The human epidermal-growth-factor receptor (EGF-R) is a 170-kDa transmembrane glycoprotein that mediates the mitogenic response of cells to EGF and transforming growth factor alpha. Culture conditions have been developed for the large-scale expression of the cytoplasmic domain of the EGF-R in insect cells using a recombinant baculovirus. From 61 Sf9 cells, grown to high density using a bioreactor, 20 mg of the EGF-R kinase was purified to greater than 95% purity. Purification, which was carried out in the absence of detergents using classical purification methods, yielded an EGF-R protein that was not phosphorylated on tyrosine. This procedure has enabled us to produce high quality enzyme for both structural and biochemical studies on the EGF-R kinase. The in vitro activity of the cytoplasmic domain of the EGF-R kinase was modulated by multiple assay factors which include substrates, divalent cations and conformational modulators. Kinetic analysis in the presence of Mn2+ gave an apparent Vmax value of 20 nmol min-1 mg-1 and Km values of 4.5 microM for ATP and 1.43 mM for angiotensin II. This corresponds to a turnover number of 1.4 mol min-1 mol-1. Ammonium sulfate (1 M) resulted in an eightfold stimulation of kinase activity when assayed using angiotensin II as substrate. The specific activity of the intracellular domain of the EGF-R, when assayed at 20 degrees C in the presence of 1M ammonium sulfate, was 160 nmol min-1 mg-1. Activation of the EGF-R kinase by ammonium sulfate was found to be substrate-specific. No activation was found when assayed using polymeric substrates. Addition of Me(2+)-ATP to the purified enzyme resulted in autophosphorylation and was accompanied by retardation of SDS/PAGE migration. Kinetic constants and metal ion preferences of a number of co-polymers and peptide substrates have been compared. Dramatic differences in kinetic constants were found which were dependent on both the substrate and metal ion used. Activation of EGF-R autophosphorylation was found to be influenced by the use of charged polymers. The random polymer of Glu, Lys, Ala, Tyr (2:5:6:1), which was not phosphorylated by the EGF-R kinase, dramatically activates autophosphorylation of the EGF-R. Thus the intracellular domain of the EGF-R appears to be in a low-activity conformation which, under appropriate assay conditions, can be activated to a similar specific activity to that reported for the purified EGF-R holoenzyme.