What Is Molecular Imaging?

The field of molecular imaging emerged in the early 1990’s as scientists from multiple disciplines, including cell biology, biomedical engineering,  chemistry, mathematics, medicine, pharmacology and genetics began working toward the development of imaging instruments, imaging probes, assays, and quantification techniques to elucidate molecular mechanisms in biology and medicine.  Molecular imaging aims to non-invasively visualize, characterize and quantify normal and pathologic processes within the living organism at the cellular and subcellular level.  E.A. Zerhouni, MD, former director of the National Institutes of Health, has  described molecular imaging as having “…the potential to define itself as a core interdisciplinary science for extracting spatially and temporally resolved biological information at all physical scales from Angstroms to microns to centimeters in intact biological systems.” (Eugene P. Pendergrass New Horizons Lecture, Radiological Society of North America meeting, 2007)(1).  Even in its early stages of development, molecular imaging is revolutionizing our ability to see and monitor specific proteins and genes, and characterize molecular pathways within the living organism.

In contrast to traditional biomedical imaging by microscopy, in which excised tissues are typically examined to characterize histological changes and thus identify an underlying disease process, molecular imaging targets distinct molecular pathways in vivo, providing visual and quantitative information for diverse research applications. For example, investigators have applied molecular imaging to: 1) noninvasively characterize the stages and progression of a disease process and establish signature biomarkers; 2) assess the efficacy of standard or experimental treatment modalities in small-animal models of human disease; 3) characterize the trafficking of stem cells and immune cells; 4) analyze the biodistribution of drugs and the dynamics of drug/receptor interactions; 5) investigate the cellular and subcellular basis of brain disorders; 6) assess metabolic changes, particularly in the brain, heart and tumors; and 7) detect tissue hypoxia. Because biological systems are, at one level, highly complex and dynamically changing assemblies of molecular structures and interactive pathways, molecular imaging provides the tools to investigate in real time the distribution and activity of targeted biomolecules as they undergo changes in space and time.

Molecular imaging offers scientists in many different specialties significant advantages over traditional research paradigms.  First, molecular imaging procedures can be conducted in the living organism.  Whereas studies of tumor responsiveness to a new therapeutic agent, for example, would have traditionally involved a large cohort of animals from which subsets would be analyzed histologically at multiple time points, molecular imaging allows characterization of tumor development and response to a therapeutic, and even response to discontinuation of the therapeutic, within the same small set of animals imaged longitudinally at multiple time points.  This example illustrates two associated advantages of molecular imaging:  1) a study can be conducted with significantly fewer animals (thereby minimizing animal usage and reducing animal costs) than would otherwise be necessary; and 2) the statistical power is increased because each animal serves as its own control.  Other advantages include the ability of molecular imaging procedures to interrogate the whole body, in addition to focusing on specific regions, and to visualize the molecular target of interest in 3-dimensional space.  Finally, molecular imaging is becoming a key bridging technology to translate experimental preclinical findings into the clinical environment.

Imaging Instruments: Molecular imaging can be performed with a range of instruments, most of which utilize a specific region of the electromagnetic spectrum (See Table below). These imaging modalities include magnetic resonance imaging (MRI), x-ray computed tomography (CT), positron emission tomography (PET), single photon emission computed tomography (SPECT), and optical (bioluminescence and fluorescence imaging). In addition, a limited band of the sound spectrum is utilized for molecular imaging by ultrasound. Beginning in the early 1990′s, new imaging instruments were designed and fabricated with the aim of advancing the use of small research animals, particularly mice, in molecular imaging studies through the use of dedicated imaging instruments with higher sensitivity and resolution than their clinical counterparts. The continuing evolution of imaging technology has seen a rapid growth in the development and use of multimodality instruments, both in the preclinical and clinical settings. Dual modality, or hybrid, PET/CT scanners have become the preferred system for clinical nuclear medicine, and dedicated PET/CT and SPECT/CT scanners for small animal imaging are also now used routinely. PET/MRI instruments also are being developed (2). The powerful advantage of these blended instruments derives from melding the functional imaging strength of one technology, such as PET and SPECT, with the high resolution anatomical imaging afforded by MRI or CT. Finally, new instruments including combined Optical/X-Ray and other technologies (e.g., Photoacoustic imaging and Raman spectroscopy) have been recent additions for small animal imaging. It can be anticipated that the growing use of molecular imaging in biology and medicine will drive the development of new imaging instruments as well as the refinement of our current imaging technologies.


Imaging Modality Radiation spectrum Resolution Main Advantages Main Disadvantages
Positron Emission Tomography (PET) High energy gamma rays 1 – 2 mm High sensitivity; isotopes of biologically important atoms for substitution; quantitative; translational to clinic Cyclotron needed for short-lived isotopes; moderate resolution; radiation dose to animal
Single Photon Emission Computed Tomography (SPECT) Lower energy gamma rays 1 – 2 mm Can image multiple probes simultaneously; moderate-high sensitivity; adaptable to clinc Moderate resolution; radiation dose to animal; lower sensitivity than PET
X-Ray Computed Tomography (CT) X-rays 25 mm Good anatomical imaging, high resolution; imaging of bone, tumor, vascular density & permeability Limited soft-tissue contrast; limited use for true
Optical Bioluminescence Imaging Visible light 1 – 10 mm Highest sensitivity; no external light stimulation needed; low cost; relatively easy; high throughput; no radiation Low resolution; 2-D image; light attenuated at increased depth; need genetically manipulated cells or animals
Optical Fluorescence Imaging Visible or near infra-red light 1 – 10 mm High sensitivity; multiple reporter wavelengths for multiplex imaging; low cost; relatively easy, quick; no radiation Low resolution; light attenuated at increased depth; autofluorescence can increase noise
Magnetic Resonance Imaging (MRI) Radio waves 25 – 200 mm High spatial resolution; both functional and anatomical information; targeted molecular contrast agents; no radiation Low sensitivity; relatively long image acquisition and processing times; highest instrumentation cost
Ultrasound High-frequency sound 50 mm Real-time imaging; portability; high spatial resolution; low cost; molecular targeting; no radiation Limited imaging through bone or lungs

Imaging Targets: Given that the goal of molecular imaging is to visualize, characterize and quantify normal and pathologic processes at the cellular and subcellular level, the choice of a target molecule becomes a critical issue, particularly since a single scanning session is typically limited to imaging only one or two molecular targets.  The choice of target is generally based on a large body of investigational “bench” research that has identified key molecules, such as specific cell surface markers, genes or gene products (e.g., mRNA or protein), that are specifically associated with a particular physiologic or pathologic process.  In this regard, functional genomics and proteomics have developed increasingly sophisticated techniques to identify genes involved in disease processes, and specific proteins that are structurally abnormal or that are expressed at abnormally high or low levels in a disease state.  In medical research, the successful choice of a target molecule that is a key disease biomarker has the potential to lead to the development of both an imaging probe to visualize the target molecule, as well as a therapeutic agent to inhibit the disease process.

Imaging Contrast Agents: Once a target molecule has been designated, an appropriate imaging contrast agent (also termed imaging probe or imaging agent) must be selected.  The contrast agent is typically a molecule (e.g., small molecule, peptide or antibody) that can bind to the target molecule or be trapped enzymatically in a cellular compartment.  In addition the imaging agent is labeled with a moiety that renders it visible to a particular imaging modality.  Labels in common use include radionuclides (e.g., 11 C, 18 F and 64 Cu for PET;  99m Tc, 123 I and 111In for SPECT), fluorescent molecules (e.g.,  enhanced green fluorescence protein [EGFP], red fluorescence protein [RFP]) and paramagnetic ions (e.g., gadolinium).  A number of small particles that can be covalently bound to targeting molecules have been developed, including nanoparticles (e.g., quantum dots) , liposomes and microbubbles.  These particles may be intrinsically fluorescent  (e.g., quantum dots),  give strong ultrasound echoes (e.g.,  microbubbles), or can be tagged with radioactive ions, fluorescent molecules or paramagnetic ions (e.g., nanoparticles, liposomes).  Targeted molecular microCT imaging is limited by the need for impractically high concentrations of contrast agent, but iodine-based contrast agents may be used to enhance soft tissue contrast and improve anatomical imaging.  The development of new contrast agents for each imaging modality, and for multimodality imaging, is progressing rapidly as investigators seek new probes with the potential to characterize the stage of a disease and evaluate the response to therapy when used at diagnostic doses, as well as to deliver a molecular therapy when administered at higher doses.

Contrast agents can be non-specific, specific, or activatable.  Non-specific probes include gadolinium chelates, used to measure tissue perfusion and permeability, and indocyanine green, a flurorochrome agent that has been used for many years in cardiac function testing, and more recently, for optical imaging of tumor perfusion.  Specific contrast agents are aimed at protein targets that are typically cell surface targets or enzymes.  An example of the former is 11 C-raclopride, used to target dopamine D2 receptors on neurons, while enzyme targeting is exemplified by 18 F-2-fluoro-2-deoxyglucose (FDG), a substrate for intracellular hexokinase which is phosphorylated and trapped within cells, providing a strong PET signal in cells and tissues with elevated glucose metabolism typical of most tumors, as well as heart and brain.   FDG is commonly used as a general PET imaging agent to identify tumors or to monitor their development.

Activatable, or “smart” probes, are a relatively new addition to optical and MRI imaging.  These contrast agents provide the advantage of enhanced signal-to-noise ratio that derives from signal detection only at sites where there is target-contrast agent interaction, complemented by virtual lack of signal from circulating and non-specifically bound contrast agent.  For example, in optical imaging a fluorophore that is linked to a quenching molecule remains virtually invisible until the linker is broken by a targeted enzyme allowing the fluorophore and quencher to move apart, thereby restoring fluorescence (3).  These optical contrast agents are primarily targeted to proteases, including cathepsins and matrix metalloproteases, although other molecular targets are also being investigated.  Recently, an activatable optical probe linked to the anti-HER-2 antibody, trastuzumab, was shown to be activated only when it had been internalized within target cancer cells, thereby diminishing the background signal from blood pool and interstitial antibody, and markedly increasing the signal-to-noise ratio (4).

Activatable probes are also being developed for MRI imaging.  For example, a novel activatable MRI contrast agent is able to image gene expression by targeting the enzyme, beta-galactosidase, encoded by the lacZ gene.  The activatable probe is composed of a paramagnetic ion, gadolinium, enclosed in a molecular ‘cage’ whose lid can only be opened by cleavage with beta-galactosidase, thus allowing the interaction of water with gadolinium and changing the T1 MRI signal (5).

As would be expected in the aftermath of decoding the human and mouse genomes, an explosion of interest in the genetic basis of disease has fostered the development of techniques to image gene expression.  In one approach, mRNA from the gene of interest is targeted with a segment of RNA that is complementary and specific to the mRNA and is, in addition, tagged with a radionuclide or fluorescent molecule.  Despite a number of limitations including short plasma life, off-target effects and limited membrane penetration, short sections of these antisense oligonucleotides, termed small interfering RNAs (siRNAs), are being investigated and developed for molecular imaging and  as possible therapeutic tools to silence specific genetic pathways.

Reporter Genes: A far more successful and widespread technique for imaging gene expression uses an introduced “reporter” gene linked to the gene of interest, both of which respond to the same promoter (6).  By imaging the signal from the reporter gene’s protein product, which is proportional to expression of both the reporter gene and the linked gene of interest, gene expression can be visualized and quantified.  Reporter genes must be genes that produce a protein not normally produced by the body in order for the image to reveal only the location and activity level of the gene of interest.

The reporter gene strategy requires that cells be genetically modified so that they contain the reporter gene.  Despite this limitation, the technique has been successfully applied to several areas of research.  For example, 1) cancer cells can be transfected with a reporter gene and the progression of the disease and response to therapy can be monitored over time; 2) a transgenic animal with a reporter gene  can be used to follow the expression of a particular gene throughout the animal’s lifetime; 3) bacteria and viruses with a reporter gene can be used to track and characterize the spread and time course of infection in a research animal; 4) in animal models, reporter genes can facilitate the development of effective gene vectors, including viruses, liposomes, ‘naked’ and ‘modified’ DNA, for gene therapy; in humans, the reporter gene technique may be used to assess the effectiveness of gene therapy by monitoring both the location of an administered gene as well as temporal changes in the magnitude of its expression.

For bioluminescence optical imaging in mammals, the reporter gene expresses a luciferase, most commonly firefly luciferase, which reacts with an administered (IP or IV) substrate, a luciferin.  The interaction of luciferase, luciferin and molecular oxygen results in the emission of photons of visible light which can be detected by a high resolution CCD camera.   The reporter gene for fluorescence imaging in mammals is commonly the gene for GFP or one of its derivatives.  When illuminated, GFP emits light of a longer wavelength which, if sufficiently intense, can be detected at the surface of the animal by a high-sensitivity CCD camera.  An advantage of fluorescence reporter gene imaging is that injection of a substrate is not necessary for visualization, a benefit when images at multiple time points over many hours are required, because bioluminescence imaging after luciferin injection is typically satisfactory for only 5-30 min after administration.  In addition, multiplexed fluorescence instruments are able to simultaneously image multiple fluorophores each with a unique molecular target.

The reporter gene approach for PET and MRI studies differs from optical imaging in producing a protein product that can activate or accumulate the administered contrast agent.  PET studies have commonly used the viral gene for an enzyme (e.g., mutant herpes simplex virus thymidine kinase (HSV1-tk)) as the reporter gene.  In this procedure, viruses engineered to contain a promoter linked to both the reporter gene and the gene of interest (e.g., a therapy gene) are administered intravenously.  After an appropriate uptake time, a nuclide-labeled contrast agent (e.g., 9-[4-[18F] fluoro-3-(hydroxymethyl)butyl]guanine (FHBG)), a substrate for HSV1-tk,  is administered which is phosphorylated and trapped in cells expressing the reporter gene and its linked gene of interest.  Both hepato-biliary and urinary excretion of FHBG occur, typically resulting in PET signal from liver, gut and urinary bladder, in addition to any sites where the reporter gene/gene of interest are being expressed.  The confounding effect of signal from sites of excretion can be reduced by imaging several hours after injection to allow time for partial elimination. It is advantageous to use an enzyme as the targeted molecule because a single copy of the enzyme can interact with multiple copies of its substrate, thereby providing natural signal amplification and improving the signal-to-noise ratio.

In addition to single gene reporter probes, dual- and triple-fusion reporter genes have been developed to take advantage of multimodal imaging and the individual strengths of each modality.  For example, a dual-fusion lentiviral vector expressing HSV1-tk and firefly luciferase was used to image transgene expression after fetal gene delivery in non-human primates (7).  A triple-fusion reporter gene expressing HSV1-tk, Renilla luciferase, and monomeric red fluorescent protein provided a probe to image tumors and metastases with PET and optical imaging (8).  MicroCT images of the same animals were overlaid with their PET images for precise anatomical localization of tumors, and CT images also provided accurate tumor volume measurements. The development of new reporter genes and multi-fusion reporters, as well as improvement of existing reporter probes, are active areas of research that will likely play an important role in the development of gene therapy in human medicine.

Clearly, molecular imaging is a rapidly growing, wide-ranging field of study that has tremendous potential to elucidate biological processes and pathways at the cellular and molecular levels, and to translate scientific discoveries in the research laboratory into the clinical setting.  The important insights gained from preclinical small animal molecular imaging, and their translation into clinical trials promises to accelerate the development of individualized therapies tailored to a patient’s genetic makeup.

General References

Cherry SR.  In vivo molecular and genomic imaging:  new challenges for imaging physics.  Phys Med Biol; 49:R13-R48, 2004.

Cherry SR. Multimodality in vivo imaging systems:  Twice the power or double the trouble?  Annu Rev Biomed Eng; 8:35-62, 2006.

Gambhir SS.  Just what is molecular imaging?  mi gateway; 1:5-7, 2007.

Jaffer FA, Libby P, Weissleder R.  Optical and multimodality molecular imaging; Insights into atherosclerosis.  Arterioscler Thromb Vasc Biol; 29:1017-24, 2009.

Koo V, Hamilton PW, Williamson K.  Non-invasive in vivo imaging in small animal research.  Cellular Oncol; 28(4):127-139, 2006.

Lyons SK.  Advances in imaging mouse tumour models in vivo.  J Pathol; 205:194-205, 2005.

Rowland DJ, Cherry SR.  Small-animal preclinical nuclear medicine instrumentation and methodology.  Semin Nucl Med; 38:209-22, 2008.

Tsui BMW, Kraitchman DL.  Recent advances in small-animal cardiovascular imaging.  J Nucl Med; 50:667-70, 2009.

Weissleder R, Pittet MJ.  Imaging in the era of molecular oncology.  Nature; 452:580-89, 2008.

Wu JC, Bengel FM, Gambhir SS.  Cardiovascular molecular imaging.  Radiology; 244:337-57, 2007.

Specific References in Text

  1. Cited in J Nucl Med; 49:25N-26N, 2008.
  2. Catana C, Procissi D, Wu Y, Judenhofer MS, Qi J, Pichler BJ, Jacobs RE, Cherry SR.  Simultaneous in vivo positron emission tomography and magnetic resonance imaging.  PNAS; 105(10):3705-10, 2008.
  3. Weissleder R, Tung CH, Mahmood U, Bogdanov A.  In vivo imaging of tumors with protease-activated near-infrared fluorescent probes.  Nat Biotechnol; 17:375-8, 1999.
  4. Ogawa M, Regino CAS, Choyke PL, Kobayashi H.  In vivo target-specific activatable near-infrared optical labeling of humanized monoclonal antibodies.  Mol Cancer Ther; 8(1):232-9), 2009.
  5. Louie AY, Huber MM, Ahrens ET, Rothbacher U, Moats R, Jacobs RE, Fraser SE, Meade TJ.  In vivo visualization of gene expression using magnetic resonance imaging.  Nat Biotechnol; 18(3):321-5, 2000.
  6. Yu Y, Annala AJ, Barrio JR, Toyokuni T, Satyamurthy N, Namavari M, Cherry SR, Phelps ME, Herschman HR, Gambhir SS.  Quantification of target gene expression by imaging reporter gene expression in living animals.  Nature Med; 8:933-7, 2000.
  7. Tarantal AF, Lee CCI, Jimenez DF, Cherry SR.  Fetal gene transfer using lentiviral vectors:  In vivo detection of gene expression by microPET and optical imaging in fetal and infant monkeys.  Hum Gene Ther ; 17:1254-61, 2006.
  8. Deroose CM, De A, Lowning AM, Chow PL, Ray P, Chatziioannou AF, Gambhir SS.  Multimodality imaging of tumor xenografts and metastases in mice with combined small-animal PET, small-animal CT, and bioluminescence imaging.  J Nucl Med; 8(2):295-303, 2007.


Comments are closed.