Human Embryonic Kidney Cells (HEK-293 Cells): Characterization and Dose-Response Relationship for Modulated Release of Nerve Growth Factor for Nerve Regeneration

The development of engineered constructs to bridge nerve gaps may hold the key to improved functional outcomes in the repair of injured peripheral nerves. These constructs must be rendered bioactive by providing the growth factors required for successful peripheral nerve regeneration. Previous studies demonstrated that harvested human and rat dermal fibroblasts could be genetically engineered to release nerve growth factor (NGF) both in vitro and in vivo. The use of fibroblasts, however, has the potential to cause scarring, and the expression of NGF from those cells was transient. To overcome these potential difficulties, human embryonic kidney cells were modified for use with the ecdysone-inducible mammalian expression system. These cells (hNGF-EcR-293) have been engineered and regulated to secrete human NGF in response to the ecdysone analogue ponasterone A. HEK-293 cells were transfected with human NGF cDNA with the ecdysone-inducible mammalian expression system (Invitrogen, Carlsbad, Calif.). Stable clones were then selected. Ponasterone A, an analogue of ecdysone, was used as the inducing agent. The secretion of NGF into the medium was analyzed with two different methods. After 24 hours of exposure to the inducing agent, cell medium was transferred to PC-12 cells seeded in 12-well plates, for determination of whether the secreted NGF was bioactive. Medium from untreated or ponasterone A-treated hNGF-EcR-293 cells was deemed bioactive on the basis of its ability to induce PC-12 cell differentiation. The concentrations of secreted NGF were also quantified with an enzyme-linked immunosorbent assay, in triplicate. NGF production was measured in successive samples of the same medium during a 9-day period, with maximal release of 9.05 +/- 2.6 ng/ml at day 9. Maximal NGF production was 8.46 +/- 2.1 pg/10(3) cells at day 9. These levels were statistically significantly different from levels in noninduced samples (p