The structure and activity of granulin/epithelin peptides as agonists and antagonists of proliferation of breast cancer cells.

3186

The granulin-epithelin precursor, Progranulin (PGRN), which is also called PC-cell-derivedgrowth factor, or acrogranin, is a secreted growth factor-like protein that promotes mitosis, survival, and invasion of cancer cells, including breast, ovarian, endometrial, hepatocellular, adrenocortical, renal, and bladder carcinomas, gliomas, leiomyosarcomas and myelomas. PGRN is comprised of 7 repeats, designated GrnA to GrnG, of a 6-disulfide bridge motif, the granulin-epithelin module (GEM). Our goals are to correlate the spatial structure of GEMs in solution, determined by 2D nuclear magnetic resonance spectroscopy (2D-NMR), to their activity on the proliferation of cancer cells in order to gain insight into the structure-activity relationships that govern the action of the parental PGRN. We generated the 7 GEMs using a bacterial expression system that favors disulfide-bridge formation followed by the separation of disulfide isomers by reverse-phase high performance liquid chromatography. 2D-NMR revealed that three peptides, Grn A, F and C, form well-folded structures containing a stable stack of two β-hairpins in their N-terminal subdomains, but a more flexible C-terminal subdomain. The other peptides (Grns B, D, E, and G) are mixtures of poorly folded disulfide isomers. The activities of the Grn peptides were tested on the breast cancer cell lines MDA-MB-468 and MCF7, which were grown for up to 8 days in serum free medium with or without Grn peptides. Under these conditions MDA-MB-468 cells undergo substantial cell death, which was inhibited by intact PGRN. GrnF reproduced the pro-survival actions of PGRN on MDA-MB-468 cells, but had no effect on MCF7. In contrast, GrnA inhibited proliferation of MDA-MB-468 cells within 3 days of incubation. Two additional disulfide isomers of GrnA were tested and show significantly reduced growth inhibitory activity compared to the correctly folded GrnA, demonstrating the specificity of the disulfide-bridging for biological action. Despite adopting the appropriate Grn-structure in solution, GrnC was only weakly growth inhibitory, as were the poorly folded peptides (Grn B, D, E and G). Thus, it is essential that the GEM adopts the correct granulin conformation, including the appropriate disulfide bridging, in order to elicit a biological effect, but, as shown by the lack of activity of GrnC, adopting the granulin fold is not, by itself, sufficient for activity. Generating Grn peptides in large amounts, defining their structures by 2D-NMR and characterizing their activity on cancer cells, will provide insight into the structure-activity relationships of the GEMs and the full-length PGRN, and may provide routes to develop new growth modulating peptides for eventual therapeutic uses.