This study confirms that stimulation of α1A AR with PHE in rat-1 fibroblasts promotes activation of a PLD activity [10, 11, 25]. Moreover, it demonstrates that PHE selectively decreases PKCζ activity and that PLD activity is regulated by a mechanism involving PKCζ. Furthermore, the pseudosubstrate domain of PKCζ appears to play an important role in the regulation of PLD. Our previous study has ruled out the involvement of classical or novel PKC isoforms in α1A AR-stimulated PLD activation .
In rat-1 fibroblasts expressing the α1A AR subtype, PHE causes a transient Ca2+ increase, cAMP accumulation and activation of PKA, p38 mitogen-activated protein kinase, p70 S6 kinase and PLD activation [10, 11, 23, 34]. In contrast, PHE inhibits basal levels or agonist-induced activation of ERK, PI 3-kinase and Akt in these cells [11, 34]. In addition, PHE slightly decreases basal PtdInsP2 and PtdInsP3 levels . Despite these negative effects on proliferation (ERK) and survival (PI 3-kinase, Akt) pathways in rat-1 fibroblasts, α1A AR stimulation does not significantly promote apoptosis . It should be noted that in these cells, PLD activity is extremely low both in serum-deprived or serum-treated cells; PLD activation by PHE is not affected by the presence of serum and PHE decreases cell proliferation (Parmentier JH, Saeed AE and Malik KU, our unpublished observation). These observations, together with our finding that PHE-induced decrease in PKCζ activity is associated with an increase in PLD activity, support an anti-proliferative effect of α1A AR stimulation in rat-1 fibroblasts, involving PLD activation. In contrast, in VSMC expressing high levels of the PLD2 isoform [30, 32], norepinephrine-induced cell proliferation was dependent on PLD activation. In addition, norepinephrine selectively stimulates the PLD2 isoform in VSMC  through the catalytic activation of PKCζ , in contrast to the effect of PHE in rat-1 fibroblasts. The cell phenotype, VSMC vs. rat-1 fibroblasts, or the adrenergic receptor subtype, α1A AR in rat-1 vs. several α1 and α2 AR subtypes in VSMC, may be responsible for this difference.
To our knowledge, this is the first report demonstrating inhibition of PKCζ activity in response to receptor stimulation. However, treatment with PHE decreases the activity of many mitogenic indices in rat-1 fibroblasts, including DNA synthesis, ERK, phosphatidylinositol 3-kinase and Akt [11, 34]. A noteworthy fact is that there is elevated mitogenic signaling in rat-1 fibroblasts expressing α1A AR cultured in serum-free medium. The decrease in PKCζ activity elicited by PHE could result from a physical interaction between PKCζ and K10 keratin, causing sequestration of PKCζ within the cytoskeleton and preventing its intracellular translocation, thus impairing its activation, as reported by Paramio et al. , although translocation is often associated with activation for other PKC isoforms. In addition, a recent study shows that a small pool of PKCζ is constitutively active and bound to 14-3-3 zeta in the brain , a mechanism that could account for the persistent activation of PKCζ in rat-1 fibroblasts. PI 3-kinase, an upstream activator required for PKCζ activation, is also inhibited by PHE in rat-1 fibroblasts . Recently, it has been demonstrated that atypical PKC activation is dependent on PLD activity [27–29]. However, our data show that PLD activation with PHE was associated with a decrease in PKCζ activity in rat-1 fibroblasts.
Our data also shows that the decrease in PKCζ activity elicited by PHE is not matched by a decrease in the phosphorylation of Thr 410. PKCζ is maintained in an inactive state by direct binding of the N-terminal pseudosubstrate domain to the C-terminal catalytic domain . Phosphorylation of the activation loop at Thr 410 is necessary and sufficient to activate the kinase function of PKCζ after autophosphorylation of Thr 560 [36–38]. In VSMC, we observed a simultaneous increase in PKCζ phosphorylation and activity in response to norepinephrine . However, at 5 min of stimulation, PKCζ phosphorylation continued to increase whereas PKCζ activity was already decreasing . Therefore, although phosphorylation at Thr 410 is a prerequisite for PKCζ activation as shown by different studies [36–38], it is not clear if dephosphorylation of Thr 410 is a prerequisite for its inactivation. It has been recently reported that in fact phosphorylation of Thr 410 and subsequent autophosphorylation of Thr 560 targets PKCζ towards proteosomal degradation . Moreover, proteins that bind to PKCζ may directly inhibit its activity. For example, PAR-4, product of a gene induced during apoptosis, inhibits atypical PKCζ and PKCλ/ι activity through direct protein-protein interaction . This mechanism of inhibition of PKCζ activity may not require dephosphorylation of the enzyme. Thus, phosphorylation at Thr 410 seems to be required for PKCζ activation whereas inhibition of PKCζ activity seems to be independent of Thr 410 dephosphorylation and may involve other proteins. Since PKCζ immunoprecipitation and assay were carried out in non-denaturing conditions, it is likely that the interaction of PKCζ with other proteins, such as with an endogenous inhibitor, is conserved during the assay.
The activation of PLD elicited by PHE may be independent of PKCζ catalytic activity since constitutively active T410E PKCζ did not alter the extent of PHE-induced PLD activation. It is widely accepted that PKCζ activates downstream targets through a phosphorylation-dependent mechanism. However, it has been reported that atypical PKCζ may also directly stimulate MEK5/ERK5 pathway through its N-terminal regulatory domain (containing the pseudosubstrate site), independent of its catalytic activity . Therefore, PKCζ may stimulate downstream targets independent of its activity and phosphorylation state. Moreover, the actual proposed mechanism of PLD1 activation by the classical isoform PKCα is independent of the catalytic activity of PKC . In contrast, the fact that catalytically inactive PKCζ potentiated PLD activity and PKCζ activity was decreased after PHE stimulation indicate that an initial decrease in PKCζ activity may be required for a non-catalytic PKCζ-dependent activation of PLD activity. On the other hand, the decrease in PKCζ activity could be a consequence and not a prerequisite for this potential mechanism of action.
Regulation of PLD activity by PKCζ may involve the pseudosubstrate domain or a regulatory domain of PKCζ. PSζ is utilized to inhibit PKCζ activity [14, 30]. However, we show that PHE is a better inhibitor than PSζ and PSζ further blocks the decrease in PKCζ activity elicited by PHE. These data and the lack of decrease in PKCζ phosphorylation suggest that other proteins activated by PHE may be involved in reducing PKCζ activity. In addition, PSζ blocked PHE-induced PLD activation. Therefore, PSζ may alter PKCζ function(s) through two different actions. First, it inhibits kinase activity through binding to the catalytic site, which mimics the endogenous pseudosubstrate domain [14, 33]. Second, it may act a competitive inhibitor of the interaction of the regulatory N-terminal domains (PB1, PS, C1) [42, 43] of PKCζ with other effectors, such as PAR, MEK5, p62/ZIP, tubulin or other proteins [20, 21, 40, 41, 44, 45]. PKCζ directly stimulates the MEK5/ERK5 pathway through its N-terminal regulatory domain independently of its catalytic activity . PSζ may competitively inhibit the interaction of the regulatory domain with a protein containing an aPKC interaction domain, such as MEK-5 or the scaffold protein p62, in addition to its effect on catalytic activity. This mechanism of action would explain the effect of PSζ on PLD activity and PKCζ activity. It should be noted that a PKCα pseudosubstrate inhibitor did not alter PHE-induced PLD activity (Fig. 8), underscoring the selectivity of myristoylated PSζ.
An alternative hypothesis involves a PKCζ-mediated PLD phosphorylation that would keep PLD inactive when PKCζ is active. However, our data do not support this, since kinase inactive PKCζ or myristoylated PSζ did not stimulate PLD activity in the absence of PHE stimulation. Moreover, PLD phosphorylation is probably not involved in its mechanism of activation .
PKCα, the only classical isoform found in rat-1 fibroblasts expressing α1A AR, was not activated, as previously shown in these cells . A previous study by Taguchi et al.  showed an increase in PKCα translocation in response to PHE in the same cell model. The main difference between the two studies resides in the PHE concentration, 2 μM PHE in our work vs. 100 μM PHE , indicating that PHE is able to significantly stimulate PKCα translocation at concentration higher than 2 μM.