Preparation of milligram amounts of [32P]p42mapk, phosphorylaled at Tyr185 or diphosphorylated at Tyr185/Thr183, for use as specific protein phosphatase substrates is described. Tyr‐ but not Thr‐phosphorylated p42mapk, accumulates when ATP is limiting. Furthermore, Tyr185‐phosphorylated p42mapk exhibits an apparent 10‐fold decrease in apparent K m (46.6 ± 6.6 nM) for MAP kinase kinase compared to that for the dephospho form (∼476 nM). We conclude that Tyr185 precedes Tyr185 phosphorylation, and that this is prerequisite, dramatically increasing the affinity of p42mapk for MAP kinase kinase.

1. Recombinant, activated mitogen‐activated protein kinase (3.3 microM; p42mapk) phosphorylated caldesmon in phasic (rabbit portal vein) and tonic (rabbit femoral artery) smooth muscle strips permeabilized with Triton X‐100. 2. Phosphorylation of caldesmon by p42mapk neither induced contraction of relaxed smooth muscle nor affected the Ca2+ sensitivity of submaximally contracted permeabilized phasic or tonic smooth muscle.

reported the structure of a ternary complex of the catalytic subunit of CAMP-dependent protein b a s e (cyclic A kinase), MgATP and a 20-residue inhibitor peptide, at a resolution of 0.27 nm. This structure has since been refined to 0.2-nm resolution and the orientation of the nucleotide and interactions of MgATP with numerous conserved residues at the active site defined [Zheng, J., Knighton, D. R., Eyck, L. F. T., Karlsson, R., Xuong, N., Taylor, S. S. & Sowadski, J. M. (1993) Biochemistry, in the press]. These studies revealed that the adenosine portion of ATP is buried deep within the catalytic cleft, with the a, B and y phosphates protruding towards the opening of the cleft. The unique spatial positioning of MgATP within the catalytic cleft of cyclic A kinase and its interactions with conserved amino acids found in all protein kinases, led us to reconsider the use of ATP as an affinity ligand for the purification of these enzymes. In this paper, we describe a straightforward method for the synthesis of [y-32P]adenosine-5'-(y-4-aminophenyl)triphosphate for the covalent linkage of ATP to Sepharose through its y phosphate. In the presence of 20 pM ATP, adenosine-5'-(y-4-aminophenyl)triphosphate exhibited apparent Ki values of 103.6, 75.18, 176.28 and 120.00 pM against cyclic A kinase, mitogen-activated protein kinase (p42"'*pk), mitogen-activated protein kinase kinase and p60'-=, respectively. To illustrate the effectiveness of adenosine-5'-( y-4-aminopheny1)triphosphate -Sepharose as an affhity column for protein kinases, we have used the resin to purify rabbit skeletal muscle mitogen-activated protein kinase kinase over 19000-fold to homogeneity.Protein kinases can be categorized into three distinct classes, based on the amino acids they phosphorylate. These are the serinehreonine protein kinases, eg. cyclic AMP-dependent protein kinase (Walsh et al., 1968., Taylor et al., 1990 or mitogen-activated protein (MAP) kinase (Sturgill, and Wu, 1991;Boulton et al., 1990), the tyrosine protein kinases, eg. p60"-'" (Anderson et al., 1985; Martinez et al., 1987) and the recently discovered 'dual specificity' protein kinases that phosphorylate exogenous substrates on both tyrosine and serineh-eonine amino acids, eg. MAP kinase kinase (MAPKK; Ahn et al

Molecular steps in endocytosis and degradation of the c‐fms protein were analyzed by following the fate of mutated c‐fms molecules after M‐CSF binding. A mutant c‐fms protein lacking tyrosine kinase activity was rapidly internalized after M‐CSF binding but not degraded. Another mutant c‐fms molecule that lacked most of the kinase insert region was similarly internalized after M‐CSF binding and also not degraded. This indicates that the signal for internalization is separate from that directing degradation of the receptor. It has been shown previously that a c‐fms mutant in which the kinase insert domain is deleted retains tyrosine kinase activity but lacks two major sites of autophosphorylation. The degradation step therefore requires both kinase activity and the kinase insert region whereas the internalization step is independent of these factors. The major sites of tyrosine autophosphorylation within the kinase insert region were next mutated to determine whether autophosphorylation in the kinase insert region of c‐fms might be the signal that triggers degradation of internalized receptors. These mutant receptors were still rapidly degraded in response to M‐CSF. Therefore, ligand‐induced degradation of c‐fms may require tyrosine phosphorylation of a protein other than the c‐fms receptor itself and the kinase insert region may be necessary for recognition of this substrate.

No abstract