Figure 1. Molecular imaging of herpes simplex virus 1 thymidine kinase (HSVTK) with 18F-FHBG and microPET. Hela tumor cells were stably transfected with HSVTK (+) or control vector (-) and tumor xenografts grown for ~1 week on the flanks of nude mice. One hour after intravenous injection of 18F-FHBG, transaxial microPET images show uptake of the tracer in the HSVTK(+) tumor.

Completion of a first draft of the human genome is expected to lead to new medical treatments, drugs, and insight into disease processes.1 In the emerging postgenomic era, wherein functionality will be added to this vast array of genetic information, opportunity exists for imaging to play a significant role in both basic and translational research within this new domain of functional genomics. Molecular imaging will likely have major impact. Molecular imaging is broadly defined as the characterization and measurement of biological processes in living animals, model systems, and humans at the cellular and molecular level using remote imaging detectors. The goal is to advance our understanding of biology and medicine through noninvasive in vivo investigation of cellular and molecular events involved in normal and pathologic processes.

Molecular imaging is focused on monitoring gene expression in vivo. The target genes can be either endogenous or exogenous. To meet the goal of monitoring endogenous genes in vivo, a strategic choice must be made regarding whether it is best to image DNA per se, messenger ribonucleic acid (mRNA transcripts), the protein product of gene expression, or functional activity of the expressed protein. The best strategy may depend on the biochemical context of the target gene under investigation and the desired endpoint of each experiment. Similarly, for monitoring exogenous gene expression in vivo such as gene therapy or reporter genes, the choice of measuring DNA, mRNA, protein, or function is fundamental to designing optimal imaging strategies and probes. Ultimately, these choices will be influenced by the characteristics of the biological pathways and their potential as imaging targets in vivo.

One consideration relates to the number of target molecules and their impact on generating sufficient signal-to-noise for imaging. For example, direct imaging of DNA would require imaging just two molecules per cell, a considerable challenge for remote imaging devices (positron emission tomography [PET], single photon emission computed tomography [SPECT], MRI, optical). Furthermore, any two DNA molecules may not be identical (heterologous polymorphisms). Nonspecific and nontarget binding of imaging probes is likely to overwhelm specific signals arising from target DNA. Similarly, mRNA is typically present at only 50 to thousands of molecules per cell, and again, direct imaging approaches requiring one-to-one correlation with the target molecules face considerable challenges. Conversely, proteins can be present at significantly higher levels, perhaps thousands to millions of copies per cell, and thus, direct imaging of proteins is achieved readily. Indeed, direct molecular imaging of subtypes of receptor proteins with radiopharmaceuticals is already a laboratory and clinical reality (eg, somatostatin type 2 receptor imaging with In111 DTPA-Octreotide or GPIIb/IIIa receptor imaging with peptide-based agents such as 99mTc-p748).2,3

Finally, imaging protein function has the potential for massive signal amplification when the target protein is, for example, an enzyme that can intensify the signal through metabolic conversion of a precursor substrate or transport of a prodrug into an intracellular compartment. Building on this concept, molecular imaging as now envisioned actually exploits creative synergies between these hierarchies of signal sources. For example, it is now feasible to exploit enzymatic amplification of protein function to amplify characteristic signals arising from expression of specific DNA and translation of mRNA. Thus, creative development, application, and validation of molecular biology probes and fusion constructs are important aspects of molecular imaging research.

Heeding the medical question

One underlying premise of molecular imaging is that this emerging field is not defined by the imaging technologies that underpin acquisition of the final image per se, but rather is driven by the underlying biological and medical questions. Thus, selection of an imaging technology will be driven primarily by the medical question and the advantages of a specific imaging modality, rather than attempting to use one technique for all investigations. In addition, most (but not necessarily all) molecular imaging approaches make use of image enhancing or contrast agents that provide molecular specificity to the imaging signal. These have inherent physi-cochemical differences producing their respective image contrast that in turn provide advantages or limitations for certain types of molecular imaging queries. In practice, choice of imaging modality and probe is usually reduced to high spatial resolution vs high sensitivity. For example, MRI contrast agents that affect water relaxivity inherently provide higher spatial resolution than PET or SPECT imaging. However, because of the indirect nature of enhancement produced by MRI contrast agents, higher concentrations of material, on the order of 10 to 100 micromolar concentrations, generally are necessary to produce sufficient image contrast. These pharmacological levels of compound result in a stringent standard for MRI contrast agents in regard to clinical risks of toxicity, cross-reactivity, and pharmacodynamic effects of the agents in vivo. Pharmacological concentrations of contrast agents also may perturb the underlying molecular signal that is being monitored by MRI. Conversely, PET and SPECT agents inherently are synthesized at sufficiently high specific activity to enable use of tracer concentrations of compound (picomolar to nanomolar) for detecting molecular signals and providing desired levels of image contrast. However, the physics of gamma ray detection result in lower spatial resolution. Thus, when high spatial resolution is desired, MRI may provide certain advantages for selected molecular imaging applications, while tracer technologies provide high sensitivity. Similar concepts and trade-offs could extend to ultrasound contrast agents, optical imaging techniques, and even x-ray contrast agents. Because the goal of molecular imaging is to interrogate specific molecular signals, the underlying biological question, rather than the technology itself, should drive the choice of imaging technique(s). Detailed below are a few examples of molecular imaging using PET and SPECT, but the reader is encouraged to explore other exciting molecular imaging applications using MRI and optical techniques.4-6

On the Horizon

Figure 1. Molecular imaging of herpes simplex virus 1 thymidine kinase (HSVTK) with 18F-FHBG and microPET. Hela tumor cells were stably transfected with HSVTK (+) or control vector (-) and tumor xenografts grown for ~1 week on the flanks of nude mice. One hour after intravenous injection of 18F-FHBG, transaxial microPET images show uptake of the tracer in the HSVTK(+) tumor.

Imaging Gene Expression with Reporter Genes. Reporter genes are commonly used in molecular biology to monitor expression and/or repression of a gene of interest. Typical reporter genes consist of a chimeric gene linking an endogenous or exogenous gene promoter to an enzyme (luciferase or b-galactosidase) or fluorophore (green fluorescent protein [GFP]). In all cases, the reporter must be introduced into the target cell or tissue, using a variety of methods including transfection of DNA, transduction with viral vectors, or incorporation into the genome of transgenic animals. The reporter gene can then be used to detect activation of an endogenous promoter of interest or exogenous gene as with viral gene therapy. Ideally, the magnitude and time course of reporter gene activity should parallel the strength and duration of expression of the target endogenous or viral gene.

Reporter genes used for studies in cultured cells and tissue specimens have limited utility for detecting and quantifying gene expression in intact animals. Hence, researchers adapted and characterized different reporter systems for use in molecular imaging. For detection of gene expression with PET, most studies have adapted either heterologous enzymes such as herpes simplex virus 1 thymidine kinase (HSVTK)7,8 or receptors such as the dopamine-2 receptor9 or somatostatin type-2 receptor.10 As compared with direct imaging of small quantities of mRNA, these systems have the potential to provide signal amplification at two biological levels and thereby increase sensitivity for detecting and localizing gene expression. First, a single mRNA transcript is translated into many copies of a receptor or enzyme. Second, each molecule of an enzyme can catalyze intracellular trapping of a large number of reporter probe molecules, thus theoretically providing greater signal compared to a receptor system, which binds a single reporter ligand at a time. However, enzyme systems as well as receptors have been used successfully for molecular imaging of gene expression.9 Another fundamental advantage of reporter gene/reporter probe systems is that once validated, the reporter gene can theoretically be cloned into an appropriate vector (virus) and any gene of interest can be interrogated with the same validated reporter probe. For PET, this provides a significant efficiency compared to the time and constraints inherent to traditional routes of synthesizing, radiolabeling, and validating a new and different radioligand for every new receptor or protein of interest.

The underlying principle driving use of HSVTK for therapeutic and imaging applications is its relaxed specificity for phosphorylating nucleoside analogs, such as ganciclovir, compared with endogenous thymidine kinases normally present in mammalian cells. Following active transport of a nucleoside analog across the cell membrane, the agent is selectively phosphorylated by viral thymidine kinase. The monophosphorylated nucleoside is trapped within cells and subsequently converted to a nucleoside triphosphate. Approximately 50% of cell-associated nucleoside is converted to the triphosphate form and incorporated into DNA within 1 hour,11 where it acts to terminate DNA synthesis. These metabolic steps result in trapping of nucleoside analogs like ganciclovir selectively within cells that express HSVTK. In the absence of the viral thymidine analogs, nucleoside analogs are not phosphorylated by native enzymes and do not accumulate within cells. In pharmacological doses, incorporation of nucleoside analogs into DNA causes cell death. However, tracer amounts of radiolabeled nucleoside analogs are trapped inside cells without causing toxicity, thus enabling repetitive imaging of HSVTK activity with SPECT or PET.

Use of radiolabeled nucleoside analogs to localize expression of HSVTK in living animals first was reported almost two decades ago by Saito et al.12,13 These authors administered the nucleoside analogue 14C-FMAU to rats infected with HSV-1 and then used selective trapping of the drug to quantify and map sites of viral infection by autoradiography of brain sections. Accumulation of FMAU was correlated highly with sites of focal infection, as determined by immunoperoxidase staining for viral antigens. The authors also proposed that appropriately labeled nucleoside analogs could be used for clinical imaging of thymidine kinase activity. Since these initial reports, several different technical innovations and improvements have been made to enable imaging of HSVTK in vivo.7 Radiolabeled nucleoside analogs appropriate for in vivo detection of HSVTK by SPECT or PET have been developed and characterized, including uracil nucleoside derivatives (such as FIAU labeled with 124I or 131I)14 and acycloguanosine derivatives (such 18F-FHBG)15 (Figure 1, page 15). To enable imaging of gene expression and other biologic processes in small animals such as mice, the development of micro-PET has been a significant advance in technology, allowing imaging of reporter gene activity at the level of an organ or tissue.16

Ready for Prime Time?

Imaging Endogenous Genes: Multidrug Resistance P-glycoprotein. Emergence of multidrug resistance (MDR) is a major obstacle to successful chemotherapy of cancer. Several of the first characterized mechanisms of MDR include transporter-mediated resistance conferred by increased expression of the transmembrane glycoprotein, P-glycoprotein (Pgp), the product of the MDR1 gene17 and a related membrane glycoprotein, the multidrug resistance-associated protein (MRP1).18 Pgp and MRP1 confer resistance to an overlapping array of structurally and functionally unrelated chemotherapeutic agents, toxic xenobiotics, and natural product drugs.19 Cells in culture exhibiting MDR generally show reduced net drug accumulation and altered intracellular drug distribution.

Figure 2. Molecular imaging of multidrug resistance (MDR1) P-glycoprotein (Pgp) transport activity in vivo. Pgp has been a pharmacological target for cancer therapy wherein the transporter is inhibited by a variety of drugs known as MDR modulators. One such modulator in clinical trials is valspodar. The figure shows an example of mapping in a living human the effect of valspodar on the pharmacokinetics of 99mTc-sestamibi (left panel), a radiopharmaceutical recognized as a transport substrate by Pgp. Two whole body posterior planar images of a patient pre- and post-treatment with valspodar are shown in the right panel. Both images were obtained with a nuclear medicine gamma camera 120 minutes after intravenous injection of 99mTc-sestamibi. The pretreatment planar image (PRE) shows significant clearance of the radiotracer from the liver and kidneys, a result of prompt Pgp-mediated clearance of the tracer from these tissues. The post-treatment image (POST) was obtained following 24 hours of treatment with valspodar. This image now depicts significantly enhanced retention of the radiopharmaceutical in the liver and kidneys as their respective Pgp transport activities were blocked. The heart, a Pgp negative tissue, serves as an internal control and shows little difference in retention of the radiopharmaceutical, a sign of molecular specificity. Imaging data such as these can be applied to assessment of Pgp transport activity in tumors and distribution of drugs in normal tissues.

MDR1 Pgp and related transporters have been targets for cancer therapy on two fronts. First, reversal of multidrug resistance in tumor cells by nontoxic agents that block the transport activity of these proteins has been an important target of pharmaceutical development.20 When coadministered with a cytotoxic agent, these antagonists, known as MDR modulators, enhance net accumulation of cytotoxic compounds within the tumor cells. Several high potency modulators discovered by targeted synthesis or combinatorial chemistry in combination with high-throughput screening are now in clinical trials.21, 22 Second, transgenic expression of the MDR1 gene has been explored for hematopoietic cell protection in the context of cancer chemotherapy,23 wherein Pgp could protect hematopoietic progenitor cells from chemotherapy-induced myelotoxicity. Hematopoietic cells transduced via retroviral-mediated transfer of the MDR1 gene have shown preferential survival after treatment of the animal with MDR drugs,23 and recent pilot clinical data support the approach.24 For proper use of MDR modulators as well as monitoring MDR gene therapy in chemotherapeutic protocols, identification of transporter-mediated resistance could guide the choice of agents and provide important prognostic information for cancer patients. Thus, non-invasive molecular imaging with a transport substrate serving as a surrogate marker of chemotherapeutic agents may identify those tumors and tissues in which transporter proteins are expressed and active.

To achieve this goal, several gamma- emitting compounds have been synthesized, validated, and characterized as transport substrates for MDR1 Pgp.21,25 One of the best characterized is hexakis(2-methoxyisobutyl isonitrile)99mTc(I) (99mTc-sestamibi), a widely available radiopharmaceutical,26 which may enable scintigraphic analysis of MDR with imaging cameras commonly available in nuclear medicine facilities. 27-31

The surrogate 99mTc-sestamibi is cationic (positively charged) and modestly hydrophobic, similar to many chemotherapeutic drugs in the MDR phenotype. In the absence of Pgp, 99mTc-sestamibi accumulates within cells driven by the negative mitochondrial and plasma membrane potentials.32 However, functional MDR1 Pgp mediates net outward transport of 99mTc-sestamibi from cells, reducing cellular accumulation of the radiopharmaceutical, just like a chemotherapeutic agent. Thus, cellular accumulation of 99mTc-sestamibi into drug-sensitive tumor cells is high and translates into a “hot spot” on scintigraphic images or a slow washout rate from a tumor focus. Conversely, expression of MDR1 Pgp transports the tracer out of cells, thereby resulting in reduced net accumulation. This will be detected either as a “cold” tumor or as a rapid washout rate from a tumor focus.

One approach to analyzing MDR in vivo and the efficacy of modulator therapy can involve a two-step nuclear medicine imaging procedure. Indeed, blockade of Pgp-mediated extrusion of 99mTc-sestamibi in tissues and tumors has been observed by treatment with Pgp modulators such as valspodar. 28, 29 For example, this two-step protocol for imaging MDR reversal in patients is under evaluation (Figure 2, page 16). After baseline imaging, administration of a potent modulating agent and reinjection of the tracer are performed. After administration of the modulator, those tissues or tumors showing higher accumulation of the tracer and/or reduced washout rates would indicate specific blockade of MDR1 Pgp.29 This enables noninvasive mapping of specific molecular events in vivo and provides a basis for further exploration of molecular imaging of Pgp in cancer and gene therapy.


Overall, the goals of molecular imaging are to translate the expanding body of knowledge obtained from molecular and genomic research into care for patients by integrating the techniques and technologies of the imaging sciences with the power of molecular biology. To achieve these goals, initiatives must 1) encourage the incorporation of molecular biology and structural biology into the design, synthesis, validation, and application of new imaging pharmaceuticals; 2) support attempts to develop labeled probes based on specific interactions with receptors, enzymes, structural proteins, and reporter genes within cells; and 3) support efforts encouraging imaging clinicians, researchers, structural biologists, molecular biologists, and chemists to undertake interdisciplinary, cooperative, and integrated projects. It is hoped that this brief review stimulates further enthusiasm for exploration of this new and exciting field.

Acknowledgements: The author wishes to thank colleagues in the Washington University Molecular Imaging Center for insightful discussions. This educational project and work from the Molecular Imaging Center were supported by US National Institutes of Health grant P20 CA86251.

David Piwnica-Worms, MD, PhD, is director, Molecular Imaging Center, Mallinckrodt Institute of Radiology, Washington University Medical School, St Louis


  1. Pennisi E. Human genome: finally, the book of life and instructions for navigating it. Science. 2000;288:2304-2307.
  2. Krenning E, Kwekkeboom D, Bakker W, et al. Somatostatin receptor scintigraphy with [In-111-DTPA-D-Phe] and [I123-Tyr]octreotide: the Rotterdam experience with more than 1000 patients. Eur J Nucl Med. 1993;20:716-731.
  3. Lister-James J, Vallabhajosula S, Moyer B, et al. Pre-clinical evaluation of technetium-99m platelet receptor-binding peptide. J Nucl Med. 1997;38:105-111.
  4. Weissleder R, Tung C, Mahmood U, Bogdanov A. In vivo imaging of tumors with protease activated near-infrared fluorescent probes. Nat Biotechnol. 1999;17:375-378.
  5. Sweeney T, Mailander V, Tucker A, et al. Visualizing the kinetics of tumor-cell clearance in living animals. Proc Natl Acad Sci USA. 1999;96:12044-12049.
  6. Louie A, Huber M, Ahrens E, et al. In vivo visualization of gene expression using magnetic resonance imaging. Nat Biotechnol. 2000;18:321-325.
  7. Tjuvajev J, Finn R, Watanabe K, et al. Noninvasive imaging of herpes virus thymidine kinase gene transfer and expression: a potential method for monitoring clinical gene therapy. Cancer Res. 1996;56:4087-4095.
  8. Luker G, Luker K, Pica C, et al. In vitro and in vivo evaluation of a dual-function green fluorescent protein-HSV1-thymidine kinase reporter gene for molecular imaging. J Nucl Med. 2001;42:76P.
  9. Herschman H, Barrio J, Satyamurthy N, et al. Progress toward in vivo imaging of reporter gene expression using positron emission tomography. American Society of Clinical Oncology Educational Book Spring. 2000:169-177.
  10. Zinn K, Buchsbaum D, Chaudhuri T, Mountz J, Grizzle W, Rogers B. Noninvasive monitoring of gene transfer using a reporter receptor imaged with a high-affinity peptide radiolabeled with Tc-99m or Re-188. J Nucl Med. 2000;41:887-895.
  11. Haubner R, Avril N, Hantzopoulos P, Gansbacher B, Schwaiger M. In vivo imaging of herpes simplex virus type 1 thymidine kinase gene expression: early kinetics of radiolabelled FIAU. Eur J Nucl Med. 2000;27:283-291.
  12. Saito Y, Price R, Rottenberg D, et al. Quantitative autoradiographic mapping of herpes simplex virus encephalitis with a radiolabeled antiviral drug. Science. 1982;217:1151-1153.
  13. Saito Y, Rubenstein R, Price R, Fox J, Watanabe K. Diagnostic imaging of herpes simplex virus encephalitis using a radiolabeled antiviral drug: autoradiographic assessment in an animal model. Ann Neurol. 1984;15:548-558.
  14. Doubrovin M, Ponomarev V, Beresten T, et al. Imaging transcriptional regulation of p53-dependent genes with positron emission tomography in vivo. Proc Natl Acad Sci USA. 2001;98:9300-9305.
  15. Alauddin M, Conti P. Synthesis and preliminary evaluation of 9-(4-[18F]-fluoro-3-hydroxymethylbutyl)guanine ([18F]FHBG): a new potential imaging agent for viral infection and gene therapy using PET. Nucl Med Biol. 1998;25:175-180.
  16. Cherry S, Shao Y, Silverman R, et al. MicroPET: a high resolution PET scanner for imaging small animals. IEEE Trans Nucl Sci. 1997;44:1161-1166.
  17. Ambudkar S, Dey S, Hrycyna C, et al. Biochemical, cellular, and pharmacological aspects of the multidrug transporter. Annu Rev Pharmacol Toxicol. 1999;31:361-398.
  18. Cole SPC, Bhardwaj G, Gerlach JH, et al. Overexpression of a transporter gene in a multidrug-resistant human lung cancer cell line. Science. 1992;258:1650-1654.
  19. Keppler D, Leier I, Jedlitschky G. Transport of glutathione conjugates and glucuronides by the multidrug resistance proteins MRP1 and MRP2. Biol Chem. 1997;378;787-791.
  20. Ford JM, Hait WN. Pharmacology of drugs that alter multidrug resistance in cancer. Pharmacol Rev. 1990;42:155-199.
  21. Sharma V, Piwnica-Worms D. Metal complexes for therapy and diagnosis of drug resistance. Chem Rev. 1999;99:2545-2560.
  22. Newman M, Rodarte J, Benbatoul K, et al. Discovery and characterization of OC144-093, a novel inhibitor of P-glycoprotein-mediated multidrug resistance. Cancer Res. 2000;60:2964-2972.
  23. Hanania E, Fu S, Roninson I, Zu Z, Gottesman M, Deisseroth A. Resistance to taxol chemotherapy produced in mouse marrow cells by safety-modified retroviruses containing a human MDR-1 transcription unit. Gene Ther. 1995;2:279-284.
  24. Moscow J, Huang H, Carter C, et al. Engraftment of MDR1 and NeoR gene-transduced hematopoietic cells after breast cancer chemotherapy. Blood. 1999;94:52-61.
  25. Piwnica-Worms D. Functional identification of multidrug resistance gene expression in vivo. American Society of Clinical Oncology Educational Book Spring. 2000:178-184.
  26. Piwnica-Worms D, Chiu M, Budding M, Kronauge J, Kramer R, Croop J. Functional imaging of multidrug-resistant P-glycoprotein with an organotechnetium complex. Cancer Res. 1993;53:977-984.
  27. Del Vecchio S, Ciarmiello A, Potena MI, et al. In vivo detection of multidrug resistance (MDR1) phenotype by technetium-99m-sestamibi scan in untreated breast cancer patients. Eur J Nucl Med. 1997;24:150-159.
  28. Chen C, Meadows B, Regis J, et al. Detection of in vivo P-glycoprotein inhibition by PSC 833 using Tc-99m-Sestamibi. Clin Cancer Res. 1997;3:545-552.
  29. Luker GD, Fracasso PM, Dobkin J, Piwnica-Worms D. Modulation of the multidrug resistance P-glycoprotein: detection with Tc-99m-Sestamibi in vivo. J Nucl Med. 1997;38:369-372.
  30. Ciarmiello A, Del Vecchio S, Silvestro P, et al. Tumor clearance of technetium-99m-Sestamibi as a predictor of response to neoadjuvant chemotherapy for locally advanced breast cancer. J Clin Oncol. 1998;16:1677-1683.
  31. Zhou J, Higashi K, Ueda Y, et al. Expression of multidrug resistance protein and messenger RNA correlate with 99mTc-MIBI imaging in patients with lung cancer. J Nucl Med. 2001;42:1476-1483.
  32. Piwnica-Worms D, Rao V, Kronauge J, Croop J. Characterization of multidrug-resistance P-glycoprotein transport function with an organotechnetium cation. Biochemistry. 1995;34:12210-12220.