Increased phosphorylation of myosin light chain (MLC) is necessary for the dynamic membrane blebbing that is observed at the onset of apoptosis. Here we identify ROCK I, an effector of the small GTPase Rho, as a new substrate for caspases. ROCK I is cleaved by caspase-3 at a conserved DETD1113/G sequence and its carboxy-terminal inhibitory domain is removed, resulting in deregulated and constitutive kinase activity. ROCK proteins are known to regulate MLC-phosphorylation, and apoptotic cells exhibit a gradual increase in levels of phosphorylated MLC concomitant with ROCK I cleavage. This phosphorylation, as well as membrane blebbing, is abrogated by inhibition of caspases or ROCK proteins, but both processes are independent of Rho activity. We also show that expression of active truncated ROCK I induces cell blebbing. Thus, activation of ROCK I by caspase-3 seems to be responsible for bleb formation in apoptotic cells.

The potent vasodilator action of cyclic GMP-dependent protein kinase (cGK) involves decreasing the Ca

We have observed that stimulation of human natural killer cells with dibutyryl cAMP (Bt2cAMP) reproduced the effects of ADP ribosylation of the GTP binding protein RhoA by Clostridium botulinum C3 transferase: both agents induced similar morphological changes, inhibited cell motility and blocked the cytolytic function. We demonstrate here that cAMP‐dependent protein kinase A (PKA) phosphorylates RhoA in its C‐terminal region, on serine residue 188. This phosphorylation does not affect the ability of recombinant RhoA to bind guanine nucleotides, nor does it modify its intrinsic GTPase activity. However, treatment of cells with Bt2cAMP results in the translocation of membrane‐associated RhoA towards the cytosol. Experiments using purified membrane preparations indicated that Rho‐GDP dissociation inhibitor, which can complex phosphorylated RhoA in its GTP‐bound state, was the effector of this translocation. Taken together, these data suggest that PKA phosphorylation of RhoA is a central event in mediating the cellular effects of cAMP, and support the existence of an alternative pathway for terminating RhoA signalling whereby GTP‐bound RhoA, when phosphorylated, could be separated from its putative effector(s) independently of its GTP/GDP cycling.