encouragements, guidances and sup-ports during my study at UVic.
Genome BC Proof-of-Concept grant and Mitacs Accelerate grant for fund-ing me throughout my program of study.
DEDICATION To my grandma.
Introduction
Lung cancer is the most common cause of cancer deaths in the world. Overall, 15% of people diagnosed with lung cancer survive five years after the diagnosis [1, 2]. Early stage lung cancer such as dysplasia and carcinoma in situ are only a few cell layers thick, they can be very difficult to visually detect by conventional diagnos-tic methods including the Carcinoembryonic antigen (CEA) and the Squamous cell carcinoma antigen (SCC) tests. Due to the drawbacks of traditional early lung can-cer test, researchers have introduced a new test by detecting Acetyl Amantadine (AcAm), a small molecule biomarker. Amantadine (Am) [3], a well-known clinically approved anti-viral drug, is acetylated by spermidine / spermint N1 Acetyltrans-ferase (SSAT) in order to produce AcAm. Therefore, the amount of AcAm excreted in urine serves as a simple proxy for in vivo SSAT activity in patients [4, 5, 6]. AcAm production is also up-regulated in other types of cancers such as pancreatic cancer and a neuroendocrine tumor.
The methods most frequently used for urinalysis are fluorescence spectroscopy and high performance liquid chromatography (HPLC) using UV, fluorescence [7, 8] or mass spectroscopy [9, 10]. A reported method requires multiple extractions such as solid phase extraction (SPE) and liquid-liquid extraction (LLE) followed by HPLC with fluorescence detection [11, 12]. The main drawback of this method is delays and cost because of multiple processing steps. Perhaps, there is a need for low-cost, high-speed approaches to urinalysis.
14]. It gives a fingerprint enabling characterization and identification of molecules. Raman scattering, however, su↵ers from extremely low intensity with typical cross section from 10 28 to 10 24 cm2 per molecule [15, 16]. Here we use surface
en-hanced Raman scattering (SERS) to enhance the weak Raman signal. SERS has been developed into a useful tool in trace analyte detection and even single molecule detection [17]. The SERS e↵ect has been attributed to an enhancement in the scat-tering efficiencies that occur for molecules localized near metal surfaces, namely, free electron metals such as silver, and gold. Generally, SERS substrates are based on Ag and Au in form of dispersed nanoparticles or arranged nanowires and nanorods arrays [18, 19, 20, 21, 22]. How to capture the molecules of interest e↵ectively and homogenously on the surface of the SERS substrate remains a big challenge. A supermolecular host, like Cyclodextrins (CDs), can be incorporated on the metal surface to capture the detection of at trace amount levels without having to phys-ically separate out interfering species. CDs are a family of supramolecular hosts, which consist of six or more a-D-glycopyranose units. Due to its internal hydropho-bicity, CDs can capture a variety of poorly water-soluble organic compounds [23]. Previous work shows analytes, such as PCBs [24] and canbendazim [25] etc., may be analyzed by decorating the gold or silver nanorods with CDs that are capable of trapping analytes. Thus, AcAm is a small hydrophobic molecule, so we can also use CD to capture it in urine samples.
Past clinical studies have looked at samples containing typically around 10 ng/mL of AcAm in urine [6]. These studies used Liquid chromatography mass spectrometry (LC-MS) with some extraction procedures. In this thesis, I will try to achieve the desired dynamic range, which is from 1 ng/mL to 300 ng/mL [26].
The research presented in this thesis was carried out at University of Victoria (UVic) Nanoplasmonics group under the supervision of Dr. Reuven Gordon. The author thanks the support from a Genome BC Poof-of-Concept grant and a Mitacs Accelerate grant. A portion of research was also conducted at Biomark Technologies Inc. Vancouver.
1.1
Outline of the thesis
In this work, I follow two di↵erent approaches, investigating the possibility for the early lung cancer detection using SERS.
Chapter 1 of this thesis is a brief introduction about the meanings and methods about this work.
Chapter 2 is a review about the SERS, lung cancer and the cancer diagnostics. Chapter 3 is based on the work published:
• Guangyi Cao, Maura Deway, Reuven Gordon. Quantification of Lung Can-cer Biomakers Using Surface Enhanced Raman Spectroscopy. In: 10th BiopSys network meeting, Toronto, September 24–26, 2013.
This poster was conducted under the supervision of Dr. Reuven Gordon. Maura Deway assisted in the experiments, sample preparation and data processing. Dr. Gor-don guided the author through all aspects of the work and provided editorial sug-gestions on the abstract and the poster.
Chapter 4 is based on the work published:
• Guangyi Cao, Ghazal Hajisalem, Wei Li, Fraser Hof and Reuven Gordon. Quantification of an Exogenous Cancer Biomarker in Urinalysis by Raman Spectroscopy. Analyst 139, 5375-5378 (2014).
The paper was conducted under the supervision of Dr. Reuven Gordon and Dr. Fraser Hof. Supervisors guided the author through all aspects of the work and provided editorial suggestions on the papers. Ghazal Hajisalem performed the AFM and SEM experiments. Wei Li performed the synthesis of the Acetyl Amantadine. Dr. Reuven Gordon also assisted in manuscript preparation.
The thesis describe the quantification of early lung cancer biomarker in urine us-ing SERS. The author used the di↵erent Raman substrates, includus-ing gold nanorods
and commercial Raman substrate Klarite. The author also performed the measure-ment using di↵erent Raman platforms, Renishaw confocal system and fiber-coupled Raman system.
Chapter 2
Background
2.1
Lung Cancer
Lung cancer is the most common cause of cancer deaths in the world. Overall, 15% of people diagnosed with lung cancer survive five years after the diagnosis, which is much lower than that for patients with cancers in other organs, such as bladder, breast, colon, cervix and prostate.
Lung cancer is a malignant lung tumor characterized by uncontrolled cell growth in tissues of the lung. Malignant means that it can spread, or metastasize, to other parts of the body. When cancer starts in lung cells, it is called primary lung cancer. The abnormal cell do not develop into healthy lung tissue, they divide rapidly and form tumors.
For therapeutic purposes, lung cancer are divided into non-small cell lung cancer (NSCLC) and small cell lung cancer (SCLC) based on the type of cell in which the cancer started. Figure 2.2 shows the two types of lung cancer.
Non-small cell lung cancer
NSCLC usually starts in glandular cells on the outer part of the lung [27]. This type of cancer is called adenocarcinoma. NSCLC can also start in flat, thin cells called squamous cells. These cells line the bronchi, which are the large tubes, or airways, that branch o↵ from the trachea, or windpipe, into the lungs. This type of
cancer is called squamous cell carcinoma of the lung. Large cell carcinoma is another type of non-small cell lung cancer, but it is less common. There are also several rare types of non-small cell lung cancer. These include sarcoma and sarcomatoid carcinoma.
Figure 2.1: (a):Non-small-cell lung caricinoma; (b): Small-cell lung caricinoma (Reprinted from Wikipedia).
Small cell lung cancer
Usually starts in cells that line the bronchi in the centre of the lungs [28]. The main types of small cell lung cancer are small cell carcinoma and combined small cell carcinoma.
Other types of cancer
Other types of cancer can spread to the lung, but this is not the same disease as primary lung cancer. Cancer that starts in another part of the body and spreads to the lung is called lung metastasis. It is not treated in the same way as primary lung cancer.
Early cancer detection and localization with e↵ective treatment is crucial to in-creasing the survival rates. Early stage lung cancer such as dysplasia and carcinoma in situ are only a few cell layers thick, they can be very difficult to visually detect by conventional diagnostic methods including the CEA and the SCC test.
Due to the drawbacks of traditional early lung cancer test, researchers have introduced a new test by detecting acetyl amantadine, a small molecule biomarker [6, 26]. Am, a well known clinically approved anti-viral drug, is acetylated by the high levels of SSAT in order to produce AcAm. Therefore, the amount of AcAm excreted in urine serves as a simple proxy for in vivo SSAT activity in patients. AcAm is also unregulated in other types of cancers such as pancreatic cancer and a neuroendocrine tumor.
Figure 2.2: Structure of Amantadine (Reprinted from Wikipedia).
2.2.1
Amantadine (Am)
Am [29, 30] (Figure 2.3) (trade name Symmetrel, by Endo Pharmaceuticals) is a drug that has U.S. Food and Drug Administration approval for use both as an an-tiviral and an anti-Parkinson’s drug. It is the organic compound 1-adamantylamine or 1-aminoadamantane, meaning it consists of an adamantane backbone that has an amino group substituted at one of the four methyne positions. Rimantadine is
a closely related derivative of adamantane with similar biological properties. Apart from medical uses, this compound is useful as a building block, allowing the insertion of an adamantyl group.
Am is used to treat Parkinson’s disease [31] and conditions similar to those of parkinson’s disease. It also is used to prevent and treat respiratory infections caused by influenza a virus.
2.2.2
Spermidine/Spermine Acetyltransferase (SSAT)
SSAT appears to have a role in the determination of tumor sensitivity to a new class of antinnaplastic agents [32]. It is ubiquitously distributed in mammalian tissues and plays a role in catabolism and elimination of polyamines from cells. SSAT is an inducible enzyme that catalyzes the transfer of an acetyl group from acetyl-coenzyme A to the aminopropyl moiety of polyamines. This action by SSAT facilitates polyamine degradation, excretion, cycling and/or intracellular cycling . In this manner SSAT participates in the maintenance of polyamine homeostasis in mammalian cells. Figure 2.4 shows how Amantadine is acetylated by the high levels of SSAT to produce AcAm.
Figure 2.3: Amantadine is acetylated by the high levels of SSAT to produce AcAm. In this thesis, the AcAm I used is synthesised by the Chemistry department at University of Victoria. Here is the procedure of the synthesis method of AcAm. Synthesis method of AcAm
Acetyl amantadine were prepared using a previously developed protocol [33]. Amantadine (2.00 g, 10.7 mmol) was dissolved in 30 mL dichloromethane under nitrogen atmosphere. Triethylamine (3.77 g, 37.5 mmol) was added to the mixture drop-wise. The resulting solution was stirred for 5 minutes at room temperature. After the addition of acetic anhydride (2.17 g, 21.3 mmol), the solution was stirred for one hour at room temperature. After an addition of water, the aqueous layer was extracted. The organic extract was dried over sodium sulfate, and filtered. A rotary evaporator was used to evaporate the solvent. High vacuum was applied to obtain AcAm in white solid form.
2.3
Surface Enhanced Raman Spectroscopy
2.3.1
Raman scattering
Raman spectroscopy [34] has been utilized as a spectroscopic technique used to study vibrational, rotational and other low-frequency modes in a system. It relies on inelastic scattering of monochromatic light from target molecules. The excitation light wavelength is usually around the visible, near infrared, or near ultraviolet range. Since the Raman process is inelastic process, it will be a energy shift. The shift in energy gives information about the vibration modes in the system. In a word, Raman spectroscopy provides a fingerprint enabling characterization and identification of the molecules, so it has inherent specificity.
Theoretical basis
For spontaneous Raman e↵ects, photons excite the molecules from the ground state to a virtual energy state. When molecules return to di↵erent ground states, photons with di↵erent wavelength will be emitted and the di↵erences in energy be-tween the original states and the final states lead to shift in the emitted photon’s frequency away from the excitation wavelength. If the final vibrational states of the molecule are more energetic than the initial states, then they will result in the longer wavelength emitted photons compared to the excitation photons, which are called
Figure 2.4: Energy diagram illustrating Stokes and anti-Stokes Raman scattering. The concept of virtual state has no physical meaning but it serves as a mathematical construction of perturbation theory. (This image is from the Dr. Aftab Amed’s thesis.)
Stokes shifts. If the final vibrational states are less energetic than the initial states, then the emitted photon will be shifted to a shorter wavelength, which are called anti-Stokes shifts. A change in the molecule polarization potential with respect to the vibrational coordinate is required for a molecule to exhibit a Raman e↵ect. (Fig-ure 2.6) The amount of the polarizability change will determine Raman scattering intensity. The pattern of shifted frequencies is determined by the rotational and vibrational states of the sample.
Raman shifts are typically reported in wavenumbers, which have units of inverse length, as this value is directly related to energy. In order to convert between spectral wavelength and wavenumbers of shift in the Raman spectrum, the following formula can be used. ! = ( 1 0 1 1 ) (2.1)
where ! is the Raman shift expressed in wavenumber, 0 is the excitation
wavelength, and 1 is the Raman spectrum wavelength. Most commonly, the unit
chosen for expressing wavenumber in Raman spectra is inverse centimeter (cm 1).
Since wavelength is often expressed in units of nanometers (nm), the formula above can scale for this unit conversion explicitly, giving
!(cm 1) = ( 1 0(nm) 1 1(nm) )⇥(10 7nm) (cm) (2.2)
2.3.2
Surface enhanced Raman scattering
Raman spectroscopy provides more structural information than fluorescence. It gives a fingerprint enabling characterization and identification of molecules. Raman scattering, however, su↵ers from extremely low efficiencies with typical cross section from 10 28 to 10 24 cm2 per molecule. Surface plasmons provide the solution to
the problem and the phenomenon is commonly known as surface enhanced Raman scattering (SERS). The strong electric field enhancement mechanism, which is called SERS, was initially discovered with the adsorption of molecules to rough metal surfaces [35]. Here, I use SERS to enhance the weak Raman signal. SERS was initially discovered with the adsorption of molecules to rough metal surfaces. SERS is a surface sensitive technique; the molecules must be on (or close to ) the surface. In this thesis, I will be using metallic nanoparticles and rough ended metal surfaces as well. The electromagnetic enhancement can be viewed as a redistribution of the electromagnetic field, resulting in localized regions of high field intensities.
However, the cause of SERS still remains in debate. There are two primary theories accounting for SERS, the electromagnetic enhancement mechanism and
Figure 2.5: Schematic diagram illustrating a localized surface plasmon. the chemical enhancement mechanism. In this thesis, I will mainly talk about the electromagnetic mechanism. When a metallic nanostructure interacts with the electric field of the incident light, then if the radiation’s wavelength is optimal, the electrons oscillate coherently producing a localized surface plasmon (Figure 2.7).
Plasmon
In the modern SERS literature, many e↵ects are attributed to plasmons and plasmon resonances. The use of the term was introduced by Pines in 1956 to ex-plain the energy loss. Since then, plasmons is therefore a quantum quasi-particle representing the elementary excitation, or modes, of the charge density oscillations in a plasma. A plasmon is simply to the plasma charge density what photons are to the electromagnetic field. However, a photon is as real quantum particle while a plasmon is a quasi-particle because it is always ”lossy” and highly interacting. A charge density oscillation, if not maintained by an external source of energy, will always decay because of various loss mechanisms.
Surface plasmons are the coherent oscillation of electrons that exist at the inter-face between two materials where the real part of the dielectric constant of the metal changes sign [36]. This condition is met with metal air and metal water interfaces when the real part of the dielectric constant of the metal is negative. Localized surface plasmons are surface plasmons that are confined to metallic nanoparticles or nanostructures.
The enhancement by the localized surface plasmon resonance can be approxi-mated by a dielectric sphere. This e↵ect can be approxiapproxi-mated by a dielectric sphere of radius r in a uniform electric field E0
^
z (Figure 2.8). The permittivities of outside and inside the sphere are "medium and "metal.
Figure 2.6: A dielectric metal sphere in a uniform electric field. The electric field from the dipole in spherical coordinates, ¯Esp is
¯ E = E0
^
The electric field from the dipole in spherical coordinates, ¯Esp is ¯ Esp = ( "⇤ metal "medium "⇤ metal+ 2"medium ) E0r 3 (r + d)3(2 cos ✓ˆr + sin ✓ ˆ✓) (2.4) Where "⇤
metal is the complex dielectric constant of the metal, "mediumis the
dielec-tric constant of the medium and ✓ is the angle with respect to the incident elecdielec-tric field. This results in strong electric fields at the surface of the sphere. The electric field enhancement factor, A , at a distance d away from the sphere is the ratio of the total electric field and the incident field. From the argument above, one can consider that a metal is good for plasmonics if:
• The real part of permittivities is negative when the laser is chosen to optimise. • The imaginary part of permittivities is small in the ranges of interest.
based on those condition and the optical properties of metal in the visible range citetagkey2009iv, it is easy to find that sliver and gold is the most promising one. T these theoremcal considerations, one should add the practical issues, such as toxicity, durability, availability and ease of fabrication. Thus, gold is the most suitable metal for SERS measurement. In this thesis, gold nanoparticles and gold nano surface structure were used.
A = E0+ Esp E0 ⇡ ( "⇤ metal "medium "⇤ metal + 2"medium )( r r + d) 3 (2.5)
From the equations above, the system is resonant when "⇤metal = 2"medium.
The materials and laser frequencies are chosen to optimize this e↵ect. In this case, metallic material will be suitable for the use in SERS because metallic material usually has a negative real part of the dielectric function and small imaginary part of the dielectric function. A comparison of the optical properties (real and imaginary parts of the dielectric function) of various metals is given in Figure 2.8. Silver and gold are most commonly used for SERS. In this thesis, gold nanoparticles and gold nano surface structure were used.
The intensity of Raman scattering is directly proportional to the square of the induced dipole moment, which is in turn the product of the Raman polarizability
and the magnitude of the incident electromagnetic field. As a consequence of ex-citing the localized surface plasmon resonance of a nanostructured or nanoparticle metal surface, the local electromagnetic field E is significant enhanced. The radio-tive photons are then magnified by the same mechanism, resulting in dramatically increase in the total output signal of the experiment. Since Raman scattering cross-section is approximately proportional to the fourth power of the local electric field E4, which means E4 folds SERS enhancement will be generated. The Raman signal
will then be enhanced by a factor M :
M ⇡ "metal "medium "metal+ 2"medium 4 ( r r + d) 12 (2.6)
2.3.3
SERS substrates
Since the discovery that high-intensity Raman scattering of small molecules could be obtained on electrochemical toughened silver surface by Fleischmann in 1974 [35], who attributed the high enhancement to the large number of molecules on the roughened surface, and Creighton and Jeanmaire [37] independent discovery in 1977 that the enhancement of the Raman scattering is related to an intrinsic surface enhancement e↵ect, marking the beginning of surface enhancement Raman scattering spectroscopy, substantial interest has been focused on the research of the fabrication of SERS-active substrates and on the applications of SERS to many fields, including surface, analytical and life science.
SERS has now become a very useful tool in various fields including chemistry, physics and biology. Among these, the basic focus as well as the key step for practical SERS applications is still the successful fabrication of the SERS substrate because the application really depends on the activity and reproducibility of the substrate.
The most used SERS substrates, however, are metallic nanoparticles by wet chemical methods and nano-structured metal surface, benefitting from the devel-opment of nanoscience and nanotechnology. Moreover, nanoparticles and nano-structured surface as SERS substrates have great advantages such as low cost as well as simple and easy manipulation.
In this thesis, I mainly used the two types of SERS substrates: Gold Nanorods in solution and Klarite Raman substrate (Figure 2.10).
Nanorods
Nanoparticles are artificially created structures with dimensions between 1 and 100 nm. According to the SERS enhancement mechanism, SERS phenomena will happen on metal surfaces with a certain degree of roughness. Metal nanoparticles with dimensions between 10 and 100 nm have the specific surface plasmon reso-nance property compared to their bulk materials, which is basic for the production of SERS and there will be very great enhancement of the Raman signal for molecules on or near their surface. From this point, SERS is indeed a nanostructure-enhanced Raman scattering. Therefore, the preparation of metal nanoparticle susbtrates has been the key issue that dictates signal intensity and reproducibility. As the initial SERS substrate in solution, metal nanoparticle colloid is the most commonly em-ployed SERS substrate because of its easy preparation and manipulation. Many methods have been reported for the preparation of metal nanoparticles with vari-ous sizes, shapes and composition [38, 39, 40]. Owing to the rapid development of this field, some good reviews and books exist that describe nanoparticles as SERS substrates from di↵erent aspects.
One of the typical methods for the preparation of nanoparticles with various shapes is by seed-mediated growth. So far, silver and gold nanorods and branched metal nanoflowers have been synthesized through the seed-mediated method and used as active SERS substrates. The most obvious example can be found in the preparation of gold NRs, which involves seed formation and growth [41, 42]. Gold NRs with di↵erent aspect ratios can be obtained by controlling the concentration of AgNO3, which plays an important role in the synthesis. Because gold NRs show
that transverse and longitudinal modes have a stronger absorption band and will red shift and increase in intensity with increasing aspect ratio, SERS activity can be tuned properly on NRs of di↵erent aspect ratios according to their surface plasmon bands [43]. Researchers have reported the synthesis of branched gold nanoparticles with the assistance of citrate by the seed-mediate method. Di↵erent sizes and shapes of gold nanoparticles could be obtained and the surface plasmon bands red-shifted with increasing size. A notable red shift can be clearly seen for the branched gold particles due to this unique structure providing the high electromagnetic field for SERS enhancement [44].
Figure 2.7: (a):TEM image of nano rods [45]. (b) Klarite Raman substrate. Klarite substrates
Klarite (Renishaw diagnostics, Glasgow, UK) features a systematically designed sub-micron scale patterning of a gold coated silicon surface. Compared with regular arrays of holes, the surface patterns form photonic crystals that control the surface plasmons, which, in turn, govern the SERS amplification.
Klarite substrates show nanotextured pyramidal subunits with⇠1.8 um openings arranged in a square lattice configuration at a separation of ⇠0.4 um.
2.3.4
Applications to Analytical Chemistry
The power of SERS lies in its ability to identify chemical species and obtain structural information in a wide variety of fields including polymer and materials science, biochemistry and biosensing, catalysis and electrochemistry. I highlight here a few exciting applications of SERS.
SERS is a highly sensitive and selective technique for use in the detection of biological samples. SERS sensing, a vast topic, has been reviewed in great detail elsewhere [46, 47]. SERS biosensors are used in detection of various biological sam-ples and diseases, including various cancers [48, 49, 50, 51], Alzheimer0s desease [52],
and Parkinson0s disease. A significant medical problem of the 21th century is the
growing incidence of diabetes mellitus, a disease. The Van Duyne group has made significant progress in the development of a SERS-based in vivo glucose sensor [53]. SERS probe
SERS as a molecular vibrational spectroscopic technique can be applied to a vast number of problems in diverse disciplines, including the analytical, biophysical and life sciences. The use of SERS probes as a labeling agent in targeted research is still in its infancy, but has a high potential for applications in bioanalytics and biomedicine. Central advantages of SERS labels are quantification, sensitivity and dense multiplexing for simultaneous target detection. Future applications of SERS probes should explore their sensitivity, multiplexing and size limits in vitro and in vivo. For all in vivo experiments with nanoparticles including SERS labels, in particular animal experiments, biodistribution and toxicity have to be intestigated. The interesting option to construct and use hybrid SERS probes in living cells [54].
2.3.5
SERS in Cancer Diagnostics and Imaging
Cancer diagnostics will certainly be an important field where this innovative methodology o↵ers significant advantages over existing approaches. For example, SERS-based assays targeting cancer biomarker, detection of circulation tumor cells using SERS. Recent advances in the preparation of SERS immunoassays include the combination of magnetic beads with antid-labeled AuNPs. For instance, MUC4 might be used as a serum marker for early detection of pancreatic cancer using a quantitative SERS-based platform [55].The possibility of monitoring more than one biomarker enhanced the accuracy of lung cancer diagnostics and facilitated the characterisation of small populations of cells . (Figure 2.12 and Figure 2.13)
SERS can also be used in cancer imaging [56, 57, 58, 59, 60, 61]. SERS tech-nologies have also progressed towards application in microscopy and small-animal in vivo imaging. The potential non-invasive utility of SERS is highly valuable for live imaging. Some works have been conducted in vivo imaging. For example, gold nanoparticles derivatized with a selected NIR SERS reporter were applied for ultra sensitive detection of HER2-position tumors in vivo [62]. Current researches in in vivo imaging are directed towards exploiting its multiplex capabilities to tack the SERS biomarkers in small-animal models [63, 64].
Other potential SERS application includes targeted drug delivery [65] and photo-thermal therapy [66]. Considerable e↵ort has been put into the development of SERS probes that integrate selective cell targeting and the action of cytotoxic agents (e.g., photo thermal therapy, singlet oxygen generation, and anticancer drugs) as theranostic tools for imaging and anticancer treatment.
The emerging integration of SERS probes with complementary imaging modali-ties represents a big leap towards the translation of SERS technologies to the clinal environment. With recent examples of SERS-guided intraoperative imaging for tu-mor resection and endoscope-based imaging, we envisage an important contribution from SERS technologies in the next generation of molecular imaging tools for cancer detection and therapeutics.
2.4
Supramolecule
There is a challenge to perform SERS. To perfrom the SERS, it is required to capture the interested molecules e↵ectively and homogeneously on the surface of SERS substrates. The solution will be using the mono-thiolated supramolecule, Cyclodextrin (CDs).
Figure 2.8: Chemical structure of the three main types of cyclodextrins (The image is from Wikipedia). In this thesis, I mainly used the -cyclodextrins (Reprinted from Wikipedia).
The term supramolecule was first introduced by Karl Lothar Wolf in 1937 [67]. It is defined complex of molecules held together by noncovalent bonds. It is often used to denote large complexes of molecules that form sphere, rod, or sheet-like species. The dimensions of supramolecular assembiles can range from nanometers to micrometers. Supramolecules also used to describe the complexes of biomolecules, such as peptides and oligonucleotides composed of multiple strands.
2.4.2
Cyclodextrins (CDs)
CDs are one of the most common supramolecule which are macrocyclic oligosugars shaped like a truncated cone with a hydrophobic internal cavity and a hydrophilic outer surface. The particular significance of CDs lies in their ability to form in-clusion complexes with other molecules. Moreover, on the basis of the reactivity of their external OH groups, chemically modified CDs can be synthesized to vary their solubility, to modify their complexation properties, and/or to introduce certain specific functional groups. These characteristics facilitate the control of enzymatic activity, not only by encapsulation of substrates or products but also by generat-ing new microenvironments around the enzyme when modified CDs are used. In a word, CDs has a hydrophobic cavity and hydrophilic exterior. They are used in food, pharmaceutical, drug delivery, and chemical industries, as well as agriculture
Figure 2.9: The schematic of how the mono-thiolated -cyclodextrins encapsulate on the gold surface. In this thesis, I mainly used the -cyclodextrins.
and environmental engineering.
In this thesis, mono thiolated -CD was used to capture the hydrophobic molecule AcAm in urine samples.
Structure
Typical cyclodextrins are constituted 7 glucopyranoside units, can be topologi-cally represented as toroids with the larger and the smaller openings of the toroid exposed to the solvent secondary and primary hydroxyl groups respectively. Because of this arrangement, the interior of the toroids is not hydrophobic, but considerably less hydrophilic than the aqueous environment and thus able to host other
hy-drophobic molecules. In contract, the exterior is sufficiently hydrophilic to impart cyclodextrins water solubility.
2.4.3
Cucurbit[n]urils (CB[n])
Cucurbit[n]uril (CB[n]) [68] are a family of supramolecular hosts, which are macrocyclic, rigid and highly symmetric. Their selectivity towards the shape and charge of a wide variety of guest is considerably higher than that of CDs. It is known to be capable of simultaneously encapsulating two guest molecules inside its cavity, forming a stable yet dynamic ternary complex. With the above unique host guest binding properties, CB[n] has recently been widely employed as a linking motif to prepare supramolecular polymers, micelles, dynamic hydrogels, microcapsules, and core shell polymeric colloids.
In this thesis, only the mono-thiolated -CD is used to be incorporated on the metal surface to capture the detection of at trace amount levels without having to physically separate out interfering species. The reason we do not use CB is because it is very difficult to be synthesized and purified. Another reason is that it is also hard to add the thiol ground to the CBs. Without a thiol group, CBs can not be easily functionalized on the gold.
2.5
Spectra Processing
2.5.1
Overview of spectra processing
Raman spectra typically contain multiple peaks with a wealth of information about the physical and chemical properties of molecules. Interpretation of this information can be challenging, especially for spectra from complicated biological molecules. The past two decades has seen increasing use of mathematical and sta-tistical methods to extract information from chemical data. These methods have developed into the field known as chemometrics.
For our purposes, it is useful to think of our measured data as a mixture of information plus noise. In a ideal world, the magnitude of the information would
be much greater than the magnitude of the noise, and the information in the data would be related in a simple way to the properties of the samples from which the data is collected. In the real world, however, we are often forced to work with data that has nearly as much noise as information or data whose information is related to the properties of interest in complex way that tare not readily discernable by a simple inspection of the data. These chemometric techniques can enable us to do something useful with such data.
We use these chemometric techniques to:
• Remove as much noise as possible from the data. • Extract as much information as possible from the data.
• Use the information to learn how to make accurate prediction s about the unknown samples.
A number of processing techniques are available, which enable analysis of Raman spectra. Pre-processing techniques remove noise and standardize spectra so that comparisons can be made. A number of di↵erent spectral analysis techniques can then be used to compare spectra from di↵erent samples. The mathematics behind some of these methods form the basis of entire textbooks and the following sections deal with the topics more on a conceptual level.
2.5.2
Pre-processing
Pre-processing techniques may be used to achieve smoothing, background reduc-tion and normalizareduc-tion [69].
Smoothing
Smoothing is applied to reduce sharp peaks due to background noise, with the aim of improving signal to noise (S/N) ratio of the Raman spectrum. Poor S/N ratio results from low scattering intensities, inherent in Raman spectra of tissue. It can also result from inefficiencies of the spectrometer grating and detector. A number of methods have been applied to achieve smoothing. Examples include adjacent averaging, Savitsky-Golay smoothing.
Savitsky-Golay smoothing applies as polynomial curve to a given number of points on either side of the center weight.
A description of adjacent averaging provides the simplest explanation of smooth-ing. In this method, each point of the smoothed spectrum is an average of adjacent points in the original spectrum. An equal number of data points on each side of the target data point are included in the average. This creates a problem at the beginning and end of the spectrum. For example, in a 10-point smoothing function with 5 points on either side of the target point, the first two and last two points in the spectrum should be discarded. Weighted averaging can be used.
Background reduction
Raman spectra often appear with sloped or curved background. This is due to fluorescence, Rayleigh ”wings” or other aberrations. It may be necessary to reduce the background to emphasize Raman peaks and enable comparisons between spectra. Several mathematical methods can be used, such as derivatives and baseline flattening.
Polynomial curve fitting [70] has a distinct advantage over other fluorescence reduction techniques in its ability to retain the spectral contours and intensities of the input Raman spectra. The fit will be based on minimizing the di↵erences between the fit and the measured spectrum, which includes both the fluorescence background and the Raman peaks.
Normalization
Normalization converts the intensity range of di↵erent spectra to the same or similar scale, which allows spectra to be combined and compared. Spectra can be normalized to a given frequency or to the total area under the spectrum. Normal-ization to a given frequency involves dividing the intensity of each frequency by the intensity of the given frequency. The intensity of the given frequency becomes 1 in all normalized spectra. This method depends on features of a single band and may distort the normalized spectrum if that band is not consistent between spectra. One way to overcome this problem is to normalize the spectra to the total area under
the spectrum, such that the area under the spectrum becomes 1. This method is independent of the features of any single band. A disadvantage is that background contributions can distort the normalized spectrum, and extraneous background in-formation should be removed first.
2.5.3
Spectral analysis
Spectral analysis techniques aim to identify unique spectral features that distin-guish one compound from another, or in the case of biomedical applications, one tissue type from another. The methods for analyzing spectra involved comparison of a few representative frequencies. Sometimes peak ratios were used. Selection of peaks was subjective, and consequently so also were the results. Some spectro-scopists became proficient at recognizing spectral patterns and could readily match unknown spectra with known library spectra. These techniques were restricted because only limited spectral information was used. Current methods utilize multi-variate analysis techniques, such as principal component analysis and Partial Least Regression to take advantage of as much spectral data as possible. The advent of the personal computer greatly facilitated application of these techniques.
In this thesis, since I have already known the regions of interests from the spec-trum of powdered AcAm, I am more interested in the intensity of the signals from these regions. So no multivariate analysis is needed for this thesis. Since I am talking about the clinical methods, I will also discuss about the limit of detection (LOD).
Limit of detection (LOD)
In analytical chemistry, LOD (limit of detection) is the lowest quantity of a substance that can be distinguished from the absence of that substance (a blank value) [71].The detection limit is estimated from the mean of the blank, the standard deviation of the blank and some confidence factor. Another consideration that a↵ects the detection limit is the accuracy of the model used to predict concentration from the raw analytical signal. There are a number of di↵erent ”detection limits” that are commonly used. These include the instrument detection limit (IDL), the
method detection limit (MDL), the practical quantification limit (PQL), and the limit of quantification (LOQ). Even when the same terminology is used, there can be di↵erences in the LOD according to nuances of what definition is used and what type of noise contributes to the measurement and calibration. In this thesis, I only consider the instrument detection limit.
Instrument detection limit (IDL)
Most analytical instruments produce a signal even when a blank (matrix without analyte) is analyzed. This signal is referred to as the noise level. The IDL is the analyte concentration that is required to produce a signal greater than three times the standard deviation of the noise level. This may be practically measured by an-alyzing eight or more standards at the estimated IDL then calculating the standard deviation from the measured concentrations of those standards. The LOD is the smallest concentration or absolute amount of analyte that has a signal significantly larger than the signal arising from a reagent blank. Mathematically, the analyte’s signal at the detection limit (Sdl) is given by:
SLOD = Sreag+ 3⇤ reag (2.7)
where Sreagis the signal for a reagent blank, reagis the known standard deviation
for the reagent blank’s signal.
The stated confidence limit of LOD will be 1 % generally because based on the definition of Gaussian distribution, about 99% of the sample are within three deviations from the mean (Figure 2.11).
Figure 2.10: Illustration of the concept of detection limit and quantitation limit by showing the theoretical normal distributions associated with blank, detection limit, and quantification limit level samples (Reprinted from Wikipedia).
Chapter 3
Quantification of a Cancer
Biomarker Using -CDs
Functionalized Gold Nanorods
Amantadine is acetylated by the enzyme spermidine/spermine N1 acetyltrans-ferase (SSAT) in the body; a process which is significantly up-regulated in can-cer cells. Therefore, the acetyl amantadine (AcAm) can serve as an exogeneous biomarker for the detection of cancers, specifically lung cancer which is difficult to detect at early stages by other methods.
Liquid chromatography mass spectrometry (LC-MS) has been successfully ap-plied for the detection and quantification of extremely low concentrations of AcAm. Facilitating accurate diagnosis of cancer at an early stage. LC-MS is a powerful tool but this method of detection is costly and time consuming. Therefore an efficient and cost e↵ective alternate solution is highly desirable for rapid economical testing. In this chapter quantification of AcAm was perform by SERS in urinalysis. AuNPs were prepared with hydrophobic pocket surface capture agents -cyclodextrin that work to extract the AcAm in mock urine. The SERS analysis were preformed by Renishaw confocal Raman microscope.
SERS was used for the detection of cancer biomarker AcAm using gold nanorods with a -cyclodextrin derivative to bind this molecule. Gold nanorods were synthe-sized with an aspect ratio of 4 to match the 785 nm excitation wavelength used
in the SERS studies. SERS analysis of the inclusion complex at di↵erent concen-trations in the presence of gold nanorods a↵orded good quality Raman spectra of AcAm at nanogram concentrations.
3.1
Renishaw Raman Microspectroscopy System
Renishaw mircoPL /Raman microscope is now installed in the Raman lab (Fig-ure 3.1). It is equipped with three lasers with pumping wavelength at 532 nm (for PL measurement only), 633 nm and 785 nm (for both PL and Raman measurement). Rayleigh line rejection filters for 633 nm and 785 nm excitation allow ripple-free measurement of the Raman spectrum to better than 100 cm 1 shift. Polarization
control is available for 325 nm and 633 nm excitation. Leica DM2500M microscope is equipped with objectives from 5⇥ to 100⇥magnification, allowing confocal Ra-man spectral measurements with better than 2.5 µm depth resolution (using 100⇥ objective). UV coated Deep Depletion CCD array detector (578⇥400 pixels) allows wavelength detection from 200 nm-1050 nm. A InGaAs point detector is also avail-able for near-IR wavelength detection. A motorized XYZ microscope stage with step size of 100 nm allows high spatial resolution mapping capability of a PL or a Raman spectrum. The liquid nitrogen sample cell is available for spectral mea-surement from 77 K to 420 C (max sample height 4 mm and diameter 22 mm). For even lower temperature measurements, liquid Helium cryostat is also available for spectral measurements from 3.2 K to 225 C (max sample height 5mm and dia 22mm)
3.2
Methods
3.2.1
Synthesis of gold nanorods
Gold nanorods with an average aspect ratio of 4 (length =40 nm, width = 10 nm) were synthesized following the seed-mediated method. Seed nanoparticles were synthesized by mixing HAuCl4 solution (0.12 mL, 15 mM) with an aqueous
Figure 3.1: Renishw InVia micro-Raman spectrometer.
deionized water, and 0.60 mL of ice-cold 0.010 M NaBH4. The growth solution was
prepared by mixing CTAB solution (5.36 g, 0.20 M), 4 mL deionized water, AgNO3
(0.4 mL, 4 mM), HAuCl4 (0.5 mL, 15 mM), ascorbic acid (0.124 mL, 0.0788 M),
and 0.1 mL of the 32 mins aged seed solution. The growth solution turned from colorless to reddish brown following incubation overnight at 27 C. Excess CTAB was removed by one centrifugation cycle at 9000 rpm for 30 min (Eppendorf centrifuge 5417R).
3.2.2
Distinction between Am and AcAm
In contract to previous reports [72], characteristic SERS vibration modes of Amantadine was investigated. These bands are tentatively assigned as two strong the CC stretching modes at 715 cm 1 and 775 cm 1, the CH stretching mode at
1100 cm 1, the CH
2 scissor modes at 1485 cm 1.
The distinction between Am and AcAm can be based on the vibrational band of carbonyl group at approximately 1600 cm 1wavenumber (a C=O stretching amide I
of Am and AcAm, this Raman band is of particular interest as it is present only in the spectrum of AcAm.
Figure 3.2 shows normalized spectra of AcAm of Am. From the spectra, there are some extra peaks in AcAm at 650, 950 and 1250 cm 1. There also are the shift
in peak locations at 720, 1100, 1200 cm 1.
Figure 3.2: Normalized spectra of AcAm and Am. Extra peaks in AcAm at 675, from 950 to 1000, 1250 and around 1625 to 1650. Shift in peak locations near 1100, 1200 and 720.
Urine is a concentrated solution of many salts, polar metabolites and multiple nonpolar steroids. Expected concentration of AcAm is about 1000 times smaller
than that of Am in urine samples. An artificial urine sample was prepared and di↵erent constituents of the sample were separated using solid phase extraction (SPE- Strata X). Thus instead of only relying on the Raman spectrum for detection of AcAm, the sample is pre-treated using SPE to get rid of the impurities. Details of the SPE are given below.
3.2.3
Solid Phase Extraction (SPE)
The objective is to provide a rapid and inexpensive pre-treatment method for enriching AcAm in urine samples using a solid-phase extraction cartridge. Primary objectives are as follows:
• Remove all salts and polar impurities.
• Increase the Acetyl Amantadine to Amantadine ratio. • Minimize contamination from nonpolar steroids.
The following protocol achieves all three aims. Figure 3.3 show the procedures of the SPE approaches. SPE phase = Strata X, Polymeric Reversed Phase. Phe-nomenex, Torrance, CA. To make the solutions, Air displacement micropipettes were used. The measured volume of the pipettes were about 1000 µL while the error of the pipettes were 0.1 µL.
1. Prime: 2 mL MeOH, 2 mL deionized H2O, 2 mL 50 mM pH 7.0 phosphate
bu↵er.
2. Load: Combine 2 mL of urine sample with 2 mL of 50 mM pH 7.0 phosphate bu↵er and load onto SPE cartridge.
3. Wash 1: 2 mL deionized H2O, 2 x 1.5 mL 50mM pH 7.0 phosphate bu↵er
(All salts and polar metabolites elute with this fraction).
4. Wash 2: 2⇥ 2 mL 40% methanol in H2O (Amantadine elutes mainly with this
fraction, while Acetyl Amantadine and the less polar steroid Corticosterone is retained).
5. Wash 3: 2 mL 100% methanol. (Acetyl Amantadine should elute with this fraction, while Corticosterone is retained).
6. Dry column by flushing air through it for a few minutes.
7. Eluent: 2 mL ethyl acetate (All remaining Acetyl Amantadine and Corticos-terone will elute).
Figure 3.3: SPE procedures. (a): Prime; (b):Load; (c): Wash and Eluent; (d):Dry with the nitrogen gas.
AcAm yields from the wash 3 while Am yields from the wash 2. Analyse wash 3 and fluent by removing ethyl acetate and methanol under vacuum or by blowing a stream of nitrogen over the samples. Reconstitute samples in 200 µL of toluene or other organic solvent for analysis. If analysis of other fractions (load, wash 1, wash 2) is desired they can be lyophilized before addition of toluene.
3.3
SERS Analysis
100 µL of gold nanorod solution was combined with 0.1 nM of -cyclodextrin in deionized water. These mixtures were incubated overnight. Then varied amounts of AcAm in methanol, obtained from wash 3 of theSPE protocol above, were added to the mixtures to give the final concentrations 100 ng/mL, 300 ng/mL and 600 ng/mL. These mixtures were incubated at room temperature for 4 h. A 2 µL aliquot was allowed to evaporate on a glass slide, followed by SERS analysis using a Renishaw InVia micro-Raman spectrometer with a 50⇥ objective. The laser beam size focused
on the sample was 1.5 µm in diameter and was positioned within the edge of the evaporated sample spot.
In SERS measurements, Raman spectra excited with a 785 nm laser line were acquired with a Renishaw InVia system spectrometer coupled to a Leica microscope. The laser power was set to 1% of the full power (approximately 80 µW). The laser beam was focused on the sample by a 5⇥ objective lens (NA =0.12). The spectra were measured with a 4 cm 1 resolution. A grating of 1800 lines/mm was used with
a RenCam charge-coupled device (CCD) (1040 ⇥ 256 pixels). Line mapping was performed with a StreamLine Raman mapping system, and the following conditions were used: spectral range of 300-1800 cm 1, and an acquisition time of 10 s.
Raman map of 10 mesh points were collected with 10 seconds of integration at each mesh point. Out of the 100 spectra, only those were retained which showed Raman peaks, the rest of the spectra were neglected. The quantification was based on the 1600 cm 1 band. The control experiments were conducted using the mock
urine samples without AcAm. I further studied the influence of -CD on this Raman approach. It shows that the CDs can extract the AcAm from the mock urine samples e↵ectively.
3.4
Results
Figure 3.4 shows Raman mapping of 10 mesh points. A fifth-order polynomial was derived by least squares fitting manner and substrate from each spectrum to remove the broad spectral background. The fit is based on minimizing the di↵erences between the fit and the measured spectra., which includes both the fluorescence background and the Raman peaks. The subsequent subtraction of this fit polynomial results in a spectrum that varies about the zero baseline. Polynomial curve-fitting has a distinct advantage over other fluorescence reduction techniques in its ability to retain the spectral contours and intensities of the input Raman spectra.
Figure 3.4: Raman mapping of the 10 mesh points. The dash lines below each spectra are the baseline calculated by a fifth-order polynomial in a least-squares manner.
spectra, most of spectra have the overlaps characteristic SERS vibration modes for AcAm. For example, the C-N binding mode at around 1450 cm 1, and the C=O
binding mode at 1600 cm 1.
Figure 3.6 shows the sum of the 10 spectra from the Figure 3.5. From the summed spectra, the CC stretching modes appear as a set of several peaks for both Am and AcAm, with two strong ones at 750 and 830 cm 1; the CH
2 scissor mode
appears at 1470 cm 1; the NH
2 torsion mode appears at around 1600 cm 1, which
indicated that there is actually AcAm extracted from the mock urine. Figure 3.7 shows the spectra which were preprocessed by Savitsky-Golay smoothing [73].
Figure 3.8 shows the spectra which are from the control groups. The control groups used the mock urine samples without AcAm. Furthermore, the influence of the nanorods without -CD encapsulation also had been studied.
Figure 3.5: Spectra after the baseline removal. There are some overlaps among the spectrum. For examples, the C-N binding mode at around 1450 cm 1, and the C=O
binding mode at 1600 cm 1.
Figure 3.6: Sum of all of the 10 spectra of the drop. There are some new feature peaks in the spectrum after the sum of the 10 spectra. It is more appropriate to take the 10 spectra from the di↵erent locations of the drop than to take just 1 spectrum from the drop.
Figure 3.7: Summed spectra were preprocessed by Savitsky-Golay smoothing. After the smooth of the spectra, the sharp peaks due to background noise in the spectra disappeared and the signal to noise ratio of the spectrum improved.
Figure 3.9 shows the calibration curve of di↵erent concentrations of AcAm in mock urine samples. It shows the intensity increases with the concentration of AcAm in mock urine. The minimum concentration could be detected is 100 ng/mL. From the figure, it also shows the linear fit at the concentration range between 100 to 600 ng/mL.
Figure 3.9: The calibration curve of di↵erent concentrations of AcAm in mock urine samples. It shows the linear fit at the concentrations between 100 ng/mL and 600 ng/mL.
3.5
Conclusion and Discussion
From the results section above, this approach is not good enough for us to quantify the cancer biomarkers AcAm urine samples. The reasons are as follows:
1. From the Figure 3.8, it is difficult to distinguish the AcAm from the Am in the mock urine samples. The spectrum of the mock urine without AcAm (the black dash line) looks almost the same as the spectrum of the mock urine which has AcAm in it (the red line). It is also obvious to see that there is a signal at 1600 cm 1 in the control samples. The reason we can think of is because of the
interferences from the mock urine solution, for instance, creatine or steroids. It seems that the SPE process does not help us remove these interferences from mock urine thoroughly and efficiently.
2. The concentration of AcAm at 100 ng/mL mininum could only be detected, which is not in the clinical range. According to some scientific papers, the clinical level for quantification is around 10 ng/mL [26]. Our result is 10 times larger than the desired clinical level. Moreover, it is note that the spectrum became inconsistent in the high wavenumber range. Figure 3.6 shows the inconsistent spectra when the wavenumber was higher than 1400 cm 1.
3. The distributions of functionalized nanorods on the substrate are not uniform, which makes the quantification approach less reproducible. Figure 3.10 shows the bright image of the nanorods on the substrates. With the di↵erent en-hancement factors at the di↵erent locations, it is almost impossible for us to get the consistent and reproducible results.
4. The Renishaw Raman microscope is too expensive, nearly 5 million dollars. The goal of this project is to develop a low-cost Raman platform for early lung cancer detection. To use the Renishaw spectrometer for the SERS measure-ment is not the good option because this setup is too expensive.
5. This approach is too slow for the project. It always takes 2 days to finish all the procedures. The SPE processing has further costs and delays.
Therefore, it is highly desired to lower the costs and delays in processing by exploring di↵erent quantification approaches, ideally without the need for SPE pro-cessing and using a low-cost Raman setup.
Figure 3.10: Bright field images on the substrates. (a) The image taken by 20 ⇥ objective. (b) The image taken from the red square area in the left figure using 100 ⇥ objective.
Chapter 4
Quantification of an exogeneous
cancer biomarker in urinalysis by
Raman spectroscopy
In this section, Raman substrates were prepared with hydrophobic pocket sur-face capture agents -CD that work to extract the AcAm directly from the urine, thereby removing the need for SPE. I also investigate the influence of hydrophobic interferences on this detection approach. The figures in this chapter are mainly from my paper which was published in the Analyst [74].
4.1
Introduction
Previous work has used solid phase extraction (SPE) and gold nanorods with -CD encapsulation as the SERS substrate using the Renishaw Raman system to quantify AcAm from urine samples at approximately 100 ng/mL levels. The use of Renishaw makes the process slow, costly and unefficient. Further costs and de-lays result from the requirement for SPE. Therefore, it is highly desirable to lower the costs and delays in processing by exploring di↵erent quantification approaches, ideally without the need for SPE processing.
Figure 4.1: Setup of low-cost Raman system. (1), Laser source, (2), Current Driver, (3), Temperature Controller, (4), Emitting/Collecting Fiber probe, (5), Spectrome-ter.
Recent reports show an increasing trend in applying Raman spectroscopy in clin-ical applications, such as medclin-ical diagnosis and chemclin-ical sensing. Raman scattering provides a fingerprint enabling characterization and identification of the molecules, so it has inherent specificity. Despite the success of Raman spectroscopy, the tech-nique still su↵ers from extremely low intensity with typical cross section from 10 28
to 10 24cm2 per molecule. Surface enhanced Raman scattering (SERS) can be used
to improve sensitivity in quantification.
4.2
Setup
Raman analysis was performed using this low-cost Raman system: A 785 nm fiber-coupled laser diode (Innovative Photonic Solution, 23 mW) was used for exci-tation, a QE 65 pro (Ocean Optics) portable spectrometer was used for detection, and a fiber optic probe (InPhotonics) was used for excitation, collection and
filter-ing. Twenty spectra were acquired for each sample with 30 s integration time for each spectrum at di↵erent random locations. The QE 65 pro spectrometer allows for up to 15 minutes exposure.
4.3
Materials
Sodium chloride, potassium chloride, potassium sulfate, urea, creatinine and cor-ticosterone were purchased from Sigma Aldrich. Amantadine was purchased from Tokyo Chemical Industry. Triethylamine and acetic anhydride are both from Ana-chemia. Mono thiolated beta-cyclodextrin was obtained from Shandong Zhiyuan Bio-Technology. Milli-Q deionized water (resistivity: 18.2 M⌦) was used for all preparations. All materials were used as received without further purification.
4.4
Method
Figure 4.2: Substrate preparation. Note that the Klarite sample needs to be rinsed with DI water and dried before it is ready to be immersed to the urine samples. After 4 h’s immersion in the urine sample, Klarite needs to be rinsed and dried again.
