Hybrid Transgenic Mice Reveal In Vivo Specificity of G Protein–Coupled Receptor Kinases in the Heart
Abstract—G protein–coupled receptor kinases (GRKs) phosphorylate activated G protein-coupled receptors, including α1B-adrenergic receptors (ARs), resulting in desensitization. In vivo analysis of GRK substrate selectivity has been limited. Therefore, we generated hybrid transgenic mice with myocardium-targeted overexpression of 1 of 3 GRKs expressed in the heart (GRK2 [commonly known as the β-AR kinase 1], GRK3, or GRK5) with concomitant cardiac expression of a constitutively activated mutant (CAM) or wild-type α1BAR. Transgenic mice with cardiac CAMα1BAR overexpression had enhanced myocardial α1AR signaling and elevated heart-to-body weight ratios with ventricular atrial natriuretic factor expression denoting myocardial hypertrophy. Transgenic mouse hearts overexpressing only GRK2, GRK3, or GRK5 had no hypertrophy. In hybrid transgenic mice, enhanced in vivo signaling through CAMα1BARs, as measured by myocardial diacylglycerol content, was attenuated by concomitant overexpression of GRK3 but not GRK2 or GRK5. CAMα1BAR-induced hypertrophy and ventricular atrial natriuretic factor expression were significantly attenuated with either concurrent GRK3 or GRK5 overexpression. Similar GRK selectivity was seen in hybrid transgenic mice with wild-type α1BAR overexpression concurrently with a GRK. GRK2 overexpression was without effect on any in vivo CAM or wild-type α1BAR cardiac phenotype, which is in contrast to previously reported in vitro findings. Furthermore, endogenous myocardial α1AR mitogen-activated protein kinase signaling in single-GRK transgenic mice also exhibited selectivity, as GRK3 and GRK5 desensitized in vivo α1AR mitogen–activated protein kinase responses that were unaffected by GRK2 overexpression. Thus, these results demonstrate that GRKs differentially interact with α1BARs in vivo such that GRK3 desensitizes all α1BAR signaling, whereas GRK5 has partial effects and, most interestingly, GRK2 has no effect on in vivo α1BAR signaling in the heart.
- adrenergic receptors, α1
- protein-coupled receptor kinase
- myocardial biology
- myocardial hypertrophy
The α1B-adrenergic receptor (AR) is a member of the G protein–coupled receptor family and is the predominant α1AR subtype expressed in adult rodent myocardium.1 2 α1AR agonists, including phenylephrine (PE), have been shown to mediate intracellular responses through α1AR activation of the heterotrimeric G protein Gq, which in turn activates the effector enzyme phospholipase C (PLC). Activation of the α1AR-Gq-PLC pathway results in the cellular accumulation of inositol 1,4,5-trisphosphate and diacylglycerol (DAG), leading to increased intracellular calcium and protein kinase C activity.3 The role of α1ARs in the heart is not well understood; however, the Gq-PLC–protein kinase C pathway is important in initiating the hypertrophic response.4 In fact, we have recently shown, using transgenic (Tg) mice, that signaling through Gq is the final common trigger of in vivo pressure overload ventricular hypertrophy.5
Activation of α1ARs in cultured neonatal ventricular myocytes has been shown to induce an embryonic program of gene expression, including ventricular expression of atrial natriuretic factor (ANF), and cell hypertrophy without hyperplasia.6 7 Moreover, adult Tg mice expressing a constitutively activated mutant (CAM) of the α1BAR in a cardiac-specific manner have elevated myocardial DAG content, ventricular ANF expression, and myocardial hypertrophy, as measured by increased heart-to-body weight ratio and myocyte cross-sectional area.8 Thus, constant stimulation of the α1BAR in vivo is sufficient to induce a hypertrophic phenotype independent of hemodynamic changes.
Signaling through α1ARs is regulated, like that mediated by many other G protein–coupled receptors, via phosphorylation and the triggering of desensitization mechanisms.9 Phosphorylation of agonist-occupied α1ARs is accomplished by members of the serine/threonine G protein–coupled receptor kinase (GRK) family. Three predominant GRKs are expressed in the mammalian heart, as follows: GRK2 (commonly known as the β-AR kinase [βARK1]), GRK3 (βARK2), and GRK5.10 Although many studies have been done in vitro concerning the actions of GRKs on receptor signaling, almost nothing in vivo has been done to elucidate substrate specificity of GRKs. In fact, in vitro studies with several receptor systems important in the heart have revealed no definitive GRK selectivity.11 12 13 14 In vitro studies examining α1BARs have found that GRK2 and GRK3 could both increase agonist-induced phosphorylation of the α1BAR and promote desensitization of signaling, whereas GRK5 increased the basal phosphorylation of α1BARs without any effect of agonist-stimulated phosphorylation.12 More recently, however, GRK substrate selectivity has been identified in vitro, as Iacovelli et al15 described the way in which GRK2 stably transfected into rat thyroid FRTL-5 cells desensitized endogenous thyrotropin receptors, whereas GRK5 and GRK6 did not. Additionally, GRK2 had variable effects on the Gαi-coupled A1 adenosine receptor. Adenylate cyclase inhibition was unaffected, but mitogen-activated protein kinase (MAPK) signaling was attenuated by GRK2 overexpression.15 Importantly, and in contrast to Diviani et al,12 Iacovelli et al15 also found that GRK2 overexpression did not alter α1BAR signaling. Thus, this suggests that there may be cell-type specificity of the GRKs for their various substrates, making it essential to understand the in vivo selectivity of these kinases in the appropriate cells/tissues of interest.
Recently, the use of Tg mice with cardiac-specific overexpression of GRK2, GRK3, or GRK5 have made it possible to address in vivo, tissue-specific GRK substrate selectivity.16 17 18 The CAMα1BAR and GRK Tg mice provide a unique and powerful opportunity to create hybrid Tg mice to study the specific interactions in vivo between these 3 individual GRKs and α1BARs. In the present study, the hearts of these hybrid mice were used as novel “in vivo reaction vessels” to determine the in vivo selectivity of GRK2, GRK3, and GRK5 for myocardial α1BARs. Furthermore, these mice were used to elucidate a possible role of α1BAR desensitization in the control of myocardial hypertrophy. These results reinforce that it is essential to verify in vitro findings in vivo in the context of the whole organism to begin to understand the true complexity of the mammalian cardiovascular system.
Materials and Methods
Myocardial-specific overexpression of the corresponding transgene was targeted by the α–myosin heavy chain (αMyHC) gene promoter as previously described.19 Each of the single-Tg mouse lines used as parental crossbreeders has previously been described as noted in the Table⇓, except for GRK2-20. All Tg lines were originally generated in C57BL/6 mice. To generate the GRK2-20 mouse, a GRK2 Tg construct was prepared by inserting the complete open reading frame of bovine GRK2 into pGEM containing a 5.5-kb SacI/SalI fragment from the αMyHC promoter and the simian virus 40 intron–poly A. Tg animals were screened by Southern blot and polymerase chain reaction analysis of tail clip DNA.
Hybrid double-Tg mice were created by mating single-Tg mice together as described in the Table⇑. The animals in this study were handled according to approved protocols and animal welfare regulations at Duke University Medical Center.
Tg mouse heart crude membranes were prepared as described.19 For determination of myocardial αAR binding density, 250 pmol/L (±)-β-([125I]Iodo-4-hydroxyphenyl)-ethyl-aminomethyl-tetralone (125I-HEAT; New England Nuclear) in the absence (total binding) or presence (nonspecific binding) of 50 μmol/L prazosin was used.20
Immunodetection of GRK2 and GRK3 was carried out first with an immunoprecipitation using a monoclonal GRK2/GRK3 antibody, as described previously.18 GRK5 was detected in crude membrane protein extracts.17
Heart Weight–to–Body Weight Ratio
Mice were first anesthetized21 and weighed, and their hearts were quickly excised, blotted dry, weighed, and frozen in liquid N2. Heart weight–to–body weight ratios were calculated and expressed in mg/g.21
Ventricular ANF mRNA Analysis
The apical portion of the left and right ventricles of frozen hearts obtained as described above was homogenized, total RNA was extracted using an Ultraspec solution (Biotecx Laboratories), and Northern analysis was performed as previously described.8 20 After ANF detection, all blots were stripped and reprobed with a rat GAPDH cDNA probe. The ANF and GAPDH bands were quantified with a PhosphorImager (Molecular Dynamics), and the ANF/GAPDH signal intensity was determined.20
Lipid fractions from frozen hearts were extracted as described.8 20 32P-labeled phosphatidic acid (phosphorylated DAG) was isolated by silica gel thin-layer chromatography and quantified with the PhosphorImager. DAG content was normalized to tissue phospholipid, and the final DAG concentration was expressed as pmol of DAG/nmol of lipid phosphate, as described previously.8 20
Mice were given a 200-μL intraperitoneal injection of 500 μmol/L PE or saline. After 10 minutes, mice were anesthetized as described above and hearts were quickly removed and frozen. Excised hearts were prepared as described previously.5 18 Immunoprecipitations were performed using anti–extracellular signal–regulated kinase (ERK2) or anti–c-jun N-terminal kinase (JNK1) antibody (Santa Cruz Biotechnology). Kinase assays were carried out at 30°C for 15 minutes5 18 using myelin basic protein (MBP) (for ERK2) with glutathione S-transferase–c-jun (JNK1) as a substrate.
Data are expressed as mean±SEM. An unpaired 2-tailed Student t test was performed for all biochemical data. For all analyses, P<0.05 was considered statistically significant.
The Table⇑ lists the 15 lines of myocardium-targeted Tg mice used throughout this study, including parental single-Tg mice and the resultant hybrid double-Tg offspring. Included is the documented overexpression levels compared with the endogenous myocardial proteins. Two separate lines of GRK2 animals were used; one line, GRK2-3, has been previously described,16 17 and the other is an as-yet-unreported GRK2 mouse line, GRK2-20, with ≈20-fold enhancement of GRK activity as compared with endogenous GRK2 activity. Figure 1A⇓ illustrates the relative kinase activity that each of the single-Tg lines has in comparison with the others versus non-Tg littermate control (NLC). The primary GRK activity in the mouse heart is GRK222 ; however, all of the kinases are equally capable of phosphorylating rhodopsin. Each transgene constitutes the majority of activity when overexpressed in the heart. Whereas the Table⇑ describes overexpression levels of the various GRKs with respect to their endogenous levels, when expressed in comparison with total NLC GRK activity, Tg GRK2-3 has ≈3-fold, TgGRK2-20 has ≈20-fold, TgGRK3 has ≈5-fold, and TgGRK5 has ≈20-fold overexpression in activity (Figure 1A⇓). Hybrid Tg mice were examined to verify that both transgenes were expressed at equivalent levels as found in the single-Tg parental lines. Importantly, we found no change in expression in any of the transgene protein products in hybrid mice. For example, as shown in Figure 1⇓, protein immunoblotting revealed no difference in the level of overexpression of GRK2 or GRK3 when comparing hearts from GRK2-3 or GRK3 with CAMα1B/GRK2-3 and CAMα1B/GRK3 hearts (Figure 1B⇓). Similar results were found in hybrid Tg mice overexpressing GRK5 (Figure 1C⇓) and the GqI (data not shown). Furthermore, endogenous GRK levels were not altered in the CAMα1BAR mice (Figure 1B⇓ and 1C⇓). We also verified that GRK activity was not compromised in the hybrid Tg mice. In vitro GRK phosphorylation assays illustrate that phosphorylation activity was similar in single-Tg and hybrid Tg mice (data not shown). As shown in Figure 1D⇓, myocardial α1AR density in the hybrid CAMα1BAR and the various GRK-overexpressing lines of Tg mice was equal to single-Tg CAMα1BAR mouse myocardial α1AR levels. The myocardial α1AR density in the different GRK-overexpressing lines of Tg mice was equal to endogenous levels found in NLC mice (data not shown). The α1AR overexpression seen in the wild-type (WT) α1BAR was preserved in the presence of all second transgenes in the hybrid lines (data not shown).
In Vivo Inhibition of Hypertrophy
CAMα1BAR mice have an elevated heart-to-body weight ratio compared with NLC mice (Figure 2⇓), consistent with earlier findings that demonstrated that these mice had myocardial hypertrophy.8 When the CAMα1BAR mice were crossed with animals overexpressing GRK2 (3-fold or 20-fold), the heart-to-body weight ratio remained elevated (Figure 2⇓). However, the presence of GRK3 overexpression in the CAMα1B/GRK3 animals ablated CAMα1BAR-induced hypertrophy, as did concomitant GRK5 overexpression (Figure 2⇓). Thus, GRK3 and GRK5 overexpression were each capable of attenuating CAMα1BAR-induced hypertrophy, whereas increased cardiac GRK2 activity had no effect.
To verify that myocardial hypertrophy seen in the CAMα1BAR animals is the result of enhanced Gq signaling, we created hybrid Tg mice with CAMα1BAR overexpression and overexpression of a peptide inhibitor of the receptor-Gq interface. This peptide (GqI) represents the last 54 amino acids of murine Gαq(305–359) and has been shown to specifically inhibit Gq signaling in vivo.5 Similar to CAMα1B/GRK3 mice, CAMα1B/GqI animals had a heart-to-body weight ratio equal to that of NLCs, demonstrating that signaling through Gq is responsible for the CAMα1BAR phenotype.
CAMα1BAR Induced DAG Activation
Specific interactions between GRKs and α1BARs in vivo should result in attenuated signal transduction because of receptor desensitization, and this could account for the inhibition of CAMα1BAR-induced hypertrophy by overexpression of GRK3 and GRK5. Signaling through α1BAR-Gq proceeds through PLC activation, resulting in DAG accumulation. Therefore, basal DAG content was quantified in lipid fractions of hearts from Tg mice. Myocardial DAG content was significantly elevated 60% in CAMα1BAR hearts (Figure 3⇓), consistent with previous data.8 A similar elevation in myocardial DAG content was observed in both CAMα1B/GRK2-3 (3.87±0.31 pmol DAG/nmol lipid phosphate, n=4) and CAMα1B/GRK2-20 (4.93±0.49 pmol DAG/nmol lipid phosphate [n=4] versus NLC 2.53±0.31 pmol DAG/nmol lipid phosphate [n=4]) (Figure 3⇓). Thus, like hypertrophy, CAMα1BAR signaling is not altered by GRK2 overexpression. In contrast to these results, myocardial DAG content was significantly lowered in hybrid CAMα1B/GRK3 mice compared with CAMα1BAR animals demonstrating that signaling through CAMα1BARs is attenuated by GRK3 overexpression (Figure 3⇓).
Interestingly, and in contrast to the results seen in the heart-to-body weight ratios, hybrid CAMα1B/GRK5 mice still had significantly elevated myocardial DAG content, similar to levels seen in CAMα1BAR and CAMα1B/GRK2 animals (Figure 3⇑). To further study the interactions between GRK5 and α1BARs in vivo, we generated hybrid myocardium-targeted Tg mice overexpressing the WTα1BAR and GRK5 (Table⇑). As with CAMα1BAR overexpression, WTα1BAR overexpression in Tg mouse hearts leads to significant increases in myocardial DAG content compared with NLC animals (in pmol DAG/nmol lipid phosphate; NLC, 2.77±0.19 [n=4] versus WTα1BAR, 4.18±0.28 [n=4]; P<0.05), consistent with previous results.20 In hybrid WTα1B/GRK5 mice, myocardial DAG content was still significantly elevated above NLC at a level similar to that of WTα1BAR mice (4.98±0.54, n=6). Thus, it appears that although GRK5 expression was capable of inhibiting the development of α1B-induced hypertrophy, it does not affect basal α1BAR/PLC signaling in the heart as examined in vivo after either WT or CAMα1BAR overexpression.
Ventricular ANF mRNA
A central molecular characteristic of ventricular hypertrophy is the upregulation of a number of genes normally expressed in fetal myocardium. This includes ventricular expression of ANF. To investigate this in our series of hybrid Tg mice, we performed Northern blots on ventricular RNA, normalizing mouse ANF expression to an internal control, GAPDH (see Materials and Methods). Ventricles from NLC mice and the single GRK–overexpressing Tg mice demonstrated very faint or undetectable ANF mRNA expression, which was consistent with inactivation of this gene in normal adult ventricular myocardium (data not shown). We found that ventricular ANF mRNA levels were increased ≈700% in CAMα1BAR mice (Figure 4⇓), which is similar to previous findings.8 Hybrid CAMα1B/GRK2 ventricles still exhibited enhanced ANF mRNA levels that actually were, for reasons unknown, even higher than those of CAMα1BAR mice for the TgGRK2-3, whereas CAMα1B/GRK2-20 mice had ANF levels equivalent to those of CAMα1BAR mice (Figure 4⇓). Conversely, CAMα1B/GRK3 animals had ventricular ANF mRNA levels equal to NLC values (Figure 4⇓). This attenuation of ANF mRNA levels induced by CAMα1BAR overexpression was also seen in hybrid CAMα1B/GqI mice (Figure 4⇓), demonstrating that Gq coupling is responsible for triggering this ANF response. Thus, as with hypertrophy and DAG signaling, GRK3 overexpression attenuates the CAMα1BAR phenotype, whereas GRK2 overexpression does not significantly alter the CAM phenotype.
Interestingly, CAMα1B/GRK5 mice exhibited lower ventricular ANF expression (350±70% of NLC, n=3) compared with that of CAMα1BAR mice (570±50% of NLC, n=9), although this attenuation was not statistically significant (Figure 4⇑). Moreover, the ANF mRNA levels were still significantly higher than levels seen in NLC mice, suggesting that GRK5 over-expression does not totally block CAMα1BAR-induced ANF expression. To further examine the effects of GRK5 on α1BAR-Gq–mediated ventricular ANF induction, we studied ANF expression in WTα1BAR mice and WTα1B/GRK5 animals. WTα1BAR mice also exhibited elevated ventricular ANF mRNA levels (505±51% of NLC [n=5]; P<0.05). Interestingly, we have previously shown that this occurs without any myocardial hypertrophy.20 Consistent with findings shown in Figure 4⇑, hybrid WTα1B/GRK5 mice also had significantly enhanced ventricular ANF mRNA (492±53% of NLC [n=5]; P<0.05), demonstrating that GRK5 does not eliminate ANF expression after enhanced α1BAR-Gq signaling. WTα1B/GRK2-3 also had maintained elevated ANF expression (555±49% of NLC [n=3]; P<0.05), suggesting that regulation of α1BAR by GRKs is similar for both the WT and CAMα1BAR and that GRK2 does not affect α1BAR signaling in the heart.
ERK and JNK Activity in Response to α1AR Activation and GRK Expression
Because the data presented above demonstrate that there is apparent specificity among GRK2, GRK3, and GRK5 in desensitizing α1BARs in vivo in the heart, we examined the effects of these 3 GRKs on endogenous myocardial α1AR signaling. A relevant signaling pathway that has been demonstrated to be activated by α1AR stimulation in the heart is the MAPK pathway, including ERK1/ERK2 and JNK1.23 24 25 In this study, we used ERK and JNK activity assays to determine whether endogenous in vivo myocardial α1AR-MAPK signaling is altered in GRK Tg mice. Mice were injected intraperitoneally with either saline (basal signaling) or the α1-agonist PE, and ERK and JNK activity induced by PE (over basal activity) was determined in the GRK Tg animals and compared with values in NLC mice. After 10 minutes, injected mice were euthanized, their hearts extracted and homogenized, and ERK and JNK were immunoprecipitated for an in vitro kinase assay. Basal ERK and JNK activity measured after saline injection was similar or equivalent in NLC and single GRK–overexpressing mice (data not shown). In NLC mice, PE induced a significant 40% increase in ERK activity over basal levels of activity (Figure 5A⇓). As shown in Figure 5A⇓, GRK2 overexpression in either the GRK2-3 or GRK2-20 animals did not inhibit myocardial PE-induced ERK activity, whereas GRK3 or GRK5 overexpression significantly attenuated PE-induced ERK activity. Similar to the ERK activity results, JNK activity was attenuated in mice overexpressing GRK3, whereas GRK2 overexpression at either low or high amounts had no impact on JNK signaling induced by the α1AR agonist PE (Figure 5B⇓). Similar to what was seen with ANF signaling (Figure 4⇑), GRK5 overexpression attenuated JNK signaling only partially as compared with NLC activation (Figure 5B⇓). These data further suggest that GRK3 is the primary GRK responsible for desensitizing in vivo α1AR signaling in the heart. GRK5 had variable effects on α1BAR signaling, and the complexity of GRK5 regulation of in vivo α1BAR signaling remains to be fully elucidated.
The results of this study, which uses 15 distinct Tg mouse lines, reveal 2 significant findings. The first is the demonstration of the feasibility and power of using hybrid Tg mice with myocardial overexpression of G protein–coupled receptors with concomitant GRK overexpression to specifically study their in vivo interactions. Secondly, this hybrid Tg strategy revealed that myocardial α1ARs are in vivo targets for GRK3-mediated desensitization but not for GRK2, the GRK most abundantly expressed in the myocardium. These differing effects of GRK2 and GRK3 were confirmed in hybrid Tg mice overexpressing either the WT α1BAR or a CAM α1BAR, as well as in studies designed to examine endogenous myocardial α1AR signaling. The effects of these GRKs were assessed on the hypertrophic phenotype of these mice, Gq-PLC signaling (via myocardial DAG content), and endogenous α1AR-MAPK signaling in the heart. In addition to the differing results found with these 2 GRKs, overexpression of GRK5 caused variable effects on in vivo α1BAR signaling. GRK5 was capable of attenuating endogenous α1AR-mediated ERK activity and CAMα1BAR-induced hypertrophy. However, it had a lesser effect on CAM and WTα1BAR-mediated DAG levels, ventricular ANF mRNA expression, and JNK activity.
The hearts of these hybrid Tg mice were used in this study as novel “in vivo reaction vessels,” making it possible to study the biochemical and physiological consequences of the actions of one transgene product (ie, GRK) on another (ie, α1BAR). This study has demonstrated the power of this simple crossbreeding strategy that can be used to address questions regarding GRK specificity on other receptors or for dissecting individual phenotypes of other Tg models. Importantly in this study, expression levels of α1BARs and individual GRKs driven by the same αMyHC promoter, when expressed concomitantly in the hybrid mice, did not differ from the levels of overexpression seen in the individual Tg parental lines. This is an important finding, given that promoter competition might be expected to occur, limiting the overexpression of 1 or both of the transgenes, and thus limiting the usefulness of this strategy. Furthermore, our previous studies of Tg mice using the αMyHC promoter have demonstrated that transgene expression was homogenous in nature throughout the heart8 16 ; thus, both transgenes should be expressed in a similar fashion when together.
Using these hybrid Tg mice in this study allowed us to investigate the in vivo specificity of GRK2 (also known as βARK1), GRK3 (βARK2), and GRK5 on α1BARs in the heart. This is an area of study concerning the 6-member GRK family, for which definitive information has been lacking. A majority of in vitro experiments using heterologous cell culture expression systems with a variety of overexpressed G protein–coupled receptors have shown limited substrate selectivity for these 3 ubiquitously expressed GRKs.10 11 12 13 14 Importantly in the present study, we have revealed that these 3 GRKs differ in their ability to desensitize α1ARs in vivo in the heart. Our study differs from previous in vitro findings regarding the α1BAR using heterologous cell culture expression systems, in which overexpression of either GRK2 or GRK3 promoted α1BAR desensitization via agonist-induced phosphorylation.12 Our results demonstrate that GRK2 is ineffective in regulating either the overexpressed CAMα1BAR or the WTα1BAR, and more importantly, signaling through endogenous myocardial α1ARs (of which the predominant subtype in the mouse heart is the α1B1 ) was also unaltered by GRK2 overexpression. Interestingly, signaling through each type of α1BAR in the different hybrid Tg mice was significantly attenuated by GRK3 overexpression demonstrating an in vivo selectivity between GRK2 and GRK3 in the heart. Similar to Diviani et al,12 we found variable effects of GRK5 on agonist-mediated desensitization of myocardial α1BARs. Importantly, levels of GRK overexpression, as compared with total endogenous NLC GRK activity, between the different Tg mice could not explain the lack of GRK2 effect, given that neither ≈3-fold GRK2 overexpression nor ≈20-fold GRK2 overexpression was sufficient to attenuate in vivo α1BAR signaling, whereas ≈5-fold GRK3 overexpression (≈12-fold overexpression when compared with endogenous GRK3 levels) effectively abrogated all examined forms of α1BAR signaling and eliminated the myocardial hypertrophy induced by CAMα1BAR expression. Recently, it was shown by Iacovelli et al15 in FRTL-5 cells stably overexpressing GRK2 that endogenous signaling through α1BAR was not attenuated, whereas signaling through endogenous thyrotropin receptors was. Thus, there are now data to suggest that even in certain cell types in vitro, GRK2 does not desensitize the α1BAR, which is consistent with our in vivo findings in the heart.
Despite their initial characterization as highly homologous isozymes, a pattern of in vivo differences between GRK2 and GRK3 is now emerging. In addition to our current findings concerning α1BAR signaling, we have demonstrated previously that GRK3 does not, whereas GRK2 does, regulate endogenous βAR signaling in the heart.18 As is the case with α1BAR signaling, these in vivo βAR findings differ from in vitro results that demonstrate that both GRK2 and GRK3 phosphorylate and desensitize β1ARs, the primary βAR subtype expressed in the myocardium.11 Furthermore, studies with TgGRK3 animals reveal GRK specificity for the endogenous myocardial thrombin receptor.18 The thrombin receptor is one receptor in which a distinct difference in GRK-mediated desensitization has been shown to exist in vitro between GRK2 and GRK3.26 Differences in the in vivo myocardial regulation of angiotensin II receptors have also been seen in TgGRK3 and TgGRK2 mice.17 18 Thus, the use of Tg technology has begun to clarify GRK specificity in the in vivo heart. Collectively, these data suggest that GRK2 and GRK3 have distinct substrates in the intact heart and do not exhibit redundancy in the normal regulation of myocardial function.
The regulation of GRK2 and GRK3 may provide insight into why these highly homologous kinases differ in their in vivo substrate selectivity in the heart. Both GRK2 and GRK3 are cytosolic enzymes that undergo a membrane-targeting event before their phosphorylation of agonist-occupied receptors.9 10 For these 2 GRKs, this is accomplished by a specific protein-protein interaction between the carboxyl-terminal domain of GRK2 and GRK3 and Gβγ subunits released from activated heterotrimeric G proteins.27 28 Interestingly, the most divergent region between GRK2 and GRK3 is within the mapped Gβγ binding domain.28 29 30 This may indicate differential affinities of these 2 GRKs for Gβγ subunits, and the availability of Gβγs may be influenced by cell type and by the G protein–coupled receptor activated. Thus, Gβγs released after myocardial α1BAR stimulation may bind to GRK3 with higher affinity than that for GRK2, and vice versa for Gβγs released by β1ARs in vivo. Importantly, in vitro evidence supports the notion that GRK2 and GRK3 can be targeted to membranes in a receptor- and Gβγ-dependent manner.31
Although the in vivo G protein–coupled receptor specificity for GRK2 and GRK3 is becoming clearer, the understanding of in vivo GRK5 actions in the heart are more complex. It has been shown that in contrast to GRK2 overexpression, myocardial GRK5 overexpression, like that of GRK3, does not alter in vivo angiotensin II signaling.17 However, like GRK2 overexpression, βAR signaling in vivo was significantly attenuated in the hearts of TgGRK5 mice.17 The present results reveal conflicting findings regarding the in vivo actions of GRK5 on myocardial α1BAR signaling. Whereas in vivo GRK5 attenuated CAMα1BAR-induced hypertrophy and endogenous α1BAR agonist-stimulated ERK2 activity, GRK5 was incapable of inhibiting overexpressed CAMα1BAR- or WT α1BAR-induced DAG and TgWTα1BAR-induced ANF levels, and only partially attenuated ANF levels in the TgCAMα1BAR hearts and JNK1 activity in the hearts expressing endogenous α1BARs. These results suggest that DAG signaling due to α1BAR stimulation does not result in hypertrophy and that ventricular ANF expression, although concomitant with it, is not sufficient for hypertrophy. In fact, recent studies associate ANF with inhibition of proliferation of nonmyocardial cells and antihypertrophic effects in cardiomyocytes.32 33 34 35 In contrast, it appears that MAPK activation in response to α1BAR-Gq stimulation may be important for hypertrophy. In fact, these findings may shed light on the specific signaling pathways responsible for the progression of α1BAR activation to myocardial hypertrophy, although this remains to be determined.
G protein–coupled receptors play integral roles in cardiac function, and examination of their regulation by GRKs is important for understanding cardiac homeostasis and the regulation of compensation during disease states. For example, heart failure is associated with a constellation of changes, including increases in circulating catecholamines, decreases in βAR density, and increases in GRK2.36 Additionally, α1AR density has been shown to be elevated when heart failure develops.37 Therefore, in disease states when βARs are downregulated, α1ARs become a more predominant population of the myocardial ARs. Importantly, as the findings regarding in vivo GRK substrate specificity in this report reveal, potentially increased α1AR signaling in the compromised heart would be insensitive to the elevated GRK2 levels associated with heart disease. This may provide a mechanism that attempts to maintain cardiac output in response to catecholamines. This lack of GRK2 effect on α1ARs could lead to enhanced Gq signaling responsible for the initial adaptive hypertrophy response in the compromised heart. Testing these hypotheses will be the subject of future studies.
In summary, results from this study definitively demonstrate that GRK3 desensitizes α1BAR-mediated signaling in vivo, whereas the highly homologous GRK2 had no effect on signaling through this G protein–coupled receptor. These results illustrate that although in vitro studies set the foundation for understanding receptor-kinase interactions, in vivo studies are required to fully elucidate in vivo selectivity. Furthermore, it is becoming clear from Tg studies that GRK2, GRK3, and GRK5 play distinct roles in the normal regulation of myocardial signaling and function.
This work was supported in part by National Institutes of Health Grants HL-16037 (to R.J.L.), HL-61690 and HL-59533 (to W.J.K.), GM44944 (to J.L.B.) and by a Grant-in-Aid from the North Carolina Affiliate of the American Heart Association (to W.J.K.). R.J.L. is an investigator of the Howard Hughes Medical Institute. J.L.B. is an Established Investigator of the American Heart Association.
- Received October 6, 1999.
- Accepted October 20, 1999.
- © 2000 American Heart Association, Inc.
Rokosh DG, Stewart AF, Chang KC, Bailey BA, Karliner JS, Camacho SA, Long CS, Simpson PC. α1-Adrenergic receptor subtype mRNAs are differentially regulated by α1-adrenergic and other hypertrophic stimuli in cardiac myocytes in culture and in vivo. J Biol Chem. 1996;271:5839–5843.
Puceat M, Hilal-Dandan R, Strulovici B, Brunton LL, Brown JH. Differential regulation of protein kinase C isoforms in isolated neonatal and adult rat cardiomyocytes. J Biol Chem. 1994;269:16938–16944.
Akhter SA, Luttrell LM, Rockman HA, Lefkowitz RJ, Koch WJ. Targeting the receptor-Gq interface to inhibit in vivo pressure overload myocardial hypertrophy. Science. 1998;280:574–577.
Chien KR, Knowlton KU, Zhu H, Chien S. Regulation of cardiac gene expression during myocardial growth and hypertrophy: molecular studies of an adaptive physiologic response. FASEB J. 1991;5:3037–3046.
Simpson PC. Transcription of early development isogenes in cardiac myocyte hypertrophy. J Mol Cell Cardiol. 1983;72:732–738.
Milano CA, Dolber PC, Rockman HA, Bond RA, Venable ME, Allen LF, Lefkowitz RJ. Myocardial expression of a constitutively active α1B-adrenergic receptor in transgenic mice induces cardiac hypertrophy. Proc Natl Acad Sci U S A. 1994;91:10109–10113.
Lefkowitz RJ. G protein-coupled receptors, III: new roles for receptor kinases and β-arrestins in receptor signaling and desensitization. J Biol Chem. 1998;273:18677–18680.
Inglese J, Freedman NJ, Koch WJ, Lefkowitz RJ. Structure and mechanisms of the G protein-coupled receptor kinase. J Biol Chem. 1993;268:23735–23738.
Freedman NJ, Liggett SB, Drachman DE, Pei G, Caron MG, Lefkowitz RJ. Phosphorylation and desensitization of the human β1-adrenergic receptor: involvement of G protein-coupled receptor kinases and cAMP-dependent protein kinase. J Biol Chem. 1995;270:17953–17961.
Diviani D, Lattion A-L, Larbi N, Kunapuli P, Pronin A, Benovic JL, Cotecchia S. Effect of different G protein-coupled receptor kinases on phosphorylation and desensitization of the α1B-adrenergic receptor. J Biol Chem. 1996;271:5049–5058.
Oppermann M, Freedman NJ, Alexander RW, Lefkowitz RJ. Phosphorylation of the type 1A angiotensin II receptor by G protein-coupled receptor kinases and protein kinase C. J Biol Chem. 1996;271:13266–13272.
Freedman NJ, Ament AS, Oppermann M, Stoffel RH, Exum ST, Lefkowitz RJ. Phosphorylation and desensitization of human endothelin A and B receptors: evidence for G protein-coupled receptor kinase specificity. J Biol Chem. 1997;272:17734–17743.
Iacovelli L, Franchetti R, Grisolia D, DeBlasi A. Selective regulation of G protein-coupled receptor-mediated signaling by G protein-coupled receptor kinase 2 in FRTL-5 cells: analysis of thyrotropin, α1B-adrenergic, and A1 adenosine receptor-mediated responses. Mol Pharmacol. 1999;56:316–324.
Koch WJ, Rockman HA, Samama P, Hamilton R, Bond RA, Milano CA, Lefkowitz RJ. Cardiac function in mice overexpressing the β-adrenergic receptor kinase or a βARK inhibitor. Science. 1995;268:1350–1353.
Rockman HA, Choi D-J, Rahman NU, Akhter SA, Lefkowitz RJ, Koch WJ. Receptor-specific in vivo desensitization by the G protein-coupled receptor kinase-5 in transgenic mice. Proc Natl Acad Sci U S A. 1996;93:9954–9959.
Iaccarino G, Rockman HA, Shotwell KF, Tomhave ED, Koch WJ. Myocardial overexpression of GRK3 in transgenic mice: evidence for in vivo selectivity of GRKs. Am J Physiol. 1998;275:H1298–H1306.
Koch WJ, Milano CA, Lefkowitz RJ. Transgenic manipulation of myocardial G protein-coupled receptors and receptor kinases. Circ Res. 1996;78:511–516.
Akhter SA, Milano CA, Shotwell KF, Cho M-C, Rockman HA, Lefkowitz RJ, Koch WJ. Transgenic mice with cardiac overexpression of α1B-adrenergic receptors: in vivo α1-adrenergic receptor-mediated regulation of β-adrenergic signaling. J Biol Chem. 1997;272:21253–21259.
Iaccarino G, Dolber PC, Lefkowitz RJ, Koch WJ. β-Adrenergic receptor kinase-1 levels in catecholamine-induced myocardial hypertrophy: regulation by β- but not α1-adrenergic stimulation. Hypertension. 1999;33:396–401.
Choi DJ, Koch WJ, Hunter JJ, Rockman HA. Mechanism of β-adrenergic receptor desensitization in cardiac hypertrophy is increased β-adrenergic receptor kinase. J Biol Chem. 1997;272:17223–17229.
Thorburn J, Frost JA, Thorburn A. Mitogen-activated protein kinases mediate changes in gene expression, but not cytoskeletal organization associated with cardiac muscle cell hypertrophy. J Cell Biol. 1994;126:1565–1572.
Bogoyevitch MA, Andersson MB, Gillespie-Brown J, Clerk A, Glennon PE, Fuller SJ, Sugden PH. Adrenergic receptor stimulation of the mitogen-activated protein kinase cascade and cardiac hypertrophy. Biochem J. 1996;314:115–121.
Glennon PE, Kaddoura S, Sale EM, Sale GJ, Fuller SJ, Sugden PH. Depletion of mitogen-activated protein kinase using an antisense oligodeoxynucleotide approach downregulates the phenylephrine-induced hypertrophic response in rat cardiac myocytes. Circ Res. 1996;78:954–961.
Ishii K, Chen J, Ishii M, Koch WJ, Freedman NJ, Lefkowitz RJ, Coughlin SR. Inhibition of thrombin receptor signaling by a G-protein coupled receptor kinase: functional specificity among G-protein coupled receptor kinases. J Biol Chem. 1994;269:1125–1130.
Pitcher JA, Inglese J, Higgins JB, Arriza JL, Casey PJ, Kim C, Benovic JL, Kwatra MM, Caron MG, Lefkowitz RJ. Role of βγ subunits of G proteins in targeting the β-adrenergic receptor kinase to membrane-bound receptors. Science. 1992;257:1264–1267.
Koch WJ, Inglese J, Stone WC, Lefkowitz RJ. The binding site for the βγ subunits of heterotrimeric G proteins on the β-adrenergic receptor kinase. J Biol Chem. 1993;268:8256–8260.
Chuang TT, Pompili E, Paolucci L, Sallese M, DeGioia L, Salmona M, DeBlasi A. Identification of a short sequence highly divergent between β-adrenergic receptor kinases 1 and 2 that determines the affinity of binding to βγ subunits of heterotrimeric guanine-nucleotide-binding regulatory proteins. Eur J Biochem. 1997;245:533–540.
Touhara K, Koch WJ, Hawes BE, Lefkowitz RJ. Mutational analysis of the plekstrin homology domain of the β-adrenergic receptor kinase: differential effects on Gβγ and phosphatidylinositol 4,5-bisphosphate binding. J Biol Chem. 1995;270:17000–17005.
Daaka Y, Pitcher JA, Richardson M, Stoffel RH, Robishaw JD, Lefkowitz RJ. Receptor and Gβγ isoform-specific interactions with G protein-coupled receptor kinases. Proc Natl Acad Sci U S A. 1997;94:2180–2185.
Cao L, Gardner DG. Natriuretic peptides inhibit DNA synthesis in cardiac fibroblasts. Hypertension. 1995;25:227–234.
Itoh H, Pratt RE, Dzau VJ. Atrial natriuretic polypeptide inhibits hypertrophy of vascular smooth muscle cells. J Clin Invest. 1990;86:1690–1697.
Silberbach M, Gorenc T, Hershberger RE, Stork PJ, Steyger PS, Roberts CT Jr. Extracellular signal-regulated protein kinase activation is required for the anti-hypertrophic effect of atrial natriuretic factor in neonatal rat ventricular myocytes. J Biol Chem. 1999;274:24858–24864.
Lohse MJ, Engelhardt S, Danner S, Bohm M. Mechanisms of β-adrenergic receptor desensitization: from molecular biology to heart failure. Basic Res Cardiol. 1996;91:29–34.
Kagiya T, Hori M, Iwakura K, Iwai K, Watanabe Y, Uchida H, Kitabake A, Inoul M, Kamada T. Role of increased α1-adrenergic activity in cardiomyopathic Syrian hamster. Am J Physiol. 1991;260:H80–H88.