Applications of Micro- and Nanoparticles in Activating Photodynamic Therapeutic Agents within Deep-seated Targets

Transcription

1 Applications of Micro- and Nanoparticles in Activating Photodynamic Therapeutic Agents within Deep-seated Targets By Erkinay Abliz B.A in Physics, July 1997, Xinjiang Normal University, P. R. China M.A in Physics, July 2004, City College of New York A dissertation submitted to The Faculty of The School of Engineering and Applied Sciences of The George Washington University in partial fulfillment of the requirements for the degree of Doctor of Philosophy January 31, 2012 Dissertation directed by Jason M. Zara Associate Professor of Engineering and Applied Science

2 The School of Engineering and Applied Science of The George Washington University certifies that Erkinay Abliz has passed the Final Examination for the degree of Doctor of Philosophy as of July 29, This is the final and approved form of the dissertation. Erkinay Abliz Applications of micro- and nanoparticles in activating photodynamic therapeutic agents within deep-seated targets Dissertation Research Committee: Jason Zara, Professor of Electrical Engineering, George Washington University, Dissertation Director Murray Loew, Professor of Electrical Engineering, George Washington University, Committee Member Shahrokh Ahmadi, Professor of Electrical Engineering, George Washington University, Committee Member Darrel Tata, Research Scientist, US Food and Drug Administration, Committee Member Ilko Ilev, Research Scientist, US Food and Drug Administration, Committee Member ii

3 Abstract of Dissertation Applications of micro- and nanoparticles in activating photodynamic therapeutic agents within deep-seated targets Photodynamic Therapy (PDT) is a therapeutic method that uses photo-sensitizers that can be preferentially localized in pathological tissue [1-3]. The dominant mode of photodynamic therapy (PDT) action is through the generation of reactive oxygen species (ROS). When the photo-sensitizer in tissue is excited by light, it interacts with molecular oxygen and transfers its energy to molecular oxygen to create highly reactive oxygen in its singlet state in tissues. Despite being non-invasive and having excellent selectivity for diseased tissue, PDT has not yet gained general clinical acceptance, largely due to the inherent limitations of light transport and penetration which restrict external light from activating photo-agents within target volumes deep inside the body. The photo-sensitizers that are approved for PDT treatment in oncology are found to maximally absorb light in the violet region of the visible spectrum, around 400 nm, and blood is a very strong absorber at this wavelength. Thus, the photo-agent s absorption characteristics inherently limit the effectiveness of PDT applications to target-sites that are shallow in depth, 2 3mm. For this reason, the clinical application of PDT has been limited to skin lesions, superficial solid tumors, or endoscopically-accessible regions [5]. One of the worldwide approved photo-sensitizers in oncology, Photofrin II, is known to have good selectivity towards diseased tissue, and its major sub-cellular target is known to be mitochondria [1-3]. In this work, both X-ray down-converting (DC) and Infrared upconverting (UC) particles were studied as platforms to generate visible luminescence to iii

4 activate the photo-sensitizer Photofrin II. Specifically, I have investigated DC particles composed of gadolinium oxysulfide doped with terbium (GdO 2 S: Tb) and UC particles composed of sodium yttrium fluoride co-doped with ytterbium and thulium (NaYF 4 : Yb/Tm). The DC and UC particles were tested in a cellular-like medium; the test tube with the DC particles was then irradiated with 120 kev X-rays, while the test tube with the UC particles was irradiated with a 980 nm laser. The ROS generation for each test tube was quantified by measuring the change in the absorbance of Vitamin C. In vitro studies on human glioblastoma cell lines were then conducted to investigate the possible cellular toxicity of these DC and UC particles through cell viability assays and an endotoxin detection assay. The therapeutic effectiveness of these particles via Photofrin II activation was also evaluated on in vitro human cancer cells through measurement of ROS levels and cell viability assays. Theoretical modeling of the experiment was generated using both analytical technique and Monte Carlo Modeling of light transport. The results obtained from cellular-like medium showed that both submicron- to micronsized DC and UC particles have great potential to activate Photofrin II and to generate substantial levels of ROS. Specifically, the results on in vitro cellular studies have shown that 20 micron-sized DC particles have great potential to activate Photofrin II in deep seated targets and to generate substantial levels of ROS and no potential cell toxicity was observed. However, the UC particles tested (50 nm) were shown to be toxic to the cell lines. The cell killing does not appear to be due to the particles' efficiency in activating the photo-sensitizer, but rather appears to be due to toxicity of the particles. iv

5 Table of Contents Abstract of Dissertation... iii Table of Contents...v List of Figures... vii List of Tables...x List of Acronyms...xi Chapter1: Introduction...1 Chapter 2: Background Mechanism of PDT Mechanism of Tumor Destruction Cellular Effects Vascular Effects Reaction of the Immune System Photosensitizer Tumor Oxygenation Chapter 3: Micro- and Nanoparticles Induced Visible Luminescence to Activate Photosensitizers within Deep- Seated Tumors Review of X-ray Production and Down Converting (DC) Micro-Particles X-Ray Generation X-Ray Down-Converting Particles Review of Up Converting (UC) Nanoparticles Chapter 4 : Experimental Quantification of ROS Generation Spectroscopic Characterization of Photofrin II, Up-converting (UC), and Down- Converting (DC) Particles Experimental Quantification of ROS Generation from DC and UC Particles v

6 4.3. ROS Generation from Photofrin II Activated by 405 nm and 633 nm Lasers ROS generation from X-ray down-converters ROS Generation from IR Up-convertors Chapter 5: Safety Evaluation of Rare-earth Based Materials and Therapeutic Efficacy on Selective Cancer Cell Lines Cell Maintenance, Cellular Metabolic Activity Measurement Techniques Therapeutic Efficacy and Cell Toxicity Results of X-ray DC Particles on Selective Cancer Cell Lines Therapeutic Efficacy and Cell Toxicity Results of Infrared UC Particles on Selective Cancer Cell Lines Chapter 6 : Theoretical modeling of ROS generation X-ray absorption coefficients of the test medium components Analytical modeling of X-ray absorbed dose and generated fluorescence light in the test medium in the presence of X-ray down convertors Statistical Modeling: Quantifying fluorescent light fluence distribution using Monte Carlo Modeling Theoretical quantification of amount of ROS generation Chapter 7: Conclusion...88 References...92 Appendix...96 vi

7 List of Figures Figure.2.1.Modes of Photodynamic killing...5 Figure 2.2.The Jablonski energy diagram photosensitizing process...6 Figure.3.1.Abrosption coefficients of whole blood...13 Figure.3.2. Two types of X-ray production...16 Figure.3.3.Principal UC processes for Lanthanide doped crystals...19 Figure.3.4.Proposed energy transfer mechanism showing UC processes...20 Figure.4.1.Spectroscopic characterization of Photofrin II...23 Figure.4.2. Spectroscopic characterization of X-ray of DC particles...25 Figure.4.3.X-ray induced light emission intensity dependence on X-ray photon energy...26 Figure.4.4.Experimental set up for measuring IR induced light emission spectrum of UC particles...27 Figure.4.5.Emission spectrum of UC particles in response to 980nm laser excitation...28 Figure.4.6.Emission intensity profile of peak values of UC particles...30 Figure.4.7.Quantum yield measurement system...32 Figure.4.8.Interaction of Vitamin C with resulting ROS in dehydroascorbic acid...33 Figure.4.9.Absorption spectrum of unoxidized Vitamin C in PBS Figure 4.10.Unoxidized Vitamin C absorbance in PBS as a function of concentration...34 Figure.4.11.ROS production from photo ii due to 633 nm and 405nm lasers...35 Figure.4.12.Experimental set up for measuring ROS generation from X-ray DC particles...37 Figure.4.13.Comparision of ROS production from Photofrin II between activation through DC particles and X-rays alone...37 Figure.4.14.Comparison of ROS production from Photo II between activation by 9mW/cm2 He-Ne laser and through the X-ray induced luminescence...38 Figure.4.15.Experimental set up for measuring ROS generation from IR UC particles...39 vii

8 Figure.4.16.ROS generation from UC particles...39 Figure.5.1.Structures of MTS tetrazolium and its formazan product Figure.5.2.X-ray exposure set up and measurement of cell viability 48 hours post exposure Figure.5.3.Human Glioblastoma cellular metabolic activity through MTS measurements taken 48 Hrs after a 15 Min diagnostic X-ray exposure Figure.5.4.Assessment on the potential cellular influence of Gd2O2S: Tb particles on human glioblastoma cell lines Figure.5.5.Infrared laser exposure set up and measurement of cell viability 48 hours post exposure Figure.5.6.Normalized Human Glioblastoma cellular metabolic activity through MTS measurements taken 48 Hrs after a 5 Min of laser exposure (5mg/ml) Figure 5.7. Normalized Human Glioblastoma cellular metabolic activity through MTS measurements taken 48 Hrs after a 5 Min of laser exposure (0.5mg/ml) Figure.5.8.Human Glioblastoma cellular metabolic activity through MTS measurements taken 48 Hrs after different exposure times and laser intensity...51 Figure.5.9.Normalized Human Glioblastoma cellular metabolic activity after 135 sec of laser(980nm) exposure...53 Figure.5.10.Typical standard curve for LAL assay Figure.6.1.Mass-energy absorption and attenuation coefficients at different X-ray photon energies for Gadolinium Oxysulfide Figure Excitation beam profile...68 Figure.6.3.Fluorescent photons are created at the point of photon absorption...71 Figure 6.4.Deflection of a photon by a scattering event Figure.6.5.Internal reflectance and transmittance Figure.6.6.Excitation photon tracking flow chart...76 Figure.6.8.Fluorescence photon tracking flow chart...78 viii

9 Figure 6.9.Results of X-ray photon simulation...79 Figure 6.9.Results of infrared photon simulation...80 ix

10 List of tables Table.2.1. Clinically approved photosensitizers in oncology...9 Table.5.1.Absorption values of dc particles at 405 nm using LAL assay Table.6.1.Physical properties of components of test medium at 120 kev X-ray exposure...61 Table.6.2.Comparision outcome between experiment and theory...84 Table.6.3.Connection chart between theoretical modeling and experimental measurements for DC particles Table.6.4.Connection chart between theoretical modeling and experimental measurements for UC particles 86 x

11 List of Acronyms PDT Photodynamic Therapy ROS Reactive Oxygen Species IR Infrared UC Up-conversion DC Down-conversion ESA Excited state absorption, ETA Energy transfer up-conversion PA Photon avalanche xi

12 Chapter1: Introduction Photodynamic Therapy (PDT) is a minimally invasive therapeutic method that uses photosensitizers that can be preferentially localized in pathological tissue. The ultimate result of the absorption of photons by the photosensitizer in the presence of ground tripletstate molecular oxygen results in the generation of highly cytotoxic species, including singlet oxygen ( 1 O 2 ), free radicals and peroxides, that attack the sub-cellular components of the targeted cells [1-3]. There is a general consensus in the literature that greater than 90% of cell killing is via generation of singlet oxygen. The singlet oxygen has a very short lifetime (<40 μs) in the water based biological environment and a very short free-diffusion radius of action (< 20 nm). Consequently, the damaged area is essentially confined within the tissue that contains the photosensitizer and then exposed to light with the appropriate wavelength [3]. Photodynamic therapy (PDT) has been used for many years to treat many different diseases, including macular degeneration, several skin disorders, and several types of cancers [2]. Compared to current treatments, such as surgery, radiation therapy and chemotherapy, PDT is relatively non-invasive, may be more accurately targeted, and is not subject to the total-dose limitations associated with radiotherapy. Despite these advantages, PDT has not yet gained general clinical acceptance. All of the photosensitizers that are approved for PDT treatment in oncology absorb light in the visible spectral regions below 640 nm, preventing access to more deeply seated tumors due to strong light absorption by blood in this wavelength range. As a result, the clinical application of PDT is limited to skin lesions, superficial solid tumors, or endoscopically accessible regions [4]. One of the 1

13 world-wide approved photosensitizers, Photofrin II, is known to have good selectivity towards diseased tissue and its major sub-cellular target is known to be mitochondria. Recently, there has been interest in the development of nanoparticles-based photosensitizer delivery system that is comprised of photo-agent molecules directly anchored at the surface of the nanoparticles. These nanoparticles are fabricated with an inorganic core, and absorb either incident X- ray photons (down-conversion) or multiple infrared photons (upconversion) and then relax to emit visible light at specific wavelengths determined by the material composition of the nanoparticles. The attached photo-sensitizer molecules could be activated directly through emitted light from the core of the nanoparticles or through direct-energy transfer schemes, resulting in copious reactive oxygen species (ROS) production [6-8]. However, the sub-cellular distribution of the photo-agent when ligated to the nanoparticles is expected to be vastly different than the free photo-agent. Free Photofrin II is known to have a high affinity for the mitochondria [1-3][9], whereas, the Photofrin II bound to nanoparticles is expected to be concentrated within the lysosomes of the cells[10]. As far as Photofrin II is concerned, the mitochondrion is its critical sub-cellular target from which apoptotic signals are delivered [1-3]. The dominant cell killing mechanism of action for Photofrin II is through the generation of singlet oxygen. Singlet oxygen molecules are very short lived (< 40 s with a free diffusion path length of < 20 nm). Therefore due to the very short singlet oxygen's diffusion path length, it becomes crucial that the Photofrin reaches the mitochondrial targets. In order to ensure that Photofrin II reaches mitochondrial target and effectively contributes to singlet oxygen generation, in my experimental design, the Photofrin II is separated from the nanoparticles. 2

14 In this dissertation, I shall begin Chapter 2 by providing background on the topics essential to the research: mechanisms of PDT, mechanisms of tumor destruction, mechanism of singlet oxygen generation and photosensitizer followed by background on micro/ nanoparticles in Chapter 3. In Chapter 4, I will describe results of spectroscopic characterization of Photofrin II, up-converting (UC) and down-converting (DC) particles; experimental quantification of ROS generation from X-ray induced DC luminescence and IR induced UC luminescence in cellular like medium. Chapter 5 includes experimental results on safety and effectiveness of rare earth based UC and DC particles in activating Photofrin II on human glioblastoma cell lines. Chapter 6 describes steps and results of theoretical modeling of ROS generation from Photofrin II activated through UC and DCinduced fluorescence. Finally Chapter 7 summarizes results, conclusions and proposed future work. 3

15 Chapter 2: Background Three critical elements are required for the photodynamic therapy process to occur: a photosensitizer agent that can be localized into pathological tissue and activated by light; a narrowband light exposure with wavelengths corresponding to the main excitation peak known as the Soret peak of the photosensitizer, and the presence of molecular oxygen [1-4]. The main controllable parameters that can influence the outcome of PDT treatment are the photo-agent s concentration and the light dose. For localized tumors it is also important to examine subtle parameters such as drug bio-distribution, localization, aggregation, oxygen supply and consumption and tissue optical properties, to enhance therapeutic efficacy, shorten treatment time, and eliminate skin photosensitization completely [2] Mechanism of PDT The cell killing mechanism in PDT is known to be predominantly through enhanced generation of reactive oxygen species (ROS) through type 1 and type 2 mechanistic pathways as shown in Figure.2.1 below[11]. Upon visible light activation, the excited photo-agent either transfers its excited electrons to molecular oxygen forming superoxide anions O 2 - and H 2 O 2 (type 1 pathway), or transfers its excited energy onto the ground (triplet) state 3 O 2, resulting in an excited oxygen molecule in its singlet state 1 O 2 (type 2 pathway). There is a general consensus in the literature that greater than 90% of cell killing is via type 2 pathways [11]. 4

16 Type I 3 * 3 P O 1 * P S P 2 S P O 2 1 h P 3 * P 3 1 Type II * P O P O O S 2 2 S( O) 2 Figure.2.1. Modes of Photodynamic killing. h: Photon energy at given frequency, P: photosensitizer, S: substrate, from[11] Figure.2.2 is a diagram of the energy level pathway for a photosensitizer in the photosensitizing process [12]. The molecules are excited from the ground state P 1, to the excited singlet state P1, with a probability proportional to the product of the absorption coefficient and irradiance. Once in the P 1 state, the molecule can relax by a fluorescent photon emission (with quantum yield ) or intersystem cross to the first fl triplet state, P3 (quantum yield isc ). From the triplet state, the molecule can either relax by phosphorescent photon emission (quantum yield ph ), or be quenched by interaction with a ground state triplet oxygen molecule 3 O 2, to produce excited singlet state oxygen 1 O 2. Photo bleaching of the molecule can come directly from P1 or P3 or 5

17 from P 1, P1 and P 3 in combination with 1 O 2, or from photosensitizer intermolecular interaction, resulting in destruction of the photosensitizer molecule. Figure.2.2. The Jablonski energy level diagram for a photosensitizer molecule in the photosensitizing process (indicated by dotted lines). P 1 : ground state photosensitizer in its singlet state, P1 : excited photosensitizer in its singlet state, a :Absorption coefficient of the tissue, : Irradiance, fl fluorescence quantum yield, ph :phosphorescence quantum yield, isc :inter-system crossing, P 3 :metastable triplet state photosensitizer, 3 O 2 ground state triplet molecular oxygen, 1 O 2 :excited singlet oxygen, modified from [12] Mechanism of Tumor Destruction PDT mediates tumor destruction by three mechanisms: direct cytotoxic effects of free radicals and oxidation products on tumor cells, damage to the tumor-associated vasculature, and activation of immune response against tumor cells [1][2][4][13]. These three PDT 6

18 effects influence each other, and the combination of all three of these factors is required for long-term tumor control Cellular Effects PDT induces cell death via either apoptotic or necrotic pathways. The drug incubation time prior to the administration of light influences the mode of cell death: longer incubation times (1 day) result in apoptosis and shorter incubation times (1 hour) result in necrosis [1-2]. As discussed earlier, the highly reactive 1 O 2 has a short lifetime (<40μs) and short radius of action (<20 ns). For lipophilic and anionic sensitizers this will damage all membranes including plasma, mitochondrial, and lysosomal membranes and also membranes of the nucleus and endoplasmic reticulum [2] [12]. There is evidence that the inactivation of membrane transport systems, plasma membrane depolarization and the inhibition of DNA repair enzymes may precede inactivation of mitochondrial enzymes; the latter is often the key event leading to cell death [2][12] Vascular Effects The effects of PDT on micro-vascular structures are rapid and dramatic and the consequences of this vascular damage for the tumor microenvironment are severe. It has been shown that Photofrin-PDT at high fluence rates can protect normal skin microvasculature while severely damages tumor vasculature and kills tumor cells. Occlusion of the tumor- surrounding vasculature can contribute to tumor control by depriving nutrients and retarding the vascular resupply of the tumor [14-19]. 7

19 Reaction of the Immune System PDT has also been shown to trigger an inflammatory response and enhance specific antitumor responses. Infiltration of lymphocytes, leukocytes, and macrophages into PDT treated tissue is an indication of activation of the immune system in response to PDT. The strength of the inflammatory response varies with the photosensitizer. For instance, Photofrin II-PDT induces a strong inflammatory response and rapid influx of neutrophils, which is critical to long-term tumor control [12-13]. Tumor tissue disruption is a direct effect of PDT, whereas the immune response is required to eliminate the surviving cells Photosensitizer Photosensitizers which undergo efficient intersystem crossing into the excited triplet state, and whose triplet state is long- lived enough to allow adequate time for interaction with oxygen, produce high yields of singlet oxygen [12][20]. Most photosensitizers in clinical use have triplet-state quantum yields from 40 % to 60% [19]. Table.1 shows the list of clinically approved photosensitizers in oncology, many of which were introduced in the 1980s and 1990s [12]. However, new generation PDT photosensitizers are continually being discovered and investigated. Photosensitizers can be categorized by their chemical structures and origins. In general, they can be divided into three broad families: (i) porphyrin-based photosensitizers (e.g., Photofrin, ALA/ PpIX, BPD-MA), (ii) chlorophyllbased photosensitizers (e.g., chlorins, purpurins, bacteriochlorins), and (iii) dyes (e.g., phtalocyanine, napthalocyanine). Most of the currently approved clinical photosensitizers belong to the porphyrin family. Traditionally, the porphyrin photosensitizers and those photosensitizers developed in the late 1970s and early 1980s are called first generation 8

20 photosensitizers (e.g., Photofrin). Porphyrin derivatives or synthetics made since the late 1980s are called second generation photosensitizers (e.g., ALA, and Photofrin II). Third generation photosensitizers generally refer to the modifications such as biologic conjugates (e.g., antibody conjugate, liposome conjugate) and with built-in photo quenching or bleaching capability [12][19]. The general guidelines for comparing different photosensitizers are based on:(i)low dark toxicity but strong photo toxicity, (ii) good selectivity towards target cells, (iii) longer excitation wavelength allowing deeper light penetration, (iv) biocompatibility and rapid removal from normal healthy tissues of the body, and (v) different routes of administration. There are only a few photosensitizers in oncology that have received official approval around the world [2] [3] [19]. Table.2.1 below lists a few photosensitizers that are used worldwide [14]. Table.2.1. Clinically approved photosensitizers in oncology, from [13]. Photofrin is the first photosensitizer approved by health agencies worldwide for the treatment of cancer [21-27]. Photofrin II is commercially available from Axcan Pharma, Inc. and has the longest clinical history and patient track record [12]. Canada approved 9

21 Photofrin II applications in 1993 for the treatment of bladder cancer. Photofrin II was approved in Japan in 1994 (for early stage lung cancer). It was approved by the U.S. FDA for clinical phase II trials in December 1995 esophageal cancer, and in 1998, it was approved for the treatment of early non-small cell lung cancer. In August 2003 the FDA approved its use for Barrett's esophagus, and endobronchial lesions. It is also being considered as a potential therapy against Kaposi s sarcoma, psoriasis and cancers of the head, brain, neck and breast and early-stage cervical cancers [20-22]. A major sub- cellular target for Photofrin II is known to be mitochondria. Photofrin II is also shown to have great selectivity toward diseased tissue [1]. In general, Photofrin II doses range from 1 to 2 mg per kilogram of patient s body mass. Patients are known to become susceptible to severe burns from bright light exposure including the sunlight during Photofrin II treatment, therefore, patients and their family need to be educated prior to receiving the Photofrin II to take appropriate precautions such as wearing clothing that covers the body completely. Patients shouldn t remain in dark room during the day either as photo bleaching by lowlevel light enhances clearance of the drug from the skin [2]. The mechanisms of action of Photofrin-mediated PDT include vascular endothelial cell damage with hypoxia and thrombosis, ischemic tumor cell necrosis, and intense local inflammation associated with immune response [23] Tumor Oxygenation Tissue oxygen supply is another important factor that affects PDT treatment outcome. Any reduction of oxygen supply reduces the amount of 1 O 2 generation, causing negative outcome of PDT treatment. Reduction can arise from many different sources such as 10

22 preexisting tumor hypoxia, vascular damage, and through rapid photochemical oxygen depletion during PDT treatment which is governed by the intensity of light exposure [2]. Because of deteriorating diffusion geometry, structural abnormalities of tumor micro vessels and disturbed microcirculation, solid tumors are prone to develop hypoxic regions within the tumor volume. With photosensitizers including Photofrin II that can constrict and occlude vessels, blood-flow obstruction can be remarkably large, restricting oxygen supply to the tumor [2]. Photochemical oxygen depletion will result if the rate of photodynamic oxygen consumption is faster than that rate of oxygen resupply from the vasculature. The oxygen depletion is found to depend upon: 1) the tissue concentration of the photosensitizer and its absorption coefficient value at the wavelength of excitation, 2) the intensity of light (i.e., fluence rate), and 3) the vascular supply of the tissue. If the first two parameters are high, and the third parameter is low, 1 O2 generation will be fast and oxygen depletion occurs rapidly [2]. Photobleaching is the photochemical destruction of the photosensitizer. Destruction of the photosensitizer through photobleaching will reduce the occurrence of oxygen depletion [2]. 11

23 Chapter 3: Micro- and Nanoparticles Induced Visible Luminescence to Activate Photosensitizers within Deep- Seated Tumors Compared to current cancer treatments, such as surgery, radiation, and chemotherapy, PDT is considered to be minimally invasive. Due to the photo-agent s high degree of selectivity in the diseased tissue, the PDT strategy offers a greater capability to accurately target and destroy the target of interest, and is not subject to the total-dose limitations associated with radiotherapy [4]. Despite these advantages, PDT has not yet gained general clinical acceptance. Photofrin II has its primary excitation maxima near 400 nm. However, human blood is the dominant absorber near 400 nm as well. Thus at the present time, the absorption of the surrounding tissue of the light needed to excited the Photofrin II inherently limits the use of PDT applications to the target-sites which are shallow in depth (~ 2 3mm). Consequently, the clinical applications of PDT have been limited to skin lesions, superficial solid tumors, or endoscopically accessible regions. To increase light penetration depth, PDT treatments are traditionally made with a red He-Ne laser at 633nm wavelength, where oxygenated blood is known to absorb considerably less (three orders of magnitude) than at 400 nm, as shown Figure. 3.1 [28]. 12

24 Figure.3.1. Absorption coefficients of whole blood RED = oxy-hemoglobin, BLUE = deoxy-hemoglobin, from[28] For these reasons, additional strategies need to be designed to activate photodynamic agents within deep-seated tumor locations in the body. One possibility to reach in deepseated tumors is the use of soft energy diagnostic X-rays and infrared lasers as noninvasive tools to produce visible light emission from rare earth particles. These particles are fabricated with an inorganic core, and absorb either incident X- ray photons (downconversion) or multiple infrared photons (up-conversion) then relax to emit visible light. Photosensitizer molecules could be activated directly through emitted light from the particles [6-8]. Section 3.1 provides background on X-ray generation and X-ray down converting particles; Section 3.1 provides background on Infrared up converting particles Review of X-ray Production and Down Converting (DC) Micro- Particles In this section principles of X-ray generation, and background on micron size DC particles will be summarized. 13

25 X-Ray Generation X-rays are produced when electrons (initially at rest) are accelerated under high electric potential difference between cathode and anode plates within a vacuum (X-ray) tube, and converting the kinetic energy of the accelerated electrons into electromagnetic radiation as a result of collisional and radiative interactions [29, 30]. The following events are required to produce X-rays. First, free electrons are required in the evacuated environment of the X- ray tube insert for electrical conduction between the electrodes; The next step involves application of high-voltage differential ( kv) by X-ray generator to the cathode and anode plates in order to accelerate the electrons to the electrically positive anode plate; X- rays are produced through the interaction between the highly energized electrons and anode plates (i.e., targets). Generally, targets are made of Tungsten due to its high atomic number (Z=74) and very high melting point. These properties facilitate efficient X-ray production and allow the target to tolerate the high-power deposition of the x-ray generation process without being destroyed [29]. There are two possible interactions of electrons with the target, resulting in the production of Bremsstrahlung (breaking radiation) and characteristic radiation, as shown in Figure 3.2 (a). The Bremsstrahlung X-rays result from the conversion of kinetic energy to electromagnetic radiation when the incident electrons are decelerated through the interaction in the vicinity of the target nucleus. Closer interactions between the nucleus and the electrons cause greater decelerations and result in higher X-ray energy. X-ray energy is at a maximum when electrons give up all their kinetic energy when stopped by target nuclei and are determined by the peak potential difference. The spectrums of X-rays are produced with minimum number at peak energy, and linearly increasing in number with 14

26 decreasing energy (see unfiltered Bremsstrahlung spectrum) [31]. It was theoretically formulated by Kramer that when the electron s velocity vector is perpendicular to its deceleration vector during the collision within the anode atoms, the spectral X-ray intensity distribution as a function of wavelength is given by[32, 33]: 2 I C. V. i. Z{ / 1}.(1/ ) (3.1) 0 0 Where Z is the atomic number of the anode material and i is the X-ray tube current. o = hc / (ev o ), where h is the Planck s constant and c = speed of light. o is known as the cut-off wavelength. Where V o is the electric plate potential difference. Characteristic X-rays result when the incident electron interacts with the target atom and removes the electron from its innermost K shell (Figure 3.2(b)). Because now the atom is energetically unstable, electrons from the other shells (L, M, N, and O) will make the transition to fill the K-vacancy. As a result, a discrete energy X-ray photon is created with energy equal to the difference in binding energies (for Tungsten, binding energies of the K, L, M,). The emitted X-ray energies are characteristic of the element (Tungsten) [29]. These characteristic X-rays add mono-energetic peaks to continuous spectrums. (See Figure.3.2 (b)). 15

27 (a) (b) Figure.3.2. X-ray production; ( a) X-ray production by energy conversion. Events 1, 2, and 3 depict incident electrons interacting in the vicinity of the target nucleus, resulting in Bremsstrahlung x-rays with the emission of a continuous energy spectrum of x-ray photons. Event 4 demonstrates characteristic radiation emission, where an incident electron with energy greater than the K-shell binding energy collides with and ejects the inner electron creating an unstable vacancy. An outer shell electron transitions to the inner shell and emits an x-ray with energy equal to the difference in binding energies of the outer electron shell and K shell that are characteristic of tungsten. (b) Bremsstrahlung and characteristic radiation spectra are shown for a tungsten anode with x-ray tube operation at 80, 100,120, and 140 kvp and equal tube current, from [29] X-Ray Down-Converting Particles Down conversion is a process in which the absorption of a high frequency photon (X-ray) yields to emission of output radiation in the visible range [34-38]. The phosphor Gd 2 O 2 S: Tb has been widely used in radiographic intensifying screens (scintillating screens) in medical imaging systems such as X-ray fluoroscopy, X-ray Computed Tomography, Single Photon Emission Tomography, and Positron Emission Tomography due to its high absorption of X-ray energy and efficiency in converting it into visible light [34-37]. X-ray scintillation of Gd 2 O 2 S requires some minute levels of crystal lattice packing defects for 16

28 significant visible light emission to occur. Small lattice packing defects are achieved through the introduction of a second rare-earth element dopant such as Tb [34-36] Review of Up Converting (UC) Nanoparticles Up conversion (UC) is a nonlinear process in which successive absorption of two or more near infrared wavelength photons leads to the emission of output radiation at shorter wavelength within the visible range, through intermediate long-lived energy states [39-41]. Lanthanide-doped phosphor UC nanoparticles were first utilized by Zijlman and coworkers to study biological recognition events in which submicron-size phosphor crystals ( μm) surface labeled with antibodies were utilized as a novel luminescent reporter for the sensitive detection of antigens in tissue sections or on cell membranes [39]. The UC technique significantly minimizes background auto-fluorescence, photo-bleaching, and photo-damage to the biological specimens and offers remarkable sample penetration depths that are much higher than those obtained by UV or visible excitation [41]. UC processes can be induced by low power (intensity is about 1mW/cm 2 ), cost-effective, continuous wave lasers. This is advantageous because low power lasers are required for biological applications in order to minimize surrounding tissue damage [41]. In recent years, biological applications of UC nanoparticles have been rapidly expanded to in vitro detection, in vivo imaging, molecular sensing, and drug delivery [41-46]. Inorganic crystals exhibit UC luminescence when lanthanide dopants are added to the crystalline host lattice in low concentrations. Efficient UC only occurs by using a small number of wellselected dopant host combinations. Rare earth fluorides are regarded as excellent host lattices for up-conversion luminescence of lanthanide dopants due to their high refractive 17

29 index and low phonon energy and ability to exhibit adequate thermal and environmental stability. Among the investigated fluorides, NaYF 4 has been found to be one of the most efficient UC host lattices and has attracted more attention in the field of materials science over the past two decades. The dopants are in the form of localized luminescent centers. The dopant ion radiates upon its excitation to a higher energetic state obtained from the non-radiative transfer of the energy from another dopant ion. The ion that emits the radiation is called an activator, while the donor of the energy is the sensitizer [41]. UC processes are mainly divided into three broad classes: excited state absorption (ESA), energy transfer up conversion (ETU), and photon avalanche (PA). All these processes involve the sequential absorption of two or more photons (Fig.3.3). Figure.3.3(a) shows ESA in which excitation takes the form of successive absorption of pump photons by a single sensitizer. The first pump photon promotes the dopant from the ground state (G) into a metastable state (E1). The second photon promotes the same excited dopant from E1 state to higher energy state E2. Optical transition from E2G results in higher energy photon emission. Figure.3.3 (b) shows the ETU process in which each of two neighboring ions populates the E1 level by absorbing a pump phonon of the same energy. One of the ions is promoted to the upper emitting E2 while the other ion relaxes back to ground state G by non-radiative energy transfer. Figure.3.3 (c) shows a PA process in which the meta-stable level population is established through the inverse population of the E1 level by nonresonant ground state absorption (GSA) followed by resonant ESA to populate the upper visible emitting level E2.Cross relaxation energy transfer then occurs between the excited ion and neighboring ground state ion causing both ions to occupy the E1 level. Then the 18

30 two ions populate E2 to further initiate cross relaxation, then strong UC emission is produced followed by an exponential increase in the E2 level population by ESA [41]. Figure.3.3. Principal UC processes for lanthanide-doped crystals :(a) excited state absorption, (b) energy transfer up conversion,(c) photon avalanche. The dashed/dotted, dashed, and full arrows represent photon excitation, energy transfer, and emission processes, respectively, from [41]. Requirements for efficient Lanthanide (La) luminescent bioprobes are: (i) water solubility, (ii) large thermodynamic stability, (iii) inertness, (iv) intense absorption above 330nm, (v) efficient energy transfer into La ion (vi) coordination cavity minimizing non-radiative deactivations, (vii) long excited state life time, (viii) ability to conjugate with bio-active molecules while retaining its photo physical properties without altering the bio-affinity of the host [40, 41, 46, 47]. The mechanism of up conversion for the Yb3+, Er3+ or Yb3+, Tm3+ co-doped nanocrystals has been extensively studied, and is illustrated in Figure.3.4. The absorber Yb3+ 19

31 ions absorb NIR light, followed by the energy transfer to the emitter Er3+ or Tm3+ ions that then emit visible light. Although the emitter can be excited directly, co-doping of the absorber with ions such as Yb3+ in the nanocrystals usually generates stronger up conversion fluorescence, because Yb3+ ions have a broad and strong absorption at 980 nm (the absorption cross-section of Yb3+ is 10 times larger than that of Er3+) [41]. Figure.3.4. Proposed energy transfer mechanisms showing the UC processes in Er3+, Tm3+, and Yb3+ doped crystals under 980-nm diode laser excitation. The dashed-dotted, dashed, dotted, and full arrows represent photon excitation, energy transfer, multiphonon relaxation, and emission processes, respectively. Only visible and NIR emissions are shown here, from [41] The fluorescence quantum yield (QY) can be defined as the ratio of photons absorbed to photons emitted. It gives the probability of deactivation of the excited state through the process of fluorescence. QY = no. of photons emitted / no. of photons absorbed (3.2) 20

32 The UC particle investigated in this work, NaYF 4 : Yb/Tm, is relatively new to the UC field of study and its optical and physical properties vary greatly based on how they are synthesized, and upon their surface modification properties [41, 48, 49]. It was reported previously that introduction of an inert crystalline shell of an undoped material around each doped nanocrystal can increase luminescence efficiency up to 30 times [49]. Determination the quantum yield of UC nanoparticles are very difficult because standards that show up-conversion property are not available and there have not been any reports until very recently. It was determined by Boyer et al that quantum yield of various sizes of NaYF 4 particles vary from 0.005% to 0.3 % [50]. 21

33 Chapter 4 : Experimental Quantification of ROS Generation This chapter describes the ROS generation from Photofrin II through the activation of X- ray DC particles, and infrared UC particles in a cellular-like medium. The ROS generation was quantified by measuring the change in the absorbance of Vitamin C at 266 nm. This vitamin C essay works due to the fact that when Vitamin C is oxidized it has no absorbance at 266 nm, so absorbance decreases with ROS generation. In order to understand and illustrate how the UC and DC particles physical properties are related to the ROS generation from Photofrin II, I also measured the spectral characteristics of both Gd 2 O 2 S:Tb and NaYF 4 :Yb/Tm particles and of Photofrin II Spectroscopic Characterization of Photofrin II, Up-converting (UC), and Down-Converting (DC) Particles I measured the spectral characteristics of both DC (Gd 2 O 2 S:Tb) and UC(NaYF 4 :Yb/Tm) particles and of Photofrin II to understand how these particles physical properties affect Photofrin II excitation. The light absorption spectrum of Photofrin II was measured using a Shimadzu, Inc. (UV-3101PC) scanning absorption spectrophotometer (Figure.4.1 (a)). The emission and excitation spectral characteristics of Photofrin II were measured using a Photon Technology International, Inc double monochromator fluorescence scanning spectrophotometer (Figure.4.1 (b) and Figure.4.1 (c)). The Photofrin II solution was prepared at a physiologically relevant concentration of 20 g /ml in Ca +2 free and Mg +2 free Dulbecco s PBS (In general Photo II dose ranges from 1 to 2 mg per kilogram of patient s body mass. Considering a patient with 70 Kg of body weight with 5000 ml of blood, 22

34 physiologically relevant concentrations range from 14 to 28 g /ml). For all of these measurements the samples were placed in a 4 ml quartz cuvette. Figure.4.1 (a), Figure.4.1 (b), and Figure.4.1(c) are the absorption, normalized emission and excitation spectrum of Photofrin II respectively nm nm 0.9 Absorption (OD) a Normalized emission 0.8 b nm (a) Wavelength (in nm) (b) Wavelength (nm) 1 397nm Normalized Emission (c) nm 504 nm 537nm 559nm Wavelength (nm) Figure.4.1. Spectroscopic characterization of Photofrin II. (a)absorption spectrum of Photofrin II in Dulbecco's PBS without Ca/Mg.[ Photo II ] = 10μg/ml. (b) Normalized Photofrin II emission spectrum in DPBS.[Photo II] = 20μg/ml, Excitation wavelength = 400nm. (c) Normalized Photofrin II excitation spectrum in DPBS. [Photo II] = 20 μg/ml, emission measured at 612nm, from [50]. We can see from Figure.4.1 that Photofrin II has an absorption maximum at 366 nm (Figure.4.1 (a)), an emission peak of 612 nm, (Figure.4.1 (b)) and a main excitation peak 23

35 (i.e., the Soret band) near 400 nm (Figure.4.1(c)). These results are of interest because they allow us to understand the connection between the activated amount of Photofrin II and the resulting amount of ROS generation. Since Photofrin II has Soret a band near 400 nm, we expect the highest amount of Photofrin II activation and ROS generation when it is excited with a 400 nm source. In order to measure the emission spectrum of the Gd 2 O 2 S:Tb particles, they were placed into a 15 ml polystyrene test tube in powder form and irradiated with diagnostic X-rays operating with a constant X-ray tube current of 20 ma and X-ray tube potential differences of 130 kev. The X-ray induced spectrum from the particles was measured using an Ocean Optics, Inc. fiber optic spectrometer spanning the wavelength range from 200 nm to 900 nm (Figure.4.2 (a) is the experimental set up and Figure.4.2 (b) is the result). The beam spot size was 10 cm. The sample was 1 meter away from the source and 0.5 meters away from the collimator. 24

36 Xraygenerator Beamspotsize Testtube Spectrometer (a) Collimator 1 544nm (b) 493nm 588nm 622nm Wavelength (in nm) Figure.4.2. Measuring X-ray induced light emission spectrum from Gd 2 O 2 S:Tb particles. (a) Experimental set up for measuring X-ray induced light emission spectrum from Gd 2 O 2 S:Tb particles. (b) Normalized X-ray induced light emission spectrum from Gd 2 O 2 S:Tb particles with X-ray tube settings of 130kVp operating with a current of 20mA, from [50]. Figure.4.2 (b) shows the emission spectrum of Gd 2 O 2 S: Tb particles. We could see that Gd 2 O 2 S: Tb particles have their emission peak at 544 nm. Since the Gd 2 O 2 S: Tb emission peak of 544 nm does not match the Photofrin II Soret band of 397 nm; we expect only partial activation of Photofrin II when excited by emission from Gd 2 O 2 S: Tb particles. In order to understand how the luminescence intensity is related to X-ray energy, Gd 2 O 2 S:Tb particles, in powder form, were placed into a 15 ml polystyrene test tube and 25

37 irradiated with diagnostic X-rays operating with a constant X-ray tube current of 20 ma and various X-ray tube potential differences ranging from 10 kev to 130 kev. We measured the X-ray induced emission intensity at 544 nm wavelength as a function of X- ray energy using a Newport, Inc. light power meter (The experimental set up was similar to the Figure.4.2(a), however the power meter was replaced with the spectrometer).the power meter was placed in line with and 1cm away from the polystyrene test tube. 1.2 y = -1E-12x 6 + 6E-10x 5-1E-07x 4 + 8E-06x x x R 2 = 1 1 Intensity (uw/cm^2) Maximium X-ray Photon Energy (in kvp) Figure.4.3. X-ray induced light emission intensity dependence on the maximum X-ray photon energy (with a fixed tube current of 20mAs) from rare-earth (Gd 2 O 2 S:Tb) particles, from [50]. We could see from Figure.4.3 that, above 60 kev of X-ray excitation energy, there is a relatively linear relationship between the X-ray energy and fluorescence intensity measured at 544 nm emission wavelength. From this result we expect that the amount of Photofrin II, 26

38 which is activated at 544 nm, and the generated ROS, to increase with increased X-ray energy. In order to understand the emission characteristics of NaYF 4 :Yb/Tm particles in pellet form, 0.5 cm in diameter (since the materials are lyophilized there is a water-soluble portion that has been freeze-dried, each pellet has 0.5 mg of nanocrystal) were placed on a horizontal stage 1 cm away from laser source and irradiated with 980 nm laser with an intensity ranging from 150 to 1000 mw/cm 2 (Figure.4.4 shows the experimental setup). The infrared laser-induced emission spectrum from the NaYF 4 : Yb/Tm particles were then collected by using an Ocean Optics, Inc. fiber optic spectrometer over the wavelength ranged from 200 to 900 nm (results are shown in Figure.4.5). The fiber tip was perpendicular to the laser beam direction. UVIRfiber LaserDiode P Spectrometer Signal acquisition Figure.4.4. Experimental set up for measuring IR induced light emission spectrum of NaYF 4 :Yb/Tm up- convertors. P: particles in pellet form. Figure.4.5 (a) shows the emission spectrum of the NaYF 4 : Yb/Tm up convertors. Figure.4.5 (b) is the same emission spectrum of the NaYF 4 : Yb/Tm particles not including 802 nm peaks in order to intensify other emission peaks. 27

39 Fluorescence intenity (Arbitary units) (a) Infrared induced light emission spectrum from NaYF4:Yb/Tm particles with 980nm laser operating at 950mW/cm^2 802nm nm 644nm nm 450nm Emission wavelength (nm) Fluorescence intensity(arbitary units) (b) Intensity profile of particles ecxited with 980nm Laser nm nm nm nm Emission wavelength(nm) Figure.4.5. Emission spectrum of the NaYF4:Yb/Tm up converter(a) Emission profile of NaYF 4 :Yb/Tm in response to 980nm laser excitation including 802 nm emission peak;(b) Emission profile of NaYF 4 :Yb/Tm in response to 980nm laser excitation not including 802 nm emission peak We could see from Figure.4.5 that NaYF 4 : Yb/Tm particles have several emission peaks (360 nm, 450 nm, 475 nm, 644 nm, and 802 nm) in UV-VIS-NIR. Since emission peak 28

40 values from these particles are different from the Photofrin II Soret band of 400 nm, we expect there will be a lower amount of Photofrin II activation and ROS generation when activated by NaYF 4 : Yb/Tm particles compared to activation by a 400 nm source. I attempted to obtain the absorption spectrum of NaYF 4 : Yb/Tm particles in PBS solution using a Shimadzu, Inc. (UV-3101PC) scanning absorption spectrophotometer. However, I was not able to obtain the absorption spectrum due to the strong and inevitable absorbance of water around 1000 nm. We expect that these particles would exhibit an absorption peak at 980 nm since they only emit fluorescence when they are excited with 980 nm laser. In order to see how the laser excitation power is related to emission intensity, the NaYF 4 :Yb/Tm particles in pellet form (0.5cm in diameter, since the materials are lyophilized there is a water-soluble portion that has been freeze-dried, each pellet has 0.5 mg of nanocrystal) were placed on a horizontal stage 1 cm away from Laser source and irradiated with a 980 nm laser with an intensity ranging from 150 to 1000 mw/cm 2 (setup shown in Figure.4.4). The infrared laser induced emission spectrum from the NaYF 4 : Yb/Tm particles were measured using an Ocean Optics, Inc. fiber optic spectrometer over the wavelength range nm. Figure.4.6 (a) shows emission intensity profile of peak values in response to various laser intensities. Figure.4.6 (b) is the same emission intensity profile of the NaYF 4 : Yb/Tm particles, not including values at 802 nm in order to more clearly show the variations in the other intensity values. 29

41 Fluorescence intensity profile of NaYF4:Yb/Tm particles in response to different 980 laser excitation intensity Fluorescence Intensity(A rbitary units) nm laser Intensity (mw/cm^2) 265nm 349nm 360nm 450nm 475nm 644nm Fluorescence intensity profile of NaYF4:Yb/Tm particles in response to different 980 laser excitation intensity Fluorescence intensity (Arbitary units) nm 349nm 360nm 450nm 475nm 644nm 802nm 980nm laser intensity(mw/cm^2) Figure.4.6. Emission intensity profile of peak values,(a)fluorescence intensity profile of NaYF 4 :Yb/Tm in response to different 980nm laser excitation intensity including 802 nm emission peak;(b)fluorescence intensity profile of NaYF 4 :Yb/Tm in response to different 980nm laser excitation intensity not including 802 nm emission peak. 30

42 We could see from Figure.4.6 that there is a relatively linear relationship between the Infrared excitation power (intensity) and fluorescence intensity when the excitation intensity is lower than 1W/cm 2. I attempted to measure fluorescence intensity in absolute units using a Newport, Inc. light power meter placed 1cm away from the pellet and perpendicular to the laser beam direction (the experimental set up is similar to Figure.4.4, a power meter was replaced with spectrometer), but I wasn t able to measure fluorescence intensity in absolute units due to very weak fluorescence signal. In a previously published paper it was described how very high fluorescence intensity was achieved in absolute units as a function of excitation power (they reported 40 mw of fluorescence emission for 200 mw excitation power)[48]. In that particular experimental setting the particles in solution were placed in a cuvette and irradiated with 980 nm laser on one side. Then the fluorescence intensity was collected from the other side. It seems that the achieved result may have been produced by intensity of the excitation laser. In fact I believe it was caused by the geometry of their setup. I also attempted to make quantum yield (QY) measurements using the system described in Figure.4.7. The system was composed of a barium sulfate coated integrating sphere, UV- VIS transmitting optical filter, Cuvette holder, and 980 nm Laser source. The sample was held in a quartz cuvette located in the center of the integrating sphere. The sample was excited with a 980 nm laser diode. The light was delivered to the entrance port using a high efficiency fiber and was collimated to a beam diameter of 1 mm and directed on the sample. The emission intensities were measured using a Newport, Inc. light power meter with and without the UV-VIS transmitting optical filter in order to separate the fluorescence signal from the excitation signal. 31

43 IRlaser source Integrating sphere F Spectrometer Figure.4.7. Quantum Yield measurement system. F: UV-VIS transmitting optical filter I was not able obtain Quantum Yield using the system described in Figure.4.7. This is believed to be due to the fact that the fluorescence intensity was very weak compared to the excitation light power and therefore the integrating sphere and optical detectors were not sensitive enough to measure the fluorescence intensity. As has been described earlier, only measuring the values of relative units is not adequate for evaluating the effectiveness of these particles and makes it difficult to theoretically predict the efficiency of these particles in Photofrin II activation. Determination of the quantum yield of UC nanoparticles is also very difficult because standards that show upconversion properties are not available and there have not been any reports until very recently. It was determined by Boyer et al that quantum yield of various sizes of NaYF 4 particles vary from 0.005% to 0.3 % [51]. Since my Quantum Yield measurements were not successful, I had used 0.1% for my modeling. 32

44 4.2. Experimental Quantification of ROS Generation from DC and UC Particles Quantification of ROS generation was made through Beer s Law and the change in absorbance of un-oxidized Vitamin C at 266 nm (Figure.4.8). This assay is made possible due to the fact that oxidized Vitamin C has no absorbance at 266 nm, i.e., its molar extinction coefficient () is zero at 266 nm, whereas unoxidized Vitamin C has a large of ~15,400 M -1 cm -1 in PBS. ROS + H 2 O Vitamin C (Unoxidized) Ascorbic Acid Vitamin C (Oxidized) Dehydroascorbic Acid Fig.4.8. Interaction of Vitamin C with ROS resulting in dehydroascorbic acid, modified from[51] Vitamin C with 100 M/ml concentration in Ca +2 free and Mg +2 free Dulbecco s PBS were placed in a 4 ml quartz cuvette and its light absorption spectra were measured using Shimadzu, Inc. (UV-3101PC) scanning absorption spectrophotometer. Figure.4.9 shows that unoxidized Vitamin C has its highest absorption peak at 266 nm. 33

45 Absorption spectrum of Vitamin C in PBS nm 1.2 Absorbance(OD) Wavelength (nm) Figure.4.9. Absorption spectrum of unoxidized Vitamin C in PBS. In order to quantify how the Vitamin C concentration is related to its light absorption properties, Vitamin C solutions were prepared at concentrations ranging from uM in Ca +2 free and Mg +2 free Dulbecco s PBS. For all of the above measurements the samples were placed in a 4 ml quartz cuvette. Figure.4.10 shows the absorbance of vitamin C as a function of concentration. Figure Unoxidized Vitamin C absorbance in PBS as a function of concentration. From Figure.4.10 we could see the linear relationship between the Vitamin C concentration and absorbance collected at 266 nm. Therefore, the reduction of unoxidized Vitamin C concentration due to ROS generation and interaction can be directly evaluated by taking 34

46 ratios: C = {A/A o }* C o, where A o and C o are initial absorbance and concentration of Vitamin C, respectively, at time t = ROS Generation from Photofrin II Activated by 405 nm and 633 nm Lasers Figure.4.1.c showed the Photofrin II excitation spectrum and its Soret band that peaks at 397 nm. To see how ROS generation from Photofrin II excited at 400 nm differs from 633 nm excitation, which is currently utilized clinically, ROS generation from Photofrin II was measured using both 405 nm and 633 nm lasers. Photofrin II (5 mg/ml) and Vitamin C (100 μm/ml) in Dulbecco s PB solution were placed in a 4 ml quartz cuvette and irradiated with 405 nm and 633 nm lasers. The spot size was 1 cm and the power was mw for both lasers. The ROS generation was quantified by measuring the change in the absorbance of Vitamin C. Figure.4.11 shows ROS production from Photofrin II using lasers operating at 405 nm and 633 nm. Fig ROS production from Photo II due to 633nm and 405 nm lasers operating at 0.587mW 35

47 As shown in Figure.4.11, by comparing values of ROS generation, we can see that Photofrin II has produced 8.5 times greater ROS (by looking rate of change, vs ) when irradiated near its main (Soret band) excitation peak at ~400 nm as compared to the same laser power at the clinical wavelength of 633 nm ROS generation from X-ray down-converters Figure.4.12 is the experimental setup for the measurement of ROS generation from the X- ray down convertors. Photofrin II (20g/ml) and Vitamin C (100 μm/ml) in Dulbecco s PB solution were placed into a 15 ml polystyrene test tube with and without the particles and irradiated with diagnostic 120 kev X-rays operating with a constant X-ray tube current of 20 ma. The spot size was 10 cm and the sample was 1 meter away from the source and 0.5 meters away from the collimator. The ROS generation was quantified by measuring the change in the absorbance of Vitamin C using Shimadzu, Inc. (UV-3101PC) scanning absorption spectrophotometer. Figure.4.13 shows ROS production from Photofrin II with particles and without X-rays, with X-rays and without particles, and with X-rays and particles both present. 36

48 Xraygenerator Beamspotsize Testtube Spectrometer Signal Acquisition Collimator Figure Experimental set up for measuring ROS generation from X-ray down convertors. [Photo II]=10ug/ml,[Vit C]=100uM, [GdO2S2:Tb]=10mg/ml Change in Vitamin C concentration(um) dark+vitamin C dark+photo II+particles+Vit C X-rays+PhotoII+Vit C X-rays+PhotoII+particles+Vit C Exposure time (minutes) Figure Comparison of ROS production from Photo II between activation through X-ray induced Luminescence and X-rays alone. We can see that there is greater ROS generation when X-ray and particles are both present. I believe the ROS generation from X-rays alone is due to the formation of H 2 O 2 through the Type 1 mechanistic pathway. 37

49 In order to see how comparable our ROS generation from X-ray DC particles to the clinically used 633 nm laser, ROS generation from Photofrin II was measured using a 633 nm laser. Photofrin II (10 g/ml) and Vitamin C (100 μm/ml) in Dulbecco s PB solution were placed in a 4 ml quartz cuvette and irradiated with 633 nm laser. The beam spot size was 1 cm and the laser power was set to 9 mw. The ROS generation was quantified by measuring the change in the absorbance of Vitamin C. Figure.4.14 shows ROS production from Photofrin II using 633 nm laser operating at 9 mw Comparision of ROS generatiofrom Re-He lase+photoii to X-ray+Photo II + particles Dark He-Ne laser+photoii X-rays+PhotoII+particles+Vit C ROS (in um) Exposure time (minutes) Figure Comparison of ROS production from Photo II between activation by 9mW/cm2 He-Ne laser and through the X-ray induced luminescence. By looking at Figure.4.14, we could see that the rate of ROS generation through the X-ray induced luminescence is comparable to that generated by a 9 mw/cm 2 He-Ne laser. 38

50 4.5. ROS Generation from IR Up-convertors Figure.4.15 is the experimental setup for the measurement of ROS generation from the IR up-convertors. The Photofrin II (10 mg/ml) and Vitamin C (100 μm/ml) in Dulbecco s PB solution were placed in 4 ml quartz cuvette and irradiated with a 980 nm laser, with the particles (10 mg/ml) and without the particles. The laser spot size was 1 cm and power measured was 830 mw. Laser Diode Q Spectrometer Signal acquisition Figure Experimental set up for measuring ROS generation from IR up- convertors. Q: quartz cuvette. Figure.4.16 shows ROS production from Photofrin II when there are particles with no laser exposure and particles with the 980 nm laser exposure. Change in Vit C concentration(um ) [Photo II]=10ug/ml,[Vit C]=100uM,[NaYF4:Yb,Tm]]=10mg/ml dark average exp avarage Exposure time(minutes) Figure ROS generation from NaYF 4 :Yb/Tm particles. N=2 39

51 As shown in Figure.4.16, a rapid increase in the rate of ROS generation was seen in the first five minutes of IR exposure which then leveled off. The ROS generation was 32 times greater than that of ROS generation from GdO2S2:Tb particles. The significant difference in ROS generation from Infrared UC particles and X-ray DC particles is important because developing the most efficient technique for ROS generation depends on accurate selection of the most appropriate particles for Photofrin II activation. In this chapter, I attempted to measure ROS generation from X-ray DC particles, and infrared UC particles in cellular like medium. The ROS generation was quantified by measuring the change in the absorbance of Vitamin C. In order to understand and illustrate how the UC and DC particles physical properties are related to the ROS generation from Photofrin II, I also measured the spectral characteristics of both Gd 2 O 2 S:Tb and NaYF 4 :Yb/Tm particles and of Photofrin II. The summary of the results include: Photofrin II has a main excitation peak (i.e., the Soret band) near 400 nm (Figure 4.1.(c))and it produced 8.5 times greater ROS when irradiated near its main (Soret band) excitation peak at ~400 nm as compared to the same laser power at the clinical wavelength of 633 nm; There was greater ROS generation when X-rays and particles are both present for Photofrin II activation and the rate of ROS generation through the X-ray induced luminescence was found comparable to that generated by a 9 mw/cm 2 He-Ne laser; The ROS generation from Infrared NaYF 4 :Yb,Tm up-converting particles were found to be 32 times greater than that of ROS generation from Gd 2 O 2 S:Tb downconverting particles. 40

52 Chapter 5: Safety Evaluation of Rare-earth Based Materials and Therapeutic Efficacy on Selective Cancer Cell Lines As discussed in Chapter 4, significant ROS generation results were recorded (through the vitamin C assay) when the DC (Gd 2 O 2 S: Tb) and UC (NaYF 4 : Yb/Tm) particles were irradiated in DPBS in presence of Photofrin II (20μg/ml). The results showed that both submicron- to micron-sized DC and UC particles have great potential to activate Photofrin II and to generate substantial levels of ROS. As the next step in this investigation, I investigated the therapeutic efficacy of these particles in activating Photofrin II on in vitro human brain cancer cells. In addition, the possible cellular toxicity of the DC and UC particles was also investigated. Section 5.1 describes techniques used for cell preparation, cell line maintenance, and cellular metabolic activity measurements; Section 5.2 describes the results of the therapeutic efficacy of DC particles on the human glioblastoma cancer cell lines and their cell toxicity evaluation; And section 5.3 describes the results of the therapeutic efficacy of the UC particles on the human glioblastoma cancer cell lines and their cell toxicity evaluation Cell Maintenance, Cellular Metabolic Activity Measurement Techniques Cell line Maintenance: Human malignant (brain cancer) glioblastoma cells were purchased from the American Type Culture Collection (ATCC, Manassas, VA) and grown and maintained in T-75 flasks under incubation conditions of 5% CO 2 at 37 o C. The 41

53 adherent cells were maintained in ATCC formulated DMEM / F12 growth medium with 10% of fetal bovine serum and 50 units/ml of penicillin and streptomycin antibiotics. Cell preparation: When the actively dividing glioblastoma cells reached 50 60% confluence within the T-75 flasks, the cells were trypsinized and brought into suspension. The cells were then spun down and the (trypsin) supernate was discarded. The cells were re-suspended in their respective fresh growth media at an initial working concentration of 10 K/ml. The cell suspension was then transferred into single wells of 96-well plates with a transfer volume of 0.1 ml (or 1,000 cells seeded per selected well). The cells were seeded into every other well in order to minimize possible overlap in the X-ray or laser light exposure. The cells within the 96 well plates were returned back into the incubator for approximately 40 hours before the X-ray or laser light exposure. Measuring cellular metabolic activity: The metabolic response of glioblastoma cells to X-ray/infrared laser exposure was assessed with a non-radioactive colorimetric cell metabolic assay (Tetrazolium compound (MTS), Promega, Madison, WI) in duplicate (control and exposed), three days after the X-ray/Infrared Laser exposure treatments. On the day of measurement, the 96-well plates were removed from the incubator and 20 μl of the MTS solution were added to each cell containing well. The plates were then returned to the incubator for a two hour incubation period. Functionally, the MTS readily permeates through the cell membrane and is metabolized and converted into formazan by living cells (Figure.5.1). Conversion into formazan induces a maximum change in absorption at 490 nm wavelength. 42

54 Two hours after the addition of MTS, absorption measurements were made at 490 nm with a 96 well plate reader. The average absorbance value at 490 nm of the treated cell s metabolic activity was computed with standard deviations, and X-ray treated, or infrared treated cell s metabolic activity were computed and normalized relative to the sham exposed (control) metabolic activity. Figure.5.1. Structures of MTS tetrazolium and its formazan product, from: Therapeutic Efficacy and Cell Toxicity Results of X-ray DC Particles on Selective Cancer Cell Lines Therapeutic efficacy: Figure.5.2 is experimental setup for the measurement of cellular metabolic activity in response to X-ray exposure. Equal numbers of human glioblastoma cells (10 3 / well) were seeded in six central wells of 96 well plates for X-ray exposure. The seeded glioblastoma cells were incubated for 12 hours overnight prior to Photofrin II incubation. After an additional 24 hours of Photo II incubation, the 96 well plates selected for X-ray treatments were incubated with Gd 2 O 2 S: Tb particles (5mg/ml) for 4 hours and 43

55 thereafter X-ray exposed. All X-ray exposures were done in the dark at room temperature. Cells were exposed to a 120 kev X-ray beam for 15 minutes. Beam spot size was 10 cm. The sample was 1 meter away from the source and 0.5 meters away from the collimator. The irradiated cells with their sham exposed counterparts were then returned to the incubator (and incubated in the dark) for an additional 48 hours and assayed for cell viability using the MTS assay. Results of Glioblastoma cell viability due to the 15 minute diagnostic X-ray exposure are shown in Figure.5.3. Xraygenerator Beamspotsize i i 48hours incubation Collimator 96wellplates Cellular metabolic activity analysis Signal Acquisition Figure.5.2. X-ray exposure set up and measurement of cell viability 48 hours post exposure. 44

56 % of cell viability relative to control Normalized Glioblastoma Cell Viability After 15 Min Diagnostic X-ray Exposure (120kVp, 20mAs), [Photofrin II]=20ug/ml, [Gd2O2S:Tb]=5mg/ml, MTS Incubation Time 2 Hrs cells cells+particles cells+particles+photo2 treatment conditions cells+photo2 dark X-ray exposure Figure.5.3. Normalized Human Glioblastoma cellular metabolic activity through MTS measurements taken 48 Hrs after a 15 Min diagnostic X-ray exposure (120kVp, 20mAs) [Photofrin II] =20g/ml, [Gd 2 O 2 S: Tb] =5mg/ml MTS incubation time 2 Hrs. N = 3, Avg. + SD, from [50]. As shown Figure.5.3, approximately 20% suppression in the cellular metabolic activity was realized from the X-ray alone and X-rays with Photo II treatment conditions. Interestingly, the presence of Gd 2 O 2 S:Tb particles without the Photo II appears to confer protection against the ionizing radiation as no reduction in the cellular metabolic activity was observed.this conferred protection is mechanistically conceivable due to the fact that the density of Gd 2 O 2 S:Tb is 7.44 fold greater than water [31], and the 20 μm size particles sink down on top of the glioblastoma cells, surround the periphery of cells, and fill the empty spaces between cells, thus forming a shield barrier surrounding the cells. The Gd 2 O 2 S: Tb attenuation coefficient is approximately 70 fold greater than the brain tissue 45

57 for a 120 kev photons (taking the density of glioblastoma cells ~ 1g/cm 3 and density of Gd 2 O 2 S:Tb ~ 7g/cm 3 ). Severe suppression (> 90% relative to controls) in the metabolic activity of human glioblastoma cells due to the presence of clinically relevant concentration of ([20 μg/ml]) Photo II, with Gd 2 O 2 S:Tb particles ([5mg/ml]), and (120 kev) diagnostic X-ray exposure was observed. Potential cell toxicity determination of Gd 2 O 2 S: Tb particles: The human glioblastoma cell suspension was transferred into single wells of 96-well plates with a transfer volume of 0.1 ml or 1000 cells/well. Cell toxicity of Gd 2 O 2 S: Tb particles on these glioblastoma cell lines was assessed 48 hours after the particle treatment (5 mg/ml concentration, 20 μm in size) through the MTS assay. On the day of measurement, the 96-well plates were removed from the incubator and 20 micro-liters of the MTS solution was added to each cell containing well. The plates were then returned to the incubator for a two hour incubation period. The particles absorbance values at 490 nm were subtracted from the average absorbance value at 490 nm of the particle treated cell s metabolic activity. Figure.5.4 shows the cellular effects of Gd 2 O 2 S:Tb particles by themselves on the human glioblastoma cell lines. 46

58 % of cellular metabolic activity relative to control Control Control + Particles Figure.5.4. Assessment on the potential cellular influence of 5 mg/ml Gd 2 O 2 S: Tb particles on human glioblastoma. Human glioblastoma cell were co-incubated with 5mg/ml of Gd2O2S: Tb for 48 Hrs and their cellular metabolic activity was determined through the MTS assay. N = 3, Avg. + SD, from [50]. We could see from Figure.5.4 that there is no remarkable change in the cellular metabolic activity when human glioblastoma cell lines were treated with Gd 2 O 2 S: Tb particles relative to control Therapeutic Efficacy and Cell Toxicity Results of Infrared UC Particles on Selective Cancer Cell Lines Therapeutic efficacy: Figure.5.5 is the experimental setup to measure cellular metabolic activity in response to infrared laser exposure. Equal numbers of human glioblastoma cells (10 3 / well) were seeded in every other wells of 96 well plates for infrared laser exposure. The seeded glioblastoma cells were incubated for 12 hours overnight prior to Photo II incubation. After an additional 24 hours of Photo II incubation, the 96 well plates selected for laser treatments were incubated with NaYF4: Yb/Tm and thereafter laser exposed for 5 47

59 minutes. All laser exposures were done in the dark at room temperature. The beam direction was perpendicular to the 96 well plates and beam spot size was 0.5 cm. The irradiated cells with their sham exposed (control) counterparts were then returned to the incubator (and incubated in the dark) for an additional 48 hours and assayed for cell viability using the MTS assay. Figure.5.5. Infrared laser exposure set up and measurement of cell viability 48 hours post exposure. Figure.5.6 shows the results of cellular metabolic activity measurements of the UC particles at the particle concentration of 5mg/ml. 48

60 %of cellular metabilic activity relative to control dark Laser exp 0 cells cells+uc particles cells+uc particles+photo II Treatment conditions Figure.5.6. Normalized Human Glioblastoma cellular metabolic activity through MTS measurements taken 48 Hrs after 5 Min of 980nm Laser exposure [Photofrin II] =20g/ml, [NaYF4: Yb/Tm] =5mg/ml MTS incubation time 2 Hrs. N = 3, Avg. + SD. We can see from Figure.5.6 that complete shutdown of cellular metabolic activity resulted including the background (dark condition) at all the particle treated conditions at a concentration of 5 mg/ml. Then particle concentration was reduced into 0.5 mg/ml. Figure.5.7 shows the results of the human glioblastoma cell viability due to the combined treatment of Photo II with NaYF4: Yb/Tm particles (0.5mg/ml) with 5 minutes of 980 nm laser exposure. 49

61 % of cell viability relative to control MTS results of 980 nm laser exposure Exposure time:5minutes, [Photofrin II]=20ug/ml, [NaYF4:Ym,Tm]=0.5mg/ml cells cells+particles cells+particles+photo2 Treatment conditions cells+photo2 dark laser exp Figure.5.7. Normalized Human Glioblastoma cellular metabolic activity through MTS measurements taken 48 Hrs after a 5 Min of laser exposure (980nm, 1982mW/cm^2) [Photofrin II] =20μg/ml, [NaYF4: Yb/Tm] =0.5mg/ml, MTS incubation time 2 Hrs. N = 3, Avg. + SD. As shown in Figure.5.7, while 50% reduction in human glioblastoma cell viability in all particle only treated conditions was observed, the dramatic reduction (>90%) in all laser exposed conditions was observed. We also investigated amount of optimum laser intensity and exposure times in order to ensure that they don t contribute to the cell metabolic activity measurement. Equal numbers of human glioblastoma cells (10 3 / well) were seeded in 96 well plates for infrared laser exposure. After 48 hours of incubation, the 96 well plates selected for laser treatments were laser exposed for 60,135, and 300 seconds at different laser intensities. All laser exposures 50

62 were done in the dark at room temperature. The experimental set up was the same as Figure.5.5. The irradiated cells were then returned to the incubator for additional 48 hours and assayed for cell viability using the MTS assay (Figure.5.8). MTS results of Glios at different Laser exposure times control 5min,237mw/cm2 5min,849mW/cm2 5min,1415mW/cm2 5min,1982mW/cm2 2min15sec,1982mW/cm2 exposure times and laser intensity 1min,1982mW/cm2 ave Figure.5.8. Human Glioblastoma cellular metabolic activity through MTS measurements taken 48 Hrs after different exposure times and laser intensity, MTS incubation time 2 Hrs. N = 3, Avg. + SD. As shown in Figure.5.8, there is similar cellular metabolic activity response relative to control condition when the cells are irradiated with 980 nm laser with the 1415 mw/cm 2 intensity with 5 minutes of exposure time and 1982 mw/cm 2 intensity with 2 minutes and 51

63 15 seconds exposure time. I choose 1982 mw/ cm 2 intensity with 2 minutes and 15 seconds exposure time as optimal laser parameters due to shorter exposure time. To ensure cell suppression is mainly from Photofrin II activation through the visible light from UC particles, next step was to evaluate cellular metabolic response to UC particles using optimal laser parameters with reduced particle concentration and reduced Photofrin II concentration. Equal numbers of human glioblastoma cells (10 3 / well) were seeded in wells of 96 well plates for Infrared laser exposure. The seeded glioblastoma cells were incubated for 12 hours overnight prior to Photo II incubation. After an additional 24 hours of Photo II incubation, the 96 well plates selected for laser treatments were incubated with NaYF4:Yb/Tm (0.2 mg/ml) and thereafter laser exposed for 2 minutes and 15 seconds with a 1982 mw/cm 2 laser intensity. All laser exposures were done in the dark at room temperature. The experimental set up is same as Figure.5.5. The irradiated cells with their sham exposed counterparts were then returned to the incubator (and incubated in the dark) for an additional 48 hours and assayed for cell viability using the MTS assay. Figure.5.9 is the result of cellular metabolic response to 1983 mw/cm 2 intensity 980 nm laser exposure for 2 minutes and 15 seconds. 52

64 % of cell viability relative to control cells only cells+particles cells +particles+photo II cells+photo II dark laser exp Treatment conditions Fgure.5.9. Normalized Human Glioblastoma cellular metabolic activity through MTS measurement taken 48 Hrs after 135 sec of laser exposure (980nm, 1982mW/cm 2 ) [Photofrin II]=15μg/ml, [NaYF4:Yb/Tm]=0.2mg/ml, MTS incubation time 2 Hrs. N = 3, Avg. + SD. We can see from Figure.5.9 that the laser exposure of UC particles (0.2mg/ml) did not contribute to the Photofrin II activation compared to its sham exposed (control) condition. We can also see that reducing the amount of Photofrin II concentration (15 μg/ml) resulted in increased cell metabolic activity. The results of therapeutic efficacy of UC particles had shown that there is severe cell suppression (>99%) when NaYF 4 :Yb/Tm particles were used at desired concentration(5 μg/ml) ; 50% cell suppression when the particle concentration was decreased to 0.5 mg/ml and that the cell suppression wasn t due to Photofrin II activation through laser induced luminescence from particles. In order to investigate the source of cell suppression, I 53

65 decided to carry extensive studies on toxicity measurements of the NaYF 4 : Yb/Tm particles. Potential cell toxicity determination of NaYF 4 : Yb/Tm particles: At first, the cell toxicity evaluation of NaYF 4 : Yb/Tm particles on the human glioblastoma cell lines was assessed using the same MTS assay technique as for the Gd2O2S: Tb particles. The cell suspension was transferred into single wells of 96-well plates with a transfer volume of 0.1 ml or 1000 cells/well. Cell toxicity of NaYF 4 : Yb/Tm particles on these glioblastoma cell lines was assessed 48 hours after the particle treatment (5 mg/ml concentration, 50 nm in size) through the MTS assay. On the day of measurement, the 96-well plates were removed from the incubator and 20 μl of the MTS solution was added to each cell containing well. The plates were then returned to the incubator for a two hour incubation period. The particles absorbance values at 490 nm are subtracted from the average absorbance value at 490 nm of the particle treated cell s metabolic activity. Potential cellular influences of 5 mg/ml of NaYF 4 : Yb/Tm on human glioblastoma cell lines had resulted in severe suppression (100%) in metabolic activity of the cells (The results are the same as Figure.5.6). Nano-enabled drugs and diagnostics present challenges for regulatory agencies such as the US Food and Drug Administration (FDA). At this present time, the FDA does not have specific guidance documents, but has recently published recommendations on the subject. The present medical regulations are expected to apply to nano-enabled drugs and diagnostics. Additional regulations are required, when considering nanoparticles 54

66 It has been also recently reported (by the FDA Nanotechnology Characterization Group) that as the nanoparticles size decreases, the particles have tendency to be more toxic (from: Agency Nanotechnology Draft Guidance CDRH Nanotech Reviewer Network (NRN) Meeting, CDRH/OSEL Presentation). In addition, since physiochemical properties of nanomaterials are different from those of their bulk counterparts, their interaction with biological systems is expected to be different. The effects may vary between different kinds of nanoparticles, depending on chemical composition, size, and shape. Contamination of nanoparticles may cause misleading results in toxicity screens (nanoformulations that are not inherently toxic may appear to be so due to contamination) and in efficacy tests for certain applications. Testing for endotoxin contamination and pyrogenicity which examines the ability of the nanoparticles to cause fever are also critical in vitro assessments before moving on to animal studies and clinical use. New studies have shown that Endotoxin contamination is a significant hurdle to the preclinical development of nanoparticles formulations. The large surface areas and high reactivity of nanoparticles along with the fact that nanoparticles are frequently synthesized on (dirty) equipment causes endotoxins contamination to be common among many nanoparticles formulations undergoing preclinical characterization. In recent studies, endotoxin contamination of gold nanoparticles was shown to be associated with undesired inflammatory reactions, while purified gold nanoparticles did not cause an inflammatory response. Endotoxin has been shown to cause tumor regression and was proposed as a drug in clinical oncology trials (later discontinued owing to severe immunotoxicity). 55

67 Since endotoxins may influence the results of toxicity and efficacy studies, it is important to identify the source of cell toxicity (endotoxins, or toxic particles) before such studies in order to avoid misinterpretation of study results. The LAL assay is an enzyme-based assay with a working time of 45 minutes. The LAL assay is intended for the quantitative measurement of endotoxins in culture medium, buffers, plasma, serum and other solutions. Bacterial endotoxin, like lipopolysaccharide (LPS), is a fever-producing by-product of gram-negative bacteria commonly known as pyrogen. The principle of the test is based on the fact that bacteria cause intravascular coagulation in the American horseshoe crab, Limulus Polyphemus. The agent responsible for the clotting phenomena resided in the crab's amoebocytes, or circulating blood cells and that pyrogen (bacterial endotoxin) triggered the turbidity and gel-forming reaction enzymatically. Thus, endotoxins cause an opacity and gelation in Limulus amebocyte lysate (LAL), which is based on an enzymatic reaction. The simplicity and economy of the LAL chromogenic endpoint assay encourages the testing of various biologicals (including sera), devices, (air) filters and tissue culture medium for the presence of harmful levels of Endotoxin [52]. Endotoxin contamination was assessed with the in vitro limulus amoebocyte lysate (LAL) assay. Samples at different concentrations and standards were incubated with LAL reagent. The absorbance at 405 nm was measured with a spectrophotometer. A standard curve was obtained by plotting the absorbance (linear) versus the corresponding concentrations of the E. coli standards (log). The endotoxin concentrations of samples, which are run concurrently with the standards, were determined from the standard curve. 56

68 In another study, the NaYF 4 : Yb/Tm particles were tested for endotoxin contamination and the particles were found to be endotoxin free. Figure.5.10 shows the standard curve for the detection of presence of Endotoxin. It shows the typical absorbance values at 405nm for different Endotoxin concentrations present. Endotoxin free water has absorbance value of Table.5.1 shows the results of absorbance values obtained for NaYF 4 : Yb/Tm particles at different concentrations and they found to have similar (smaller ) absorbance values to endotoxin free water which shows our sample is free of Endotoxin. LAL chromogenic endpoint essay standard curve Absorption at 405 nm Endotoxin (EU/m l) Figure Standard curve for LAL assay (Water has an absorption value of 0.105). 57

69 Table.5.1. Absorption values at 405 nm of NaYF4: Yb/Tm particles at different concentration using LAL assay Sample concentration Mean absorbance at 405nm Stock (1 mg/ml) X (0.1 mg/ml) X (0.05 mg/ml) X (0.025 mg/ml) The results of endotoxin study on NaYF4: Yb/Tm particles showed that there is no detectable endotoxin, so it suggests that the particles are toxic itself. Conclusion: The results on in vitro cellular studies have shown that 20 micron-sized DC particles have great potential to activate Photofrin II in deep seated targets and to generate substantial levels of ROS and no potential cell toxicity was observed. However, the UC particles were shown to be toxic to the cell lines. The cell killing through ROS generation appears not to have been due to the particles' efficiency in activating the photo-sensitizer, but rather due to toxicity of the particles. 58

70 Chapter 6 : Theoretical modeling of ROS generation PDT depends on the amount of light delivered (L), the amount of photosensitizing drug (S 0 ) in the tissue, and the amount of oxygen (O 2 ) in the tissue. Absorption of light converts S 0 into an activated drug (S*). Reaction of S* with oxygen yields oxidizing radicals (primarily singlet oxygen, as discussed in Chapter 1, the cell killing mechanism in PDT is known to be predominantly through enhanced generation of reactive oxygen species (ROS). Greater than 90% ROS generation is through the generation of excited oxygen molecule in its singlet state 1 O 2,). A fraction (f) of these radicals attacks critical sites within the cell causing an accumulated oxidative damage (A). When the accumulated damage exceeds a threshold, A > A th, then cell death occurs [53]. The aim of this chapter is to estimate the amount of excitation light deposited and fluorescent light produced by down-converting (DC)/ up-converting (UC) particles, in response to X-ray radiation dose/infrared laser irradiation, and to assess activation of the photosensitizer as well as the theoretical effectiveness of the produced singlet oxygen. As for X-ray DC particles, the amount of deposited X-ray radiation dose and generated fluorescent light in the test medium will be quantified using both analytical and statistical methods. The analytical modeling is based on the assumption that all of the particles are uniformly distributed, and they all receive the same X-ray energy due to the high penetration depth of X-rays at 120 kev. To ensure the results of analytical modeling were not contingent upon these specific assumptions, I created a statistical model in which the photon direction and photon absorption of the sample are determined randomly. 59

71 As for Infrared UC particles, the analytical modeling of X-ray DC particles cannot be used due to existence of significant amount of absorption and scattering of the Infrared light by the all the sample components. However, I used the same statistical modeling as X-ray DC particles for quantifying amount of absorbed infrared light and generated fluorescent light. Section 6.1 provides information on the physical properties of sample components; section 6.2 describes the steps for analytical modeling of the fluorescence light generation from X ray DC particles; section 6.3 describes steps and results of statistical modeling of fluorescence light generation from UC and DC particles respectively; section 6.4 describes theoretical results of amount of singlet oxygen generated X-ray absorption coefficients of the test medium components Prior to quantifying the amount of fluorescent light generated, the physical properties of the materials present in the test sample need to be known. Our test media is composed of water, polystyrene test tube, Gd 2 O 2 S: Tb particles and air. This section provides information on the physical properties of the materials that are present in the test sample for X-ray DC particles. Figure.6.1 and Table.6.1 show X-ray mass absorption and attenuation coefficients and some physical properties for water, dry air, Gd 2 O 2 S, and polystyrene at 120 kev X-ray exposure [5]. From Figure.6.1 and Table.6.1, we can note that Gd 2 O 2 S has 137 times stronger absorbance compare to water, and 341 times compared to polystyrene. 60

72 The values given on Table.6.1 are used in theoretical modeling. X-ray attenuation coefficients from NIST Figure.6.1. Mass-energy absorption and attenuation coefficients at different X-ray photon energies for Gadolinium Oxysulfide, from: Table.6.1. Physical Properties of several materials at 120 kev X-ray exposure Gadolinium Air Water Polystyrene Oxysulfide Mass attenuation coefficient cm2/g cm2/g cm2/g cm2/g Mass energy cm2/g cm2/g cm2/g cm2/g absorption coefficient Density g/cm3 1 g/cm g/cm g/cm3 Attenuation coefficient cm cm cm cm-1 Absorption coefficient cm cm cm cm-1 61

73 6.2. Analytical modeling of X-ray absorbed dose and generated fluorescence light in the test medium in the presence of X-ray down convertors Quantifying fluorescence intensity as a function of X-ray absorbed dose: The analytical modeling is based on the assumption that all of the DC particles are uniformly distributed, and they all receive same X-ray energy due to the high penetration depth of X-rays at 120 kev[34]. The amount of X-ray dose deposited and the generated amount of fluorescent light will be found as a function of X-ray absorption efficiency, intrinsic conversion efficiency, and molecular weight of the DC particles and the X-ray exposure rate. Absorbed dose, also known as total ionizing dose (TID), is a measure of the energy deposited in a medium by ionizing radiation. It is equal to the energy deposited per unit mass of medium. The SI unit for absorbed dose is Gray (Gy) and is defined as: 1Gy=1J/kg Another unit for absorbed dose is rad, which represents the absorption of 100 ergs of energy per gram of absorbing material. 2 1rad=100ergs/g= 10 J/kg 1Gy=100 rad In the presence of full charged particle equilibrium, the absorbed dose to air is given by [54]: 62

74 D air ( J / kg) X ( R).2.58`10 4 C / kg ( ).33.97( J / C) R 2 J / kg ( ). X ( R) R (6.1) Where the X(R) is the exposure in roentgens. The SI unit for exposure is C/kg. (1R= C / kg ). Since 1 rad= 10 2 J / kg : (rad) D air rad (0.876 ).X(R) (6.2) R We can see from Equation (6.2) that roentgen to-rad conversion factor for air, under the condition of electronic equilibrium is In the presence of full charged particle equilibrium, the absorbed dose (D) to a medium can be calculated from the energy flux and the weighted mean mass energy absorption coefficient, en [13]: D= ( en ) (6.3) The dose to the air is related to the dose to the medium by the following relationship [13]: D D med air ( en ( en ) ) med air. med air ( en ( en ) ) med air. A; (6.4) Where air is the energy fluence at point in air and med is the energy fluence at the same point when a material other than air (medium) is interposed in the beam. A is the 63

75 transmission factor that equals the ratio med air at the point of interest and it is close to 0.99 for soft tissue and it will approaches to 1 when the beam energy decreases to orthovoltage range (100 to 350 kev supplied by x-ray generators used for radiation therapy). From equations (6.3) and (6.4) we can obtain the relationship between exposure to air and absorbed dose to a medium. D med ( en ( en ) ) med air. D air rad ( en [(0.876 ) R ( en ) ) med air ]. X; (6.5) The term in brackets is represented by the symbol f med and is called the roentgen-to-radian conversion factor. We can see in Equation (6.5) that this factor depends on the mass energy absorption coefficient of the medium relative to the air. Thus, the f factor is a function of the medium composition as well as the photon energy. So the equation becomes [54]: D med = f med.x (6.6) The total absorbed dose for a given mass m will be: D med = f med.x. m (6.7) From the NIST website, for X-ray energy of 120KeV, ( en ) air = cm 2 /g, ( en ) Gd = cm 2 /g, the f factor is

76 Substituting the values I have used for my experiment to the equation (6.7), ( f med =42.78 rad/r= J/ (kg R), m=60 mg=0.06 kg, X =192 mr/s =0.192 R/s, t=15 min=900 sec), we could calculate expected X-ray dose absorbed by our sample to be: D med = f med.x. m= f med m. X. t = J The rate of X-ray energy that is converted to fluorescence light by particles then could be written as: de dt ddmed c.. m c. f medm. X (6.8) dt Where the c is the Intrinsic Conversion Efficiency of the Gd 2 O 2 S: Tb particles. Intrinsic Conversion Efficiency ( c ) is defined as the fraction of absorbed X-ray energy into light within the mass of the particles. The Intrinsic Conversion Efficiency, ICE ( c ) of a Gd2 O 2 S: Tb crystal in absorbing and converting one X-ray energy photon into numerous visible light photons is reported in the literature to be dependent on the percent of Tb doping, and the size of the crystals, it typically ranges within 15 20% [34-37]. For example, we could calculate for a Gd 2 O 2 S: Tb crystal having an ICE value of 15%, which corresponds to 20 μm size, one absorbed 120 kev X-ray photon will yield 7895 green wavelength photons (544 nm wavelength), with each photon having an energy value of 2.28 ev(equations 6.9 and 6.10). This Quantum Yield value is used when excitation photons are absorbed and fluorescence photons are generated. 65

77 E N g g h h*544nm 2.28eV.(6.9) E E x g 120keV * c *15% 7895.(6.10) 2.28eV Where Eg and are the energy and wavelength of the green wavelength photon emitted by the Gd 2 O 2 S: Tb particles, h is the Planck s constant, Ex is the energy of the X-ray photon. N g is the number of photons emitted at 544 nm wavelength. Substituting experimental values to the equation (6.8) the rate of X-ray energy that is absorbed and converted to fluorescent light found to be mj/s. Where m 0 is the molecular weight of the particle, and X is the x-ray exposure rate. Total amount of X-ray energy that is absorbed and converted to fluorescent light by a given mass m will be: E f. X m = f med mx t (6.11) c med. c. Substituting experimental values to the equation (6.11), I found total amount of fluorescence light generated to be J. This energy level would provide ROS generation ( M) on the level that was measured experimentally ( M) Statistical Modeling: Quantifying fluorescent light fluence distribution using Monte Carlo Modeling. The stochastic numerical Monte Carlo (MC) model provides a basis for simulating photon propagation in a homogeneous medium with random scatterers and absorbers[55]. This 66

78 modeling is used for modeling of X-ray and infrared laser energy deposited in the test medium. It is also used for quantifying the amount of fluorescence light fluence rate generated from X-ray DC particles and infrared UC particles. Fluorescence will depend on: (1)the fluence rate distribution of the excitation light, (2) the product of the absorption coefficient and quantum yield of fluorophores, and (3) the attenuation of the fluorescence light by absorption and scattering in sample[56]. My modeling includes the following steps: 1) Generating random numbers. I will use random number generator function to create random numbers that are uniformly distributed between 0 and 1. Random numbers are used to determine original photon position, photon step size, and probability of photon scattering, absorption, transmission, and reflection. 2) Computing light distribution produced by a finite radius, collimated excitation beam. Photons will be launched uniformly orthogonal to the sample surface from the (Xray/Infrared) source in the x-y-plane within the beam radius R, as shown figure 6.2. The radial magnitude r, the angle is chosen such that (r, ) define the launch point: r R random (6.12) =2(random) (6.13) 67

79 Figure.6.2. Excitation beam profile. Beam radius:r, radial magnitude:r, angle:, X and Z are thickness of the sample in x and z directions, n 1 and n 2 are the refractive indexes of the air and the sample. Positions x and y are chosen based on r and : x=rcos() (6.14) y=rsin() 3) Moving the photon at the air sample interface: Photons will be either transmitted through the sample or reflected back from the sample at the air-tissue interface. Since the photons are injected orthogonally, the specular reflectance is specified by: 68

80 R sp ( n n ) (6.15) ( n1 n2 ) Where n 1 and n 2 are the refractive indexes of the air and the tissue respectively. For my modeling, n 1 1 ; n ; If R sp >random, the photon is reflected back. If R sp <random, the photon is transmitted. 4) Determining step size of the photons. The movement of each photon is variable and distance corresponds to a photon travels from a scattering event to the next scattering or absorbing event. The step size of the photon while the photon is inside the sample is given by [55]: s = -ln (random)/ ( a s ). (6.16) Where a, and s are mean absorption, and scattering coefficients of the sample at excitation wavelength. They are found by : a c1 c2 c3 cn ( a 1) ( a ) ( 3)... ( ) 2 a an (6.17) n c1 c2 c3 cn s ( s ) ( ) ( )... ( ) 1 s 2 s 3 s n n 69

81 Where c 1, c 2, c n are concentrations of the components of the sample, 1, 2, n are densities of the components, a1, a2, an and s1, s 2, s n are absorption and scattering coefficients of the components at excitation wavelength. For my modeling, water, and particles are main absorbers and scatterers of excitation light (n=2). Photon direction is set by the angle of scattering from the original direction of propagations to the new direction of propagation which will be discussed in step 6. 5) Recording photon absorption Probability of photon absorption is computed by: If a a s random, photon takes the new step If a a s random, Photon is absorbed and is terminated (update the absorption at this point and fluorescent photons are created only for absorbed photons by particles) (Figure.6.3). 70

82 Excitationphoton Fluorescentphoton Excitationphotonabsorptionand fluorescentphotonisgenerated Figure.6.3. Fluorescent photons are created at the point of photon absorption by particles 6) Photon scattering Once the photon has taken a step and moved to the new position and is not absorbed, it is ready to be scattered (Figure.6.4). The selection of the deflection angle is calculated by [55]: g [1 g ( ) 2 ] cos( ) = 2g 1 g 2g if g>0 (6.18) 2 1if g=0 The azimuthal angle,, is calculated by =2 (random) (6.19) 71

83 Figure. 6.4.Deflection of a photon by a scattering event. The angle of deflection,, azimuthal angle, (Modified from [55]) Once we calculate the deflection and azimuthal angle, the new trajectory of the photon x, y, ) is calculated from the old trajectory ( x, y, z ), the deflection angle and the ( z azimuthal angle [54]: x sin 2 1 z ( xz cos y sin ) x cos y sin 2 1 z ( cos sin ) cos y z x y (6.20) z sin cos 2 1 z cos z 72

84 If the angle is close to normal ( z ), then the following is formula is used: x sin cos y sin sin (6.21) cos / z z z 7) When photons hit the boundary When photons hit the boundary they will be either transmitted or reflected back to the tissue (Figure.6.5). The internal reflectance is calculated by Fresnel s law R sin ( ) tan ( ) 2 2 (6.22) 2 sin ( ) tan ( ) i t i t [ ] i i t i t Where i and t are the angles of incidence and transmittance respectively. If R i <random number, then the photon exits the tissue and terminated. If R i >random number, then the photon is reflected. 73

85 74 Figure.6.5.Internal reflectance and transmittance: Blue lines indicate transmittance, i and t are the angles of incidence and transmittance, green lines indicate internal reflectance, X and Z are thickness of the sample in x and z directions, because of symmetry y direction is not shown here. In both cases the actual position of escape needs to be calculated using foreshortened step size: z -boundary: z z z Z s z s (6.23) x-boundary: x x x X s x X s 2 / 2 / (6.24) y -boundary: y y y Y s y Y s 2 / 2 / (6.25) ),, ( ),, ( z y x z y x ),, ( ),, ( z y x z y x ),, ( ),, ( z y x z y x ),, ( ),, ( z y x z y x ),, ( ),2, ( z y x z Z y x ),, ( ),, ( z y x z y x ),, ( ),, ( z y x z y x X ),, ( ),, ( z y x z y x X i Z t X/2 X/2

86 The reflected photon will also have a new position and trajectory. The new position is computed as following: Outside top of the surface (z<0): substitute z with z; Outside slab at bottom of the surface (z>0): substitute z with 2Z-z; Outside slab at right side of the surface(x>x/2): substitute x with X-x; Outside slab at left side of the surface(x<-x/2): substitute x with -X-x; Outside slab at outward direction of the surface(y>y/2): substitute y with Y-y; Outside slab at inward direction of the surface(y<-y/2): substitute y with Y-y; And in call cases the corresponding trajectory is reversed. A summary of the excitation photon tracking is shown in Figure

87 Start Yes EndProgram InitializePhoton No Last UpdateReflectance MovePhotonataVariableStep No GetPhotonPosition anddirection Yes PhotoninSample? Yes Isphoton absorbed? No ChangePhoton Direction Internally Reflected? Yes No UpdateReflection Storelocationof absorptionevent CallFluorescent Function Returntomain program Figure.6.6. Excitation photon tracking flow chart. 9) Fluorescence: Fluorescent photons are created at the same location where excitation photons are absorbed by particles. At the point of creation, the photon at the fluorescence wavelength is given a direction assuming isotropic generation. Step size s is given by equation (6.16) at emission wavelength. As a fluorescence photon encounters sample, the probability of its a ( fl ) absorption is determined according to. ( ) ( ) a fl s fl 76

88 Where a ( fl ), s ( fl ) are the total absorption and scattering coefficients of the sample at the emission wavelength, they are found by (6.17) at the emission wavelength. (10)Computation of fluorescent light fluence at certain point The fluence rate n (r) per unit input for small volume V(r) is given by [55]: N a ( r) n ( r) V ( r) N ( ) (6.26) a fl Where N a (r) is the number of photons absorbed in V(r), N is the total number of photons in the simulation, a ( fl ) is the absorption coefficient of the sample at the emission wavelength. A summary of the fluorescence photon tracking is shown in Figure

89 Start at x,y,z of absorbtion Isotropic generation of fluorescence Move Photon at a Variable Step No Get Photon Position and Direction Yes Internally Reflected? No Update remitted fluorescence Photon in Sample? Yes Update Photon Weight Due to Absorption Weight Too Small? Yes Survive Roulette? No No Yes Change Photon Direction Return to main function, initialize new excitation photon Main function Figure.6.7. Flourescence photon tracking flow chart. 11) Results For X-ray DC particles, photons/s (120 kev photon energy with 192 R/s exposure rate) need to be simulated. Because of computer memory, I launched photons/s and multiplied the number of absorbed photons by 10,000. For infrared UC particles, photons/s (840 mw/cm^2 laser intensity with 1.27 ev photon energy) need to be simulated. Because of computer memory, I launched photons/s and multiplied the number of absorbed photons by generated fluorescence photons, and fluence rates are shown below The distribution of absorbed photons, 78

90 Figure.6.8. Results of X-ray photon simulation. (a) Fluence rate distribution of X-ray photons in Y/2 position. (b) Distribution of absorbed X-ray photons in Y/2 position. (c) Fluence rate distribution of generated fluorescence photons in Y/2 position. (d) Distribution of absorbed fluorescence photons in Y/2 position 79

91 Figure.6.9. Results of infrared photon simulation. (a) Fluence rate distribution of infrared photons in Y/2 position. (b) Distribution of absorbed infrared photons in Y/2 position. (c) Fluence rate distribution of generated fluorescence photons in Y/2 position. (d) Distribution of absorbed fluorescence photons in Y/2 position We could see from Figures 6.8 and 6.9 that there strongest absorption of excitation photons occurred at the air-sample boundary and is attenuated over the sample depth. For X-ray irradiation, the water and DC particles both contributed to the strong absorption. As for 80