PDK1 governs thromboxane generation and thrombosis in platelets by regulating activation of Raf1 in the MAPK pathway: comment

PDK1 has gained interest in the field as a potential target for antithrombotic therapies [1]; however, the role of PDK1 downstream of the P2Y_1/12 receptors remained unknown. We therefore read with great interest a recent report published in the Journal of Thrombosis and Haemostasis, by Manne et al. [2] describing the role of PDK1 in regulating platelet activation and TxA₂ generation downstream of the P2Y_1/12 receptors. In their study, Manne et al. used both pharmacological and genetic methods to conclusively show that targeting PDK1 can significantly attenuate pulmonary thromboembolism-induced mortality in an in vivo murine model. They further show that PDK1 regulates activation of Raf1 in the MAPK pathway induced by 2MeSADP.

Cytoplasmic phospholipase A₂ (cPLA₂) is recognized as one of the main PLA₂ that is responsible for the release of AA from membrane phospholipids and subsequent TxA₂ generation in platelets. Phosphorylation of cPLA₂ at Ser⁵⁰⁵ increases its activity [3]. While initial studies of cPLA₂ in cell free kinase assays showed that cPLA₂ could be phosphorylated directly by either ERK or p38 [4]; follow-up studies using p38 inhibitors suggested that it was p38 and not ERK1/2, which is responsible for phosphorylating cPLA₂-Ser⁵⁰⁵ in platelets [5]. Unfortunately, the identity of the kinase directly responsible for phosphorylation of cPLA₂-Ser⁵⁰⁵ remained in limbo for over a decade. However, recent reports using knockout mice may have solved this conundrum. Recently, we found that Ask1 null platelets showed a loss of p38 phosphorylation, but enhanced activation of ERK1/2 when stimulated with thrombin, collagen, or ADP. Interestingly, we observed a complete absence of cPLA₂ phosphorylation despite augmented ERK1/2 activity [6]. These data suggested that it was in fact p38, and not ERK, which is responsible for cPLA₂-Ser⁵⁰⁵ phosphorylation. This result was further supported by a contemporary report by Shi et al., who found that p38α null platelets also showed an absence of cPLA₂-Ser⁵⁰⁵ phosphorylation despite ERK1/2 being hyperphosphorylated [7].

In their report, Manne et al. described a mechanism of cPLA₂ phosphorylation in platelets induced by 2MeSADP. Manne et al. report that: (1) the P2Y_1/12 agonist 2MeSADP induces cPLA₂-Ser⁵⁰⁵ phosphorylation in an ERK1/2 and p38 dependent manner, as the PDK1 inhibitor BX-795 was able to block ERK1/2 and cPLA₂ phosphorylation, while the p38 inhibitor SB203580 also blocked cPLA₂ phosphorylation. (2) p38 can be activated downstream of P2Y_1/12 independently of ASK1 as the ASK1 inhibitor MSC2032964A failed to block 2MeSADP-induced p38 activation [2]. However, previously published data from our group and others suggests that ERK1/2 is not the main kinase responsible for directly phosphorylating cPLA₂ in platelets [6, 7], and that agonist-induced p38 phosphorylation is exclusively dependent on ASK1 signaling in platelets [6]. The difference in these studies is that we and Shi et al., used ADP and not 2MeSADP as used by Manne et al.

To resolve this discrepancy, we performed a series of experiments using platelets isolated from Ask1^−/− mice. In agreement with our previous report, when Ask1 null platelets were stimulated with 2MeSADP, we found that p38 activation was completely ablated while Erk1/2 activation was augmented (Fig. 1A–B). At a low dose, 2MeSADP (2.5nM) rapidly (within 1min) phosphorylated cPla₂ in WT platelets but failed to cause any cPla₂ phosphorylation in Ask1 null platelets, even after 3 minutes. However, when the dose was increased to 50nM we found that cPla₂ was phosphorylated, albeit at a later time point (3 minutes post stimulation) (Fig. 1A–B). While our data indicates that cPla₂ can be phosphorylated in the absence of p38 activity, this phosphorylation appears to occur downstream of outside-in signaling. In fact, cPLA₂ has been shown to be associated with α_IIbβ₃ and is activated downstream of outside-in signaling [8]. Additionally, it has been shown that ADP-induced TxA₂ generation requires integrin outside-in signaling [9]. In addition, this data suggests, unlike Manne et al., that p38 is exclusively activated downstream of ASK1 in platelets.

Figure 1. PDK1 and ASK1 regulate ADP-induced cPLA₂ phosphorylation and TxA₂ generation through two distinct mechanisms.

(A) Western blot of lysates of washed murine platelets (2×10⁸ platelets/mL) isolated from WT or Ask1^−/− mice and stimulated with 2MeSADP (2.5 or 50nM). Blots were probed with phospho-specific antibodies against cPla2-Ser⁵⁰⁵, Erk1/2, and p38 as indicated. Blots were reprobed with anti-Hsc70, anti-p38, and anti-Erk1/2 to ensure equal loading. (B) Quantitation of optical density of bands from A for cPla₂-Ser⁵⁰⁵ (Bi–ii), Erk1/2 (Biii–iv), and p38 (Bv–vi) when stimulated with either 2.5nM (Bi, iii, v) or 50nM (Bii, iv, vi) 2MeSADP. (C) Western blot of lysates of washed human platelets (4×10⁸ platelets/mL) pretreated with GS-4997 for 20 minutes and stimulated with 2MeSADP (100nM) for 1, 3, and 5 minutes. Blots were processed as in A. (D) Quantitation of optical density of bands from C for ASK1-Thr⁸⁴⁵ (Di), cPLA₂-Ser⁵⁰⁵ (Dii), ERK1/2 (Diii), and p38 (Div). (E) Western blot of human platelet lysate (4×10⁸ platelets/mL) pretreated with GS-4997 and stimulated with convulxin (50ng/mL) for 3 minutes. (F) Quantitation of band intensity was expressed as percent inhibition from E. (G) Schematic representation of the signaling mechanism (based on our observations and those of Manne et al.,) describing the relationship between ASK1/p38 and PDK1/ERK1/2 as it relates to cPLA₂-Ser⁵⁰⁵ phosphorylation. Two-way ANOVA (B & D) and liner regression analysis (F) were performed using Graphpad Prism Software. Representative blots are shown in panel (A, C, and E). Results are from 3 independent experiments.

In addition to murine platelets, when we stimulated human platelets with 100nM 2MeSADP we found that ASK1 and p38 were activated, albeit weekly. We found that pretreating platelets with an ASK1 specific inhibitor, GS-4997, significantly attenuated ASK1, p38, and cPLA₂ phosphorylation while having minimal effect on ERK1/2 activation (Fig. 1C–D), suggesting again that p38 is solely activated downstream of ASK1 in human and murine platelets and that ASK1/p38 signaling regulates cPLA₂ phosphorylation.

To determine if our findings are specific to P2Y_1/12 signaling, or can be attributed to a more generalized signaling mechanism, we examined if ASK1 regulated p38 activation and cPLA₂ phosphorylation downstream of convulxin. When washed human platelets stimulated with 50ng/mL of convulxin, we observed robust activation of ASK1, p38, cPLA₂, and ERK1/2. Interestingly, except for ERK1/2, phosphorylation of ASK1, p38, and cPLA₂ was dose dependently attenuated by GS-4997 (Fig. 1E–F), suggesting that it is p38, and not ERK1/2, which is responsible for phosphorylating cPLA₂ in platelets.

Taken together, our data suggests that: (1) p38 is exclusively phosphorylated downstream of ASK1 in platelets and (2) that phosphorylation of cPLA₂ at Ser⁵⁰⁵ is primarily mediated by p38 in platelets. Given that both Pdk1 null murine platelets and human platelets pretreated with BX-795 show a dramatic reduction in the level of cPLA₂ phosphorylation and TxA₂ generation [2], it is reasonable to conclude that PDK1 is involved in the regulation of cPLA₂ in platelets. However, Manne et al.’s findings showing that both SB203580 and BX-795 can completely abolish cPLA₂ phosphorylation, suggests that both p38 and PDK1 regulate cPLA₂ phosphorylation albeit through distinct mechanisms as depicted in Figure 1G. It is possible that, in platelets, p38 is the kinase that phosphorylates cPLA₂ and PDK1 regulates cPLA₂ phosphorylation via ERK1/2 by negatively regulating a cPLA₂-Ser⁵⁰⁵ phosphatase [10].