Data reported here are the first to demonstrate direct association of Pak-1 to Erk1/2 in WT hearts. The involvement of Pak-1 and its upstream signals Cdc42 and Rac1 in cell remodeling had indicated a potential role in cardiac hypertrophy and dilatation [18
]. However, the molecular mechanisms underlying the effects produced by Pak-1 were uncertain. Based on our evidence that there is enhanced activity of Erk1/2 in the Pak-1-KO mouse heart, we propose that a significant role of Pak-1 is to normally suppress Erk1/2 activity. The mechanism is likely to involve a regulation of the β-adrenergic signaling, through a mechanism mediated by Pak-1, which, in turn, regulates PP2A activation. Enhanced activation of Erk1/2 and reduced activation of PP2A in ISO-treated Pak-1-KO mouse hearts support this mechanistic interpretation. Additionally, administration of FR180204, an Erk inhibitor, reduced the increased LV hypertrophy in ISO-treated Pak-1-KO mice. Overall, our findings indicate that in the absence of Pak-1, there is an exacerbation of the stress response of the myocardium to isoproterenol, which supports the idea that Pak-1 is an anti-hypertrophic signaling kinase and may serve the role as a natural modulator of the β-adrenergic signaling cascade and the Erks.
Previous studies support our hypothesis that a significant element in the Pak-1 signaling cascade is activation of PP2A and suppression of effects of β-adrenergic stimulation. Pak-1 has been shown previously to activate PP2A [1
], which in turn dephosphorylates cTnI [1
], myosin-binding protein C [1
], phospholamban [20
], inwardly rectifier potassium channels [21
], and L-type Ca2+
]. In the current study, we demonstrated association of Pak-1, PP2A, and Erk1/2 in vivo
. Previous studies have found β-adrenergic stimulation in cardiomyocytes [23
], HEK293 cells [24
], COS-7 cells [25
], and Chinese hamster ovary cells [26
] results in increased Erk phosphorylation. Our results demonstrate increased Erk1 phosphorylation at T202 and Y204 residues, increased Erk2 phosphorylation at T188, and increased phosphorylation of PP2A at Y307 in Pak-1-KO mice treated with ISO compared to all other groups studied (Pak-1-KO/CTRL, WT/CTRL, WT/ISO). This demonstrates that β-adrenergic stimulation triggers enhanced Erk phosphorylation due to a suppression of PP2A activation in the absence of Pak-1, promoting Erk-induced LV cardiac hypertrophy. This is in agreement with previous studies, which reported Erk1/2 to be involved in the development of cardiac hypertrophy [27
]. Thus, Pak-1 may have a role in regulating the progression to LV cardiac hypertrophy during β-adrenergic stimulation through regulation of Erks. Previously, Pak-1 has been reported to counter-regulate adrenergic stimulation by mediation of PP2A [11
], and a role for PP2A has been shown in regulating MEK and Erk in a receptor-independent manner [28
]. In the present study, intracellular phosphorylation and activation of Erk1/2 during β-adrenergic stimulation may be maximal in the Pak-1-KO mice because the loss of Pak-1 coupling to PP2A leads to the consequent loss of an association between PP2A and Erk1/2. illustrates this proposed mechanism of LV myocardial hypertrophy.
Proposed mechanism of LV myocardial hypertrophy in a Pak-1-KO mouse model
Although our data indicated that Akt phosphorylation was not significantly different among any of the groups studied, some studies report Pak-1 activation of Akt [29
]. Mao et al. reported that endogenous Pak-1 is physically associated with Akt in cardiac cells and may act as a potential phosphoinositide-dependent protein kinase-2 essential for regulation of Akt phosphorylation [29
]. However, others report Akt regulation of Pak [31
], and yet others report that these two proteins may possibly regulate each other [30
]. An interdependence of Pak and Akt may explain why we did not detect a change in Akt phosphorylation in hearts of the Pak-1-KO mouse model.
Our studies employing a KO mouse model provided a unique advantage in investigating the effects of Pak-1 on p38 MAPK and JNK1/2/3. Despite an extensive number of studies of Pak-1 effects on MAPK in various cell types, the role of Pak-1 in cardiac tissue remained uncertain. In the particular model investigated here our experiments demonstrated that phosphorylation of MAPK JNK1/2/3 was similar in all the models previously investigated. However, there was a significant depression of phosphorylation of p38 MAPK in Pak-1-KO mice in the presence and in the absence of ISO, compared to ISO-treated and untreated WT mice. This result agrees with previous studies demonstrating that CA Pak mutants activate p38 MAPK. We speculate that the reduction in phosphorylation of p38 we observed in the Pak-1-KO hearts is the consequence of the lack of Pak-1, a well-known activator of p38 [32
]. Yet our data contrast with studies that reported activation of JNK in COS-7 cells [33
]. Pak-1 and Pak-2 in HEK 293 cells induced activity of p38 MAPK but also JNK/SAPK [34
]. CA Pak-3 was also reported to activate JNK1 in COS1 cells [35
]. In that study, activated Cdc42 stimulated the activity of p38 MAPK, but was a less effective activator of Erk2 [35
]. Others found that expression of the constitutive human Pak isoform, hPak-1, a kinase in COS7 mammalian cells, led to specific activation of the JNK1 MAPK pathway, but not the Erk MAP kinase pathway [33
]. These varied results may arise from different Pak isoforms being studied in different cell types isolated from different species.
An unexpected and novel finding in our studies was the demonstration of a significant increase in tension generated by skinned fiber preparations from Pak-1-KO/ISO hearts compared to all other groups. This increase in tension generating capability indicates that in addition to the increase in cardiac mass in the Pak-1-KO/ISO hearts compared to controls, the relative increase in maximum tension was likely to contribute to the enhanced contractility reflected in a reduced ESV and EDV. To determine the mechanism for this increase in tension, we performed an analysis of protein phosphorylation in myofilament preparations from the experimental groups. Previous studies had indicated the potential for complex Pak-1 related mechanisms controlling phosphorylation of myofilament proteins. In vitro
studies demonstrated direct phosphorylation of cTnI, cTnT, and desmin by Pak-3 [36
] and phosphorylation of cTnI by Pak-1 [1
]. However, we also reported that activation of PP2A by Pak-1 induces dephosphorylation of cTnI and myosin binding protein C [1
]. The analysis in the present paper showed no changes of phosphorylation of myofilament proteins except for cTnI. In the case of cTnI it is seemed likely, and indeed our data demonstrated, that S23/S24, well known PKA-sites, would be phosphorylated in WT/ISO compared to controls. On the other hand S23/S24 residues in the Pak-1-KO hearts were nearly fully phosphorylated, and thus there was little further increase in Pak-1-KO/ISO hearts. In view of evidence that cTnI S150 is site phosphorylated by Pak-1 [36
], we assessed modifications in this residue. With ISO treatment in WT hearts, phosphorylation of S150 increased significantly. However there was no difference between in cTnI-S150 phosphorylation between WT controls and Pak-1-KO with or without ISO treatment. These data either further support the evidence of direct phosphorylation of S150 by Pak-1 or indicates discoordinate dephosphorylation of cTnI by Pak-1-PP2a. Phosphorylation at S23/S24 is known to depress [38
], whereas phosphorylation at S150 of cTnI is known to enhance myofilament Ca-sensitivity [36
]. This may account for the lack of differences in pCa-tension relations between preparations from WT controls and WT/ISO. We have no clear interpretation of the mechanism of the enhancement of maximum tension in the Pak-1 KO/ISO group. This appears to represent a novel and previously unknown state of cTnI, which may involve as yet undetermined modifications as noted in our 2-D DIGE analysis ().
Yet, the major implications of our data remain with regard to our demonstration of the significant role of Pak-1 as a determinant of growth signaling and sarcomeric function in the myocardium. We conclude that an important role of Pak-1 is its function as a natural inhibitor of the Erks and a novel anti-hypertrophic signaling enzyme with a role in modulation of β-adrenergic signaling. Thus, Pak-1 plays a significant contribution in the mechanism of adaptive control of cardiac contractility.