DESCRIPTION (provided by applicant): Positron emission tomography (PET) is a highly sensitive, noninvasive method for the detection of cancer. Although the technique is generally available in major oncology clinics, its usefulness is nonetheless restricted by limitations in radiolabeling methodology. PET makes predominant use of the positron-emitting fluorine isotope 18F. Currently, the only practical vehicle for clinical application of 18F in PET is 2-18F-2-deoxy-D-glucose (FDG), which is taken up by many types of tumor cells. The exclusive use of FDG however limits the types of cancers which can be imaged by PET. Certain peptides and small proteins are known to be remarkably sensitive and selective cancer targeting agents. However, the labeling of these peptides with 18F is laborious and impractical in the clinic. This proposal addresses this problem by the introduction of cage-like molecules which are able to bind fluoride by a combination of powerful ionic, hydrogen-bonding, and anion-pi type interactions while size excluding competing anions. These cages can be attached to the above-mentioned peptides to allow the immediate uptake of radioactive fluoride, resulting in a highly simplified labeling protocol which has the potential to revolutionize the application of PET to the cancer problem. The specific goals of this proposal are: 1) To prepare a series of highly selective fluoride-binding cage compounds. 2) To determine the association constants of these complexing agents with aqueous fluoride. 3) To attach the fluoride-binding cages to receptor-specific octreotide and RGD peptides. [4) To determine the blood plasma stability of the cage-peptide conjugates.] Once the basic methodology has been established, the longer-term objectives of this research program are to test the [18F cage]-peptide conjugates first in a mouse model, and then eventually in a research clinic. The fluoride-binding cages are 1,3,5-cylophanes linked by -CH2CH2NRCH2CH2- chains. Extensive computational modeling has shown that these hosts selectively form very stable complexes with fluoride in water. Two straightforward synthetic approaches are proposed starting from commercially available hydroxyethylcyanuric acid or diethanolamine. The cages can be attached to bioactive peptides using standard coupling reactions. The labeling mode may be monovalent (one peptide per cage) or multivalent (up to three peptides per cage).