Protein kinase Cε and protein kinase Cθ double-deficient mice have a bleeding diathesis

Summary Background: In comparison to the classical isoforms of protein kinase C (PKC), the novel isoforms are thought to play minor or inhibitory roles in the regulation of platelet activation and thrombosis. Objectives: To measure the levels of PKCθ and PKCε and to investigate the phenotype of mice deficient in both novel PKC isoforms. Methods: Tail bleeding and platelet activation assays were monitored in mice and platelets from mice deficient in both PKCθ and PKCε. Results: PKCε plays a minor role in supporting aggregation and secretion following stimulation of the collagen receptor GPVI in mouse platelets but has no apparent role in spreading on fibrinogen. PKCθ, in contrast, plays a minor role in supporting adhesion and filopodial generation on fibrinogen but has no apparent role in aggregation and secretion induced by GPVI despite being expressed at over 10 times the level of PKCε. Platelets deficient in both novel isoforms have a similar pattern of aggregation downstream of GPVI and spreading on fibrinogen as the single null mutants. Strikingly, a marked reduction in aggregation on collagen under arteriolar shear conditions is observed in blood from the double but not single-deficient mice along with a significant increase in tail bleeding. Conclusions: These results reveal a greater than additive role for PKCθ and PKCε in supporting platelet activation under shear conditions and demonstrate that, in combination, the two novel PKCs support platelet activation.


Introduction
Following damage to the blood vessel wall, components of the subendothelial matrix such as collagen become exposed and activate platelets circulating in the blood. Platelet activation involves granule secretion, activation of the major platelet integrin aIIbb 3 , actin rearrangements and generation of thrombin, which amplifies platelet activation, thrombus formation and hemostasis.
The protein kinase C (PKC) serine/threonine kinase family plays a critical role in the regulation of several processes involved in platelet activation. Broad spectrum inhibition of all PKC family members blocks platelet responses to most agonists, including collagen and thrombin [1][2][3]. The PKC superfamily consists of 10 isoforms subdivided into classical (a, bI, bII, c), novel (d, e, g, h) and atypical (n, i/k) isoforms on the basis of their domain structure and sensitivity to 1,2-diacylglycerol and Ca 2+ . Robust expression of several isoforms has been reported in human (a, b, d, h) and mouse (a, b, e, d, h) platelets, with evidence of expression of additional isoforms. In contrast, there are conflicting reports on expression of PKCe in human platelets [4][5][6][7], although a recent study using an inhouse antibody has demonstrated robust expression [8].
The role of individual isoforms in platelet activation has been investigated using isoform-specific inhibitors and mice deficient in single isoforms [4][5][6][7][9][10][11][12][13][14][15][16][17][18][19]. This has led to the conclusion that the classical isoforms play positive roles in platelet activation, with PKCa playing the predominant role, supported by PKCb, while the novel isoforms play minor or inhibitory roles [10]. There is, however, a need for caution in this generalized overview as there is increasing evidence that individual isoforms of classical and novel PKCs have agonist-specific roles. This is illustrated by PKCe, which supports activation of mouse platelets by GPVI through serine phosphorylation of the FcR c-chain, leading to increased binding activation of the tyrosine kinase Syk [6]. In contrast, PKCe has been shown to play a role in the negative regulation of G protein coupled receptor signaling, in particular in the regulation of ADPinduced platelet dense granule secretion [8,20]. In comparison, the novel isoform PKCh is required for aIIbb 3 -mediated adhesion and filopodial generation on fibrinogen [14,18] but has only a minor role downstream of GPVI and PAR receptors, with both stimulatory and inhibitory results observed, possibly reflecting subtle changes in experimental conditions [7,10,14,15,18,21,22]. The role of PKCh in platelet aggregation under flow conditions is unclear, with a stimulatory [7,15], inhibitory [10,14] and or no significant role [18] being reported, again most likely reflecting differences in the experimental conditions.
In the present study, we have investigated platelet activation in mice deficient in the two novel isoforms, PKCh and PKCe. Although mice deficient in multiple classical isoforms have been described [23], this is the first report of mice lacking two novel isoforms. Mice deficient in PKCh and PKCe show a marked reduction in aggregation on collagen at arteriolar shear and exhibit a significant increase in tail bleeding in vivo relative to wild-type (WT) mice or mice deficient in one novel isoform. These results reveal a combined positive role of PKCe and PKCh in supporting platelet activation.

Washed platelet preparation
Mouse blood was drawn either by cardiac puncture or from the vena cavae of terminally CO 2 -narcosed mice, anesthetized with gaseous isofluorane. Blood was taken into 100 lL ACD and 200 lL modified Tyrodes-Hepes buffer (138 mM NaCl, 2.7 mM KCl, 1 mM MgCl 2 , 3 mM NaH 2 PO 4 , 5 mM glucose and 10 mM HEPES) pH 7.3 and centrifuged at 200 g for 6 min; separation from PRP was by spinning at 1000 g for 6 min. Platelets were adjusted to ensure a count of at least 10 8 platelets mL )1 . Human platelets were isolated as previously described [6] with ethical approval from the Oxford Research Ethics Council (reference number 08/H0605/123).

Quantification of the PKC isoform levels
The PKC isoforms were quantified using washed platelet samples from three wild-type mice and five human donors using antibodies specific for each isoform. The relative expression of levels of each PKC isoform was determined using quantitative western blotting [24][25][26]. A range of known concentrations of reference samples, GST-tagged forms of each human PKC isoform (purchased from Enzo Life Sciences, Exeter, UK), were subjected to western blotting and bands were quantified using ECL in combination with the Biorad GelDoc system. Corresponding recognition regions are at least 97% conserved between human and mouse isoforms. Expression levels were calculated by comparing the level of PKC isoforms present in platelet samples with those of the reference samples. The level of expression of PKC isoform per platelet was calculated both as the number of molecules per platelet and as a concentration.

Aggregometry and ATP release
Aggregation and ATP secretion were monitored following stimulation by the required agonist as previously described using washed platelets [6].

Flow cytometry
Expression of cell surface glycoproteins was measured by flow cytometry [6].

Aggregate formation on collagen under shear
Anticoagulated (heparin [5 IU mL )1 ] and PPACK [40 lM]), whole blood was perfused through collagen-coated capillaries at a shear rate of 1000 s )1 . Thrombus formation was imaged using phase-contrast microscopy and expressed as the percentage of surface area covered by platelets. Capillary contents were lysed and levels of adherent platelets assessed by western blotting for actin [6].

Tail bleeding
Tail bleeding experiments were performed on 20-35 g male and female mice, anesthetized with isofluorane and injected with buprenorphine intraperitoneally. The terminal 3 mm of tail was removed using a sharp razor blade and blood collected. Mice were allowed to bleed until they lost either 15% blood volume or for a maximum of 20 min. Data are presented as ratio of amount of blood loss (mg)/mouse weight (g) and rate of blood loss (mg min )1 ).

Statistical analysis
For all results n ‡ 3 for WT, PKCh )/) and PKCh )/) /e )/) mice. Statistical analyses were carried out on data using unpaired, two-tailed StudentÕs t-test, and P < 0.05 was considered statistically significant. Values are expressed as mean ± SEM.

PKCh )/) /e )/) mouse platelets exhibit normal expression of the other PKC isoforms
Quantification of the levels of PKC isoforms in both human and mouse platelets revealed that expression varies over more than two orders of magnitude. PKCh is the most highly expressed isoform in both species, even though its role has proven difficult to define. In comparison, the level of PKCe expression in mouse platelets is < 5% of that of PKCh (Fig. S1). The level of PKCe in human platelets, however, is unclear as we were not able to detect expression using commercially available antibodies, although a recent study using an in-house antibody has reported robust expression [8].
To determine whether any functional redundancy exists between the novel isoforms PKCh and PKCe, mice deficient in both isoforms were bred and their platelet activity monitored in comparison to PKCh )/) , PKCe )/) and wild-type controls. Mice deficient in both PKCh and PKCe were indistinguishable from littermate controls for up to 30 weeks and had similar platelet counts and platelet size (data not shown). The expression of the major PKC isoforms was compared in WT, PKCh )/) and PKCh )/) /e )/) washed platelet lysates (Fig. 1). As expected, no expression of PKCh, and neither PKCh or PKCe, could be detected in platelets purified from PKCh )/) and PKCh )/) /e )/) mice, respectively. There was also no significant changes in expression of other PKC isoforms in PKCh )/) /e )/) platelets relative to PKCh )/) or WT platelets (Fig. 1). We have also previously reported that expression of other PKC isoforms is not altered in mice deficient solely in PKCe [6]. Expression of GPVI, GPIb and aIIbb3 were also similar in double-deficient platelets to those in controls (Fig. S2). Similar observations have been reported for the single nulls ( Fig. S2) [6,7,14,15,18]. This indicates that any functional differences between the PKCh )/) / PKCe )/) platelets relative to PKCh )/) and PKCe )/) or WT platelets are not due to altered expression of surface receptors or other PKC isoforms.

Distinct roles for PKCh and PKCe in platelet activation
We have previously reported a mild defect in aggregation and dense granule secretion in PKCe )/) mouse platelets to the GPVI-specific agonist collagen-related-peptide (CRP) [6]. In contrast, we found no significant difference in aggregation or dense granule secretion in PKCh )/) mouse platelets relative to controls in response to concentrations of CRP that induce partial and full aggregation ( Fig. 2A). Washed platelets from PKCh )/) /e )/) mice exhibit a similar delay in onset and reduction of aggregation and ATP secretion in response to low and high concentrations of CRP as that previously reported in PKCe )/) mouse platelets (Fig. 2B) [8]). Direct comparison of PKCh )/) /e )/) and PKCe )/) mouse platelets confirmed a similar defect in CRP-induced responses (Fig. 2C). Thus, PKCh does not play a critical role in GPVI signaling even in the absence of the novel isoform PKCe. No significant difference was observed in the rate or extent of aggregation or dense granule secretion in PKCh )/) or PKCh )/) /e )/) platelets relative to WT mice in response to threshold concentrations of thrombin (Fig. S3).
As we have previously shown [6], adhesion and filopodial generation on fibrinogen are not altered in the absence of PKCe (Fig. 3). In contrast, a reduction in adhesion and filopodia generation on a fibrinogen-coated surface was observed in PKCh )/) platelets in agreement with earlier studies [14,18]. A similar defect was also observed in PKCh )/) /e )/) platelets (Fig. 4), which was indistinguishable from that seen in PKCh )/) platelets. Thus, PKCe does not play a critical role in fibrinogen signaling even in the absence of the novel isoform PKCh.
Platelet activation is reinforced by the feedback agonists ADP and thromboxane A 2 . Aggregation induced by a low concentration of the thromboxane agonist mimetic, U46619, was not altered in platelets deficient in either PKCh or PKCe or in the absence of both novel isoforms (data not shown). We and others have previously reported that secretion in response to ADP is potentiated in the absence of PKCe, although this did not translate into a change in aggregation [6,8,20]. This result was confirmed in the present study, although interestingly potentiation was not observed in the absence of both PKCe and PKCh (Fig. 4). This highlights a positive role for PKCh in the  regulation of ADP-induced dense granule secretion, which opposes that of PKCe. These results suggest isoform-specific rather than redundant roles for the two novel PKC isoforms, PKCh and PKCe, in supporting platelet aggregation, secretion and adhesion and filopodial generation. platelet aggregation and thrombus formation under arteriolar flow rates were investigated by flowing whole blood over immobilized collagen at an arteriolar shear rate of 1000 s )1 . As we have previously shown, PKCe )/) platelets show no significant difference in aggregate and thrombus formation under these conditions (Fig. 5A) [6]. There was also no significant difference in aggregation on collagen in PKCh )/) blood under these conditions (Fig. 5A). In contrast, a marked reduction in platelet aggregation on collagen was observed at a shear rate of 1000 s )1 in PKCh )/) /PKCe )/) platelets (Fig. 5A), with platelets forming small unstable aggregates. This result demonstrates that at high shear platelet aggregation on collagen is regulated by the combined action of PKCh and PKCe.
We further investigated whether PKCh and PKCe are required for hemostasis in vivo using a tail bleeding assay. There was no significant increase in bleeding times in the single isoform null mice in comparison to WT controls. In contrast, there was a marked increase in blood lost and time to occlusion in the double-deficient mice (Fig. 5B). This reveals that the two novel isoforms also work in combination to support hemostasis.

Discussion
It has been proposed that the classical isoforms PKCa and PKCb play the dominant positive roles in the regulation of platelet activation and thrombus formation, whilst the novel isoforms are thought to play comparatively minor or in some cases negative regulatory roles [9,10]. In the present study, however, we show that the combined loss of the novel isoforms PKCe and PKCh results in a marked defect in aggregation under flow conditions and a marked increase in tail bleeding, revealing a combined net positive role for the two novel isoforms in hemostasis, which may reflect their individual roles in platelet activation by collagen and fibrinogen, respectively.
PKCh is the most highly expressed PKC isoform in both human and mouse platelets. Surprisingly, the high level expression of PKCh is not associated with a major change in GPVI and PAR4 receptor signaling in mouse platelets [7,10,14,15,18,21,22], a result that has been confirmed in the present study. Differences with regard to PKCh function are most likely due to minor changes in experimental design and the relatively mild role of the novel PKC isoform in platelet activation downstream of GPVI and PAR. On the other hand, PKCh regulates adhesion and filopodia formation on fibrinogen [14,18], although this is not associated with altered aggregation on collagen at an arteriolar rate of shear (present study) or to a change in tail bleeding time [18]. In comparison, PKCe is expressed at < 5% of the level of PKCh in mouse platelets and the variable reports of its presence in human platelets are consistent with a low level expression (see Introduction). We have shown that PKCe plays a key role in supporting platelet activation by GPVI [6], as well as secretion (but not aggregation) by ADP in mouse platelets [20].
The use of mice deficient in both PKCe and PKCh enables the net contribution of the two isoforms to be studied. The in vitro studies reveal that the isoform-specific functions of the two novel PKCs are carried over to the double-deficient platelets, with the only departure being the loss of the increased secretion to ADP in the PKCe-null platelets, suggesting that PKCh opposes this response. Nevertheless, despite the rela- tively minor phenotype in response to individual agonists, the present study demonstrates a marked reduction in aggregation on collagen under arteriolar flow rates and significant increase in tail bleeding in the combined absence of PKCh and PKCe. It is already known that the two classical isoforms of PKC, PKCa and PKCb, play a major role in supporting platelet activation under static and flow conditions [10,12]. The observation of a significant defect in aggregation on collagen at an arteriolar rate of shear and in hemostasis (tail bleeding assay) in the double-deficient mice demonstrates that, in combination, PKCe and PKCh also contribute to activation as a consequence of distinct roles in platelet activation by GPVI and integrin aIIbb3, respectively [10,12]. Thus both classical and novel PKC isoforms are required for hemostasis in the arteriolar system.

Disclosure of Conflict of Interests
The authors state that they have no conflict of interest.

Supporting Information
Additional Supporting Information may be found in the online version of this article: Figure S1. Quantification of PKC isoforms in human and mouse platelets. Figure S2. PKCh )/) /e )/) platelets express normal levels of major cell surface receptors. Figure S3. The role of PKCh and PKCe in thrombin-induced platelet activation in vitro.
Please note: Wiley-Blackwell is not responsible for the content or functionality of any supporting materials supplied by the authors. Any queries (other than missing material) should be directed to the corresponding author for the article.