Si. Hisanaga et al., PORCINE BRAIN NEUROFILAMENT-H TAIL DOMAIN KINASE - ITS IDENTIFICATIONAS CDK5 P26 COMPLEX AND COMPARISON WITH CDC2/CYCLIN-B KINASE/, Cell motility and the cytoskeleton, 31(4), 1995, pp. 283-297
Using dephosphorylated neurofilament (NF) proteins as substrates, the
kinase with a higher activity for the dephosphorylated NF-H than the p
hosphorylated form of NF-H was searched for in the porcine brain extra
ct. Most NF-H kinase activity in the brain extract pelleted with micro
tubules. The NF-H kinase purified from a high salt extract of the micr
otubule pellets was composed of cdk5 and a 26 kDa protein, a fragment
of the 35 kDa regulatory subunit of cdk5. In contrast to the associati
on of the active kinase with microtubules, each of uncomplexed cdk5 an
d the 35 kDa regulatory subunit was differently distributed in the sup
ernatant fraction and the pellet, respectively, by ultracentrifugation
of the brain extract. Dephosphorylated forms of NF-H and NF-M became
reactive to antibodies recognizing in vivo phosphorylation sites (SMI3
1, 34, and 36, JJ31 and 51) by phosphorylation with cdk5/p26. cdk5/p26
showed similar enzymatic properties to p34(cdc2)/cyclin B kinase; the
substrate specificity and inhibition by a p34(cdc2) kinase specific i
nhibitor, butyrolactone I. However, p34(cdc2)/cyclin B kinase was dist
inguished from cdk5/p26 by its binding to p13(suc1) protein and by its
reactivity to anti-p34(cdc2) antibodies. In spite of similar enzymati
c properties of cdk5/p26 and p34(cdc2)/cyclin B kinase, cdk5/p26 did n
ot display M-phase promoting activity when assayed with a cell-free sy
stem of Xenopus egg extract. (C) 1995 Wiley-Liss, Inc.