PKC Substrates

Recent work on substrate preferences for PKC isoforms is also revising perceptions of the PKC family as an overly promiscuous group. Oriented peptide libraries have revealed significant preferences for PKC isoforms from different subclasses (conventional, novel or atypical), but also within the same subclass (Nishikawa et al. 1997). PKC-e and PKC-5 have consistently been shown to have distinct substrate preferences, and a recent detailed analysis of preferred as well as disfavored residues have led to the proposal of rather specific recognition motifs for PKC-5 and PKCZ (Fuji et al. 2004). The motif for the two atypical PKC (aPKC) isoforms is different enough from all others that a selective pseudosubstrate peptide can be used to selectively inhibit these isoforms (Couet et al. 1997a). Lest we become too comfortable with the idea that substrate preferences are strictly encoded in the primary sequence of PKC isoforms, there is also emerging evidence that the substrate preferences for PKC depend on many factors (Jaken and Parker 2000). For instance, the substrate specificity of PKC-5 can change as a result of tyrosine phosphorylation on its catalytic domain (Konishi et al. 2001). Tyrosine phosphorylation of PKC-5 also greatly reduces its requirements for activation by lipids leading to a "quasi-constitutively active" PKC species no longer dependent upon being near a membrane (Kikkawa et al. 2002).

Phosphorylation of intended target substrates with high fidelity is crucial for normal cell function. Therefore, multiple inter-dependent factors contribute to phosphorylation specificity in vivo, including specific lipid-protein interactions and active site recognition motifs. Some of our perception that PKCs as a family exhibit low substrate selectivity may have arisen as a result of widespread use of (1) phorbol esters, (2) short peptide substrate motifs, (3) kinase overexpression strategies and (4) kinase catalytic domains proteolytically removed from regulatory domains. Phorbol esters potently and permanently activate many PKC isoforms as well as other signaling molecules (Griner and Kazanietz 2007), possibly contributing to non-specific phosphorylation. Full length enzymes often contain conformational features that contribute to substrate specificity (Pears et al. 1991), as do the immediate lipid and protein environment. These "secondary" factors are lost when constitutively active catalytic domains are used experimentally. Overexpression of kinases is very likely to promote excessive (and possibly non-physiological) crosstalk between signaling systems. As usual, caution and multiple independent strategies are needed for a balanced interpretation of experiments of this nature.


The pioneering work of Mochly-Rosen and colleagues (1995) introduced the concept of a class of intracellular proteins whose purpose was to localize PKC near its substrates. The first protein of this class to be discovered and thoroughly characterized was RACK1 (receptor for activated C kinase 1) (Mochly-Rosen et al. 1995). It featured isoform-specific and high-affinity binding of PKC only in its activated form, with the interesting twist that RACK1 was not a substrate for the kinase activity (Mochly-Rosen et al. 1995). In vitro, purified RACK1 was shown to bind PKC-PII with a Kd of 1 nM, and only in the presence of the PKC activators Ca2+, DAG and phosphatidylserine. The binding affinity for other PKC isoforms was much lower. Many important advances building on this concept followed from the Mochly-Rosen group, including characterization of the PKC-RACK1 binding interface and identification and use of peptides to disrupt PKC-piI anchoring in cells to address biological function (Ron et al. 1995).

Cloning and sequence analysis revealed that RACK1 was a member of a WD40 gene superfamily that included the P-subunit of heterotrimeric G-proteins (Ron et al. 1994). These WD40 repeat proteins form a P-propeller structure with a variable number of propeller blades that assemble to form a circular disc. In the case of RACK1, its seven blades form a circular disc from which loops with variable size and sequence protrude above and below the plane of the disc. It is interesting to speculate that RACK1 exhibits an analogous structural organization to GPCRs, which contain seven transmembrane a-helices bundled into a barrel-like structure with loops of variable size and sequence protruding away from the membrane surface on both sides (Escriba et al. 2007).

In short order, a second protein was identified that displayed the requisite properties of a PKC-e selective RACK and was duly christened RACK2 (Csukai et al. 1997). DAG/phosphatidylserine activated PKC-e bound RACK2 with ~10 nM Kd, was not a substrate and led directly to the development of peptide reagents that could either block or promote PKC-e anchoring (Csukai et al. 1997). Many investigators have now used these peptides successfully to explore PKC-e function in complex biological preparations, including cells (Robia et al. 2005), tissues (Johnson et al. 1996) and transgenic mice (Mochly-Rosen et al. 2000). Further characterization of RACK2 revealed it to be identical to a P'-COP subunit of the coatamer assembly involved in membrane trafficking through the Golgi apparatus (Csukai et al. 1997). RACK proteins for the other remaining ten or so PKC iso-forms have not yet been identified. This may be in part because the Mochly-Rosen group has focused its attention on therapeutic applications of its powerful peptide reagents (Kikkawa et al. 2002), and in part because understanding of kinase anchoring proteins was evolving beyond the simple idea of a specific RACK for each PKC isoform.

A good example of PKC and its scaffold being part of a vast complex or supramolecular assembly was illustrated by the recent finding that RACK1 binds tightly to the ribosome (Sengupta et al. 2004). RACK1 could therefore be considered a ribosomal protein. Further evidence suggests that ribosome-bound RACK1 recruits conventional PKC isoforms for the purpose of regulating protein synthesis (Sengupta et al. 2004). RACK1 no longer seems like a static anchor, but perhaps more like an adaptor that allows certain PKC isoforms to attach to and regulate a variety of processes, including the protein synthesis machinery. In the case of the ribosome, RACK1 is but a tiny cog in the wheel of a supramolecular assembly.

0 0

Post a comment