Abstract

Because most noncancer cells are tolerant to high micromolar concentrations of genistein (GEN), inhibitory or stimulatory effects of GEN have been claimed for a wide variety of biochemical targets that lead to a plethora of potential mechanisms. However, because GEN is present in tissues in the nanomol-per-liter range, most of these mechanisms are unlikely to be relevant in vivo. To better identify proteins that are targets of GEN, we used a GEN-agarose-affinity phase. Cytosols from human breast cancer MCF-7 cells were fractionated over a Sephadex diethylaminoethyl column, and nonabsorbed proteins in the flow-through were affinity absorbed onto a 2-carboxygenistein-agarose column. After proteins were washed with 100 mmol NaCl/L to remove weakly bound proteins, affinity elution was conducted with 1 mmol 2-carboxygenistein/L. Using this method, a p38 protein was recovered from MCF-7 cells. N-terminal chemical sequencing of the first 30 residues of the protein revealed a peptide sequence similar to those that have been discovered in human tissues (a T-cell attractant protein from synovial fluid from patients with osteoarthritis and an analogous human skin fibroblast protein using a hirudin-affinity column) as well as a cotonine-binding protein from rat brain and related proteins in plants. In each case, the corresponding gene has not been found. In conclusion, although much of the human genome has been sequenced, novel proteins that are not described by genome data remain to be found. The DING protein (N-terminal amino acid sequence Asp-Ile-Asn-Gly) that binds to genistein with high affinity is one of these. Its biological role, however, remains to be defined.