A number of years ago, we developed an affinity resin for the capture of the chaperone protein Hsp90. The resin consisted of an Hsp90 binding ligand derived from SNX-5422, a compound currently in the clinic for cancer. The binding ligand was connected to the Sepharose resin by an PEG-6 linker. The affinity resin proved to be extremely selective for Hsp90 and resistant to non-specific binding of other proteins. This observation suggested that the ligand-linker combination might also be used in solution and that a variety of agents could be coupled to the linker without perturbing its Hsp90 binding ability. One could imagine using the ligand-linker combo to label Hsp90 with imaging agents and possibly using the combo to deliver therapeutic agents, in vivo. Our first foray into testing the combo was to label it with fluorescein using FITC to give probe HS-27. As expected, HS-27 labeled Hsp90 quite well and could be used to follow the protein when it was chromatographed. However, of greater interest was the result of applying HS-27 to cells. When HS-27 was applied to cells, not all cells took up the probe even though there was substantial Hsp90 in the all cell lines. There seemed to be a correlation of uptake with the ability of the cell line to form xenographs. Because fluorescein itself would limit the cell permeability of the probe, we reasoned that there might be an uptake mechanism based on trafficking of extracellular Hsp90. This original observation has driven an extensive research effort in our lab, aimed at elucidating what’s happening here and extending the variety of agents, both for imaging and therapy, that might be selectively delivered.