Info

Cytosine deaminase

Spectroscopy

Production of 5-fluorouracil (5-FU)

(66)

Figure 1 Basic Principles of Optical CCD Imaging (Fluorescence/Bioluminescence) There are fundamentally 2 different types of optically-based imaging systems: fluorescence imaging, which uses emitters such as green fluorescent protein (GFP), wavelength-shifted GFP mutants, red fluorescent protein (RFP), ''smart'' probes, and near-infrared fluorescent (NIRF) probes, and bioluminescence imaging, which utilizes systems such as Firefly luciferase/D-Luciferin or Renilla luciferase/coelenterazine. Emission of light from fluorescent markers requires external light excitation, while bioluminescent systems generate light de novo when the appropriate substrates/cofactors are made available. In both cases, light emitted from either system can be detected with a thermoelectrically cooled charge-couple device camera (CCD) since they emit light in the visible light range (400 nm to 700 nm) to near-infrared range (—800 nm). Cooled to -120 to - 150°C, these cameras can detect weakly luminescent sources within a light-tight chamber. Being exquisitely sensitive to light, these desktop camera systems allow for quantitative analysis of the data. The image shown above the Fluorescence Imaging schematic is representative of the one obtained from a glioma model which expresses RFP (image used with permission from Anticancer, Inc.). The method of imaging bioluminescent sources of living subjects with a CCD camera is relatively straightforward: the animal is anesthetized, subsequently injected with the substrate, and immediately placed in the light-tight chamber. A light photographic image of the animal is obtained, which is followed by a bioluminescence image captured by the cooled CCD camera positioned above the subject within the confines of the dark chamber. The 2 images are subsequently superimposed on one another by a computer, and relative location of luciferase activity is inferred from the composite image. An adjacent color scale confers relative concentration of luciferase activity. Sample image above the Bioluminescence Imaging schematic is a typical image obtained with this technology (image courtesy of Bhaumik and Gambhir, Ref. 15. In this specific example, the image was obtained after intravenous injection of coelenterazine into a mouse containing intraperitoneal Renilla luciferase-expressing tumor cells. Significant bioluminescence is detected from the region of the xenograft. See the color insert for the color version of this figure.

within the cell. In contrast, FDG clears from cells or tissues that lack the ability to transport or phosphorylate FDG. The positron emitting moiety of FDG, in this example, 18F, decays by emitting a positron from its nucleus. This positron eventually collides with a nearby electron, resulting in an annihilation event where two 511,000 eV photons in the form of gamma rays are emitted —180° apart. The 2 emitted photons travel extracorporeally and are detected nearly simultaneously as they interact with the PET ''camera''—a ring array of detectors (composed of scintillation crystals and photomultiplier tubes) surrounding the subject (Fig. 2). Detection of a single annihilation event results in the ''activation'' of detectors opposing one another, which is recorded as a ''coincident event'', thus defining a set of coincident lines (25). The recording of multiple detector pair combinations yields a large number of these coincident lines. Sophisticated mathematical analyses of the coincident lines, which include filtered back projection and attenuation correction, yield the location of cell populations or tissues that have accumulated FDG. Tomo-graphic images of relative probe concentration can be reconstructed in the sagittal, coronal, and transverse imaging planes. Quantitative information obtained from the images is, in turn, related to the underlying biochemical process.

Radiolabeling molecules is not just limited to 18F. A collection of positron-emitting isotopes is available for use, which includes the more commonly utilized isotopes 15O, 13N, 11C, and 18F, and the less commonly used 14O, 64Cu, 62Cu, 124I, 76Br, 82Rb, and 68Ga. Most of these isotopes are created in a cyclotron, a device used to accelerate charged particles to create the relatively short-lived positron-emitting isotopes (for example, 18F, the half-life of which is 110 minutes) (26). Automated synthesizers can then couple the isotope to a molecule of interest to produce the molecular probe (tracer). Given the relatively short half-life of positron emitters, the process of producing isotopically labeled molecules has to be performed with great efficiency and in relatively close proximity to the hospital, clinic, or animal research facility. In this regard, a modest number of PET radiopharmacies are available worldwide, producing PET tracers on a daily basis.

Clinical PET scanners have been around for several decades and, in recent years, PET cameras for small animals have been developed for the purpose of developing molecular imaging assays in small rodents prior to their application in humans. The spatial resolution of most clinical PET scanners is —(6-8)3 mm3 with the more recent scanners achieving —(3)3 mm3 capabilites. By comparison small animal PET scanners (microPET) have a resolution of ~(2)3 mm3 with newer generation scanners attaining —(1)3 mm3 (27,28). When compared to other modalities, the sensitivity of PET is relatively high—on the order of 10"11 to 10"12 moles/liter (MRI's intrinsic sensitivity is —10"4 to 10"5 M) (29). Furthermore, the location depth of the tracer of interest does not affect sensitivity. In contrast, the imaging of many optical imaging probes is significantly affected by tissue depth. Under appropriate conditions, the smallest cluster of cells that can be visualized by a clinical PET scanner is 106 to 109 in number. Thus, radiotracer imaging techniques afford the detailed loca tion(s), magnitude, and persistence of probes or tracers for in vivo use in animals and humans.

C. Basic Principles of the Gamma Camera and SPECT imaging

Imaging with a gamma camera is similar to PET, but the radiolabel emits gamma rays instead of positrons. A variety of radiolabels, each emitting at characteristic photon energies, can be attached to molecules including 111In (171, 245 keV), 125I (27-35 keV), 131I (364 keV), and 99mTc (140 keV). Once introduced into the body, detection of these radiolabeled probes is performed with a gamma camera, a scintillation detector consisting of collimator, a sodium iodide crystal, and a set of photomutliplier tubes (Fig. 3). Upon decay, these radionuclides emit a gamma ray at their characteristic energies in different directions. Some of the gamma rays will scatter or lose energy and others may never interact with the camera. Since the gamma camera is situated only on 1 side of the subject, only rays directed toward the camera will be ''captured''. Furthermore, only those gamma rays that arise parallel to the collimator will be detected since the collimator will absorb scattered gamma rays. Those rays that successfully reach the crystal will be converted into photons of light. In turn, the photomultiplier tubes convert the light into an electrical signal that is proportional to the incidental gamma ray. Gamma rays, which arrive at the detector lower than the expected characteristic energy, are thought to be the result of scattering and summarily rejected from the analysis. Since gamma cameras acquire data in a single plane, the resultant images are a 2-dimensional representation of a 3-dimensional subject (referred to as ''planar imaging'').

While more affordable and accessible than PET, the limitations of gamma camera imaging are obvious: 1) diminished sensitivity, since many decay events are either rejected or never captured, 2) decreased signal-to-noise since overlapping foci of activity are not delineated, and 3) lower spatial resolution. Alleviating some of these problems is single photon emissions computed tomography (SPECT), which acquires volumetric data by rotating the gamma camera around the subject and/or using multidetector systems (30) (Fig. 3). As with microPET, small animal SPECT devices (microSPECT) have been created to study the use of gamma-emitting radiolabeled reporter probes in animal models of cancer and gene therapy. These instruments have resolutions on the order of 1 mm3. Thus, advantages of the SPECT systems are that they generally allow for better spatial resolution. Another advantage of SPECT is that 2 radioisotopes of different energies can be imaged simultaneously, allowing for the concurrent study of 2 distinctly radiolabeled molecules (1 radiolabeled with 99mTc and another radiolabeled with 125I). With PET imaging, such simultaneous imaging is not possible since all positron emitting events are 511keV.

A disadvantage for SPECT, however, is that it is an order of magnitude lower in sensitivity than PET. To accommodate for the loss of sensitivity, more radiolabeled probe and, thus,

Figure 2 Basic Principles of Positron Emissions Tomography (PET) Imaging Biologically active molecules such as glucose, peptides, and proteins can be radiolabeled with positron-emitting radioisotopes. This radiolabeled molecule is referred to as a probe or tracer. The positron-emitting isotope decays by emitting a positron from its nucleus. This positron eventually collides with a nearby electron, resulting in an annihilation event where 2 511,000 eV photons in the form of gamma rays are emitted —180° apart. The 2 emitted photons travel extracorporeally and are detected nearly simultaneously as they interact with a ring of detectors (composed of scintillation crystals and photomultiplier tubes) surrounding the subject. Detection of a single annihilation event results in the ''activation'' of detectors opposing one another, which is recorded as a ''coincident event.'' The recording of multiple detector pair combinations yields a large number of these coincident lines. Sophisticated mathematical analyses of the coincident lines, which include filtered back projection and attenuation correction, yield the location of cell populations or tissues that contain the molecule labeled with the positron emitter. Tomographic images of relative probe concentration can be reconstructed in the conventional sagittal, coronal, and transverse imaging planes or, actually, in any arbitrary plane. The resultant image depicts the distribution and concentration of the radiolabeled tracer. Sensitivity of PET is in the range of 10 ~11 to 10 ~12 moles/liter and is independent of the location depth of the tracer of interest. It is also important to note that all positron-emitting radioisotopes produce 2 gamma rays of the same energy, so if 2 molecular probes—each with a different positron-emitting isotope—are injected simultaneously, there is no way for the PET camera to distinguish between the 2 molecular probes. Therefore, to perform studies that look at 2 or more distinct molecular events (e.g., suicide gene therapy and imaging apoptosis, cardiac gene therapy, and perfusion 13N ammonia imaging, etc.), one has to inject molecular probes separately, which allows decay of the isotope.

higher levels of radioactivity have to be injected into the subject to maximize signal to noise.

D. Basic Principles of Magnetic Resonance Imaging

With regards to the electromagnetic spectrum, MRI works with longer wavelengths (lower energies) than those used by radionuclide or optical techniques. A comprehensive review of MRI is beyond the scope of this chapter; there are several excellent reviews and books available (31,32). A very brief, simplistic description follows here: Spinning charged nuclei generate a magnetic field, and MR imaging depends on these charged atomic nuclei, which contain an odd number of protons. Those nuclei that contain an odd number of protons always have an unpaired proton, which gives the atom a net magnetic field or ''magnetic dipole moment'' (MDM). In contrast, atoms, which have an even number of protons, have a net magnetic field of zero. Thus, the nucleus of hydrogen (1H), which possesses a single unpaired proton as its nucleus and is present in abundance (in the form of H2O and -CH2-) in cells and tissues, is primarily used for MR imaging.

The majority of available MR scanners are outfitted with a superconducting magnet in the shape of a long hollow tube. The magnet creates an ''external magnetic field'' (B0) and, within the hollow center of the tube-shaped magnet, the mag-

Figure 3 Basic Principles of Gamma Camera/SPECT Imaging Imaging with a gamma camera is similar to PET, but the radiolabel emits gamma rays instead of positrons. A variety of radioisotopes, each emitting at characteristic photon energies, can be attached to a variety of molecules. Examples of isotopes include mIn (171, 245 keV), 125I (27-35 keV), 131I (364 keV), and 99mTc (140 keV). Once introduced into the body, detection of these radiolabled probes is performed with a gamma camera, a scintillation detector consisting of collimator, a sodium iodide crystal and a set of photomultiplier tubes. Upon decay, these radionuclides emit a gamma ray at their characteristic energies in different directions. Some of the gamma rays will scatter or lose energy and others may never interact with the camera. Since the gamma camera is situated only on 1 side of the subject, only rays directed toward the camera will potentially be ''captured''. Furthermore, only those gamma rays that arise parallel to collimator will be detected since scattered gamma rays will be absorbed by the collimator. Those rays that successfully reach the crystal and are stopped by it will be converted into photons of light. In turn, the photomultiplier tubes convert the light into an electrical signal that is proportional to the incidental gamma ray. Gamma rays that arrive at the detector lower than the expected characteristic energy are thought to be the result of scattering and summarily rejected from the analysis. Since gamma cameras acquire data in a single plane, the resultant images are a 2-dimensional representation of a3-dimensional subject (referred to as ''planar imaging''). Single photon emission computed tomography (SPECT) acquires volumetric data by rotating a gamma camera around the subject and/or using multidetector systems (shown above). (From Ref. 30.) See color insert for color version of this figure.

Figure 3 Basic Principles of Gamma Camera/SPECT Imaging Imaging with a gamma camera is similar to PET, but the radiolabel emits gamma rays instead of positrons. A variety of radioisotopes, each emitting at characteristic photon energies, can be attached to a variety of molecules. Examples of isotopes include mIn (171, 245 keV), 125I (27-35 keV), 131I (364 keV), and 99mTc (140 keV). Once introduced into the body, detection of these radiolabled probes is performed with a gamma camera, a scintillation detector consisting of collimator, a sodium iodide crystal and a set of photomultiplier tubes. Upon decay, these radionuclides emit a gamma ray at their characteristic energies in different directions. Some of the gamma rays will scatter or lose energy and others may never interact with the camera. Since the gamma camera is situated only on 1 side of the subject, only rays directed toward the camera will potentially be ''captured''. Furthermore, only those gamma rays that arise parallel to collimator will be detected since scattered gamma rays will be absorbed by the collimator. Those rays that successfully reach the crystal and are stopped by it will be converted into photons of light. In turn, the photomultiplier tubes convert the light into an electrical signal that is proportional to the incidental gamma ray. Gamma rays that arrive at the detector lower than the expected characteristic energy are thought to be the result of scattering and summarily rejected from the analysis. Since gamma cameras acquire data in a single plane, the resultant images are a 2-dimensional representation of a3-dimensional subject (referred to as ''planar imaging''). Single photon emission computed tomography (SPECT) acquires volumetric data by rotating a gamma camera around the subject and/or using multidetector systems (shown above). (From Ref. 30.) See color insert for color version of this figure.

netic field is nearly homogeneous and parallel to the long axis of the tube. When a subject is placed within the hollow confines of the magnet, the MDMs of the hydrogen atoms align themselves with the main magnetic field B0- much like the way iron filings behave when placed in the vicinity of a magnet.

Once equilibrium has been established in the external magnetic field, B0, a magnetic pulse (an electromagnetic wave), otherwise known as radiofrequency (RF) pulse, is introduced perpendicular to B0. It is called an RF pulse because the frequency (and energy) of the pulse is in the radiofrequency range of the electromagnetic spectrum (3-100 MHz). This causes the hydrogen nuclei to transiently orient their MDM parallel to the new magnetic field (perpendicular to B0). After the RF pulse, they realign (relax) their MDMs to the main magnetic field B0, and, in the process give off energy in the form of radiofrequency waves that can be detected by receiver coil, which typically surrounds the subject. To reiterate, both a RF pulse and magnetic field are used to perturb the underlying subject, with an RF wave being generated in the process that is used to produce an image. Also, the time it takes hydrogen nuclei to relax to equilibrium (or a fraction thereof) can be measured. The rate at which a hydrogen nucleus relaxes is dependent upon the nature of its parent molecule such as a freely mobile water proton vs. the rigidly attached proton of hydrocarbon backbone of a fatty acid. Water protons, which randomly tumble in aqueous solution, take longer to regain equilibrium with the main magnetic field, B0, than those protons associated with much larger, more fixed molecules. The measurements of relaxation rates can be converted into a value, which translates into image pixel value, with each pixel representing a small, representative, unit volume of the subject (voxel). On a certain MR imaging protocol called a ''T1-weighted'' sequence, a voxel composed mostly of fatty (hydrocarbons) protons will have a high (bright) signal since the rate of relaxation is rapid. Compare this to the voxel that contains a large number of water protons: this voxel will have a low (dark) signal on Tl-weighted MR imaging since the rate of relaxation is much longer. Each MR image is 256 x 256 or 512 x 512 pixels, each a representative slice through the subject (Fig. 4).

Certain exogenous or endogenous atoms/molecules, like Gadolinium (paramagnetic) and iron/hemosiderin (superpara-magnetic), respectively, can influence the local magnetic field by their powerful magnetic properties, significantly alter the rate of relaxation of the protons and therefore generate contrast in the image. Paramagnetic substances have unpaired electrons. They become magnetized in the presence of an external magnetic field and contribute to an increase in the effective magnetic field. Examples include the rare-earth element, gadolinium (Gd) (7 unpaired electrons), deoxyhemoglobin (4 unpaired electrons), and methemoglobin (5 unpaired electrons). Hemosiderin, an end-stage by-product of hemorrhage, has more than 10,000 unpaired electrons and, thus, belongs to superparamagnetic group of substances. The magnetic susceptibility of superparamagnetic substances is 100 to 1000 times stronger than paramagnetic substances. Both super- and paramagnetic substances help localize reporter gene expression during MRI (see also Fig. 10). As we will see in a later section, the properties of paramagnetic and superparamagnetic compounds are exploited for MR-based reporter systems.

MRI techniques offer phenomenal spatial resolution (voxel resolutions of —10 ^m3 in vitro and —50 ^m3 in small animals) but are several orders of magnitude less sensitive than optical and radionuclide-based techniques. Sensitivity of MRI is on the order of 10- 3 M while PET imaging is 10-12 M, and thus, substantially more MRI probe has to be injected into the living subject in order to provide sufficient contrast (33).

E. Computed Tomography

Computed tomography (CT) deals in the X-ray range of the electromagnetic spectrum. In CT, the subject is placed in the center of a ring of detectors. A rotating, focused X-ray source emits radiation that penetrates the subject and reaches a set of detectors on the other side of the patient. The amount of X-ray reaching the detector depends upon the amount absorbed by the patient and is inversely proportional to the density of tissues encountered as it passes through the patient. The amount of radiation reaching a detector is given a value, and, through a complex set of back calculations, tomographic images in the transverse plane can be constructed.

While this modality is not currently utilized to monitor gene expression, efforts are under way. Rather, CT's main role in human gene therapy will grow significantly as it currently serves as an anatomical adjunct to PET imaging. Both CT and PET data sets can be coregistered, and because of its superior spatial resolution, CT gives PET information more specificity. Clinical CT scanners have recently achieved spatial resolutions under 1 mm. Small animal CT scanners (mi-croCT) have attained resolutions of 50 ^m (34). Already showing its prowess in oncology, PET/CT clinical machines as well as small animal machines should help to show gene expression coupled to anatomy. More details about PET/CT can be found elsewhere (23).

Was this article helpful?

0 0

Post a comment