Cincinnati Children’s Hospital Medical Center, University of Cincinnati, Cincinnati, Ohio, USA.
Address correspondence to: Jeffery D. Molkentin, Division of Molecular Cardiovascular Biology, Cincinnati Children’s Hospital Medical Center, 3333 Burnet Avenue, Cincinnati, Ohio 45229, USA. Phone: (513) 636-3557; Fax: (513) 636-5958. E-mail: email@example.com.
Published March 1, 2006 - More info
Previous work showed that calmodulin (CaM) and Ca2+-CaM–dependent protein kinase II (CaMKII) are somehow involved in cardiac hypertrophic signaling, that inositol 1,4,5-trisphosphate receptors (InsP3Rs) in ventricular myocytes are mainly in the nuclear envelope, where they associate with CaMKII, and that class II histone deacetylases (e.g., HDAC5) suppress hypertrophic gene transcription. Furthermore, HDAC phosphorylation in response to neurohumoral stimuli that induce hypertrophy, such as endothelin-1 (ET-1), activates HDAC nuclear export, thereby regulating cardiac myocyte transcription. Here we demonstrate a detailed mechanistic convergence of these 3 issues in adult ventricular myocytes. We show that ET-1, which activates plasmalemmal G protein–coupled receptors and InsP3 production, elicits local nuclear envelope Ca2+ release via InsP3R. This local Ca2+ release activates nuclear CaMKII, which triggers HDAC5 phosphorylation and nuclear export (derepressing transcription). Remarkably, this Ca2+-dependent pathway cannot be activated by the global Ca2+ transients that cause contraction at each heartbeat. This novel local Ca2+ signaling in excitation-transcription coupling is analogous to but separate (and insulated) from that involved in excitation-contraction coupling. Thus, myocytes can distinguish simultaneous local and global Ca2+ signals involved in contractile activation from those targeting gene expression.
Xu Wu, Tong Zhang, Julie Bossuyt, Xiaodong Li, Timothy A. McKinsey, John R. Dedman, Eric N. Olson, Ju Chen, Joan Heller Brown, Donald M. Bers
Ca2+ plays a pivotal role in both excitation-contraction coupling (ECC) and activation of Ca2+-dependent signaling pathways. One of the remaining questions in cardiac biology is how Ca2+-dependent signaling pathways are regulated under conditions of continual Ca2+ transients that mediate cardiac contraction during each heartbeat. Ca2+-calmodulin–dependent protein kinase II (CaMKII) activation and its ability to regulate histone deacetylase 5 (HDAC5) nuclear shuttling represent a critical Ca2+-dependent signaling circuit for controlling cardiac hypertrophy and heart failure, yet the mechanism of activation by Ca2+ is not known. In this issue of the JCI, Wu et al. convincingly demonstrate that the inositol 1,4,5-trisphosphate receptor (InsP3R) is involved in local control of Ca2+ for activating CaMKII in the nuclear envelope of adult ventricular cardiac myocytes (see the related article beginning on page 675). The overall paradigm that is demonstrated is the best example of a molecular mechanism whereby signaling is directly regulated by a local Ca2+ pool that is disparate or geometrically insensitive to cytosolic Ca2+ underlying each contractile cycle.
The physiologic relevance of “reactive” Ca2+ signaling in the heart is uncertain, given the highly specialized manner in which Ca2+ cycling is tightly controlled through the membrane compartments/domains, channels, and pumps that underlie excitation-contraction coupling (ECC). For example, in response to depolarization of the sarcolemma, Ca2+ enters through the voltage-dependent L-type Ca2+ channel, which directly stimulates adjacent ryanodine receptors embedded within the sarcoplasmic reticulum (SR). This priming Ca2+ from the L-type channel induces a much larger release of Ca2+ from ryanodine receptors that increase intracellular Ca2+ concentration by more than 10-fold to induce contraction. During diastole, Ca2+ is removed from the cytoplasm by resequestration back into the SR through the action of SR/ER Ca2+-ATPase (SERCA) as well as extrusion from the cell through the action of the Na+/Ca2+ exchanger within the sarcolemma (Figure 1). Thus, ECC-mediated Ca2+ cycling accounts for nearly all routinely detectable Ca2+ alterations that occur within a cardiomyocyte, which makes it difficult to explain how Ca2+-activated signaling proteins function in this background. Also to be considered, it would make little sense to regulate Ca2+-dependent signaling proteins through changes in Ca2+ cycling associated with ECC since such a mechanism would not afford a disconnection between inotropy and signaling. Indeed, near maximal Ca2+ cycling due to phospholamban deletion in mice does not induce cardiac hypertrophy or otherwise predispose the heart to dysfunction that would be associated with activation of reactive signaling pathways, indicating that inotropic drive is not a direct source of Ca2+ for intracellular signaling pathways (1). Thus, specialized pools of Ca2+ that are location specific or somehow buffered from cytoplasmic Ca2+ need to be evoked to account for the regulation of Ca2+-sensitive signaling proteins, such as Ca2+-calmodulin–dependent protein kinase (CaMK), calcineurin, and PKC. However, the existence of microdomains or region-specific buffering of Ca2+ to control signaling proteins has not been convincingly demonstrated to date in cardiac myocytes. In this issue of the JCI, Wu et al. report on their employment of a series of pharmacologic, molecular, and genetic manipulations to convincingly demonstrate that inositol 1,4,5-trisphosphate receptor–based (InsP3R-based) alterations in Ca2+ levels regulate CaMKII activation and its ability to control histone deacetylase 5 (HDAC5) translocation, thus suggesting a Ca2+-dependent reactive signaling circuit that is controlled outside of ECC-based Ca2+ cycling (2).
Schematic of potential Ca2+ sources that might be specialized to regulate reactive signaling pathways in cardiac myocytes. (i) Some L-type Ca2+ channels (ICa,L) are not associated with the junctional complex and hence could be involved in providing a local Ca2+ signal in specific membrane-associated compartments. (ii) T-type Ca2+ channels (ICa,T) are reexpressed in hypertrophic states where they could provide Ca2+ in specific microenvironments associated with the sarcolemma to affect reactive signaling pathways. (iii) Capacitative or store-operated Ca2+ entry through transient receptor potential (TRP) channels, alone or in conjunction with (iv) InsP3R-mediated release of Ca2+ from the ER/nuclear envelope, could also provide a highly localized Ca2+ pool for controlling reactive signaling pathways in cardiac myocytes. Signaling from G protein–coupled receptors (GPCRs) activates PLC and generates InsP3, causing a perinuclear Ca2+ signal through the InsP3R, resulting in CamKII activation and HDAC5 nuclear export, as proposed by Wu et al. (2). IP3, InsP3; IP3R, InsP3R; PLN, phospholamban; NCX, Na+/Ca2+ exchanger; RyR, ryanodine receptor; CaM, calmodulin; SERCA, SR/ER Ca2+-ATPase; P, phosphate.
In response to disease-causing stimuli, the myocardium is directly influenced by neuroendocrine secreted growth factors and/or cytokines that induce ventricular remodeling, hypertrophic enlargement of myocytes, and alterations in the viability of myocytes. Many of these neuroendocrine factors (e.g., angiotensin II and endothelin-1) signal through G protein–coupled receptors on cardiac myocytes to induce phospholipase C (PLC) activation, which in turn generates InsP3 and diacylglycerol (DAG) (3). In most cell types, InsP3 generation in turn leads to release of Ca2+ from the endoplasmic reticulum through the InsP3R channel, although it is uncertain whether InsP3R activity significantly regulates Ca2+ release in adult ventricular cardiac myocytes. It is known that the heart generates InsP3 and that the type 2 InsP3R (InsP3R2) is present and localized around the nucleus of ventricular myocytes while atrial myocytes also likely contain InsP3Rs in the junctional SR where they can affect ECC (4, 5).
Neuroendocrine factors and cytokines promote activation of a number of intracellular signaling pathways in cardiac myocytes, including MAPK, calcineurin, PKC, CaMK, and IGF-1 pathway constituents (3). Three of these signaling factors, calcineurin, CaMK, and PKCα, -β, and -γ, require increases in Ca2+ to become activated although, as discussed above, the source of such Ca2+ in cardiac myocytes has yet to be determined. In most other cell types, InsP3R-regulated Ca2+ mobilization in association with store-operated Ca2+ entry is required for activation of Ca2+-dependent signaling effectors (6). The importance of understanding such regulation in cardiac myocytes is underscored by the observation that both calcineurin and CaMK are potent inducers of the hypertrophic response (7–9). Hence, careful elucidation of the mechanism whereby calcineurin and CaMK become activated in cardiac myocytes should suggest the pool(s) of Ca2+ that function in reactive signaling.
The cardiac myocyte contains an organized array of sarcolemmal invaginations referred to as the T tubule network that closely apposes the SR as a means of distributing membrane depolarization and coordinated Ca2+ release throughout the cytoplasm. The complexity of this T tubule network as well as regions of the SR/ER that includes the nuclear envelope could easily accommodate more specialized Ca2+ regulatory domains that function outside of ECC and serve a signaling function. At least 4 different regulatory mechanisms for compartmentalizing Ca2+ outside of ECC have been hypothesized. One possible source of reactive Ca2+ is the voltage-regulated L-type Ca2+ channel itself, which normally triggers Ca2+-induced Ca2+ release and contraction. However, a percentage of L-type Ca2+ channels are not associated with the junctional complex and Ca2+-induced Ca2+ release, but instead could be dedicated to special membrane domains, such as lipid rafts to mediate a signaling function (Figure 1, i). Indeed, overexpression of the pore-forming subunit of the L-type Ca2+ channel in the hearts of transgenic mice led to late-onset cardiac hypertrophy (10). Another potential source of Ca2+ for reactive signaling is the voltage-dependent current mediated by T-type Ca2+ channels (Figure 1, ii). T-type channels are not normally expressed in mature ventricular myocytes, but expression is present in immature myocytes as well as in adult ventricular myocytes from hypertrophied and failing hearts (11–13). In other cell types, T-type currents play important roles in regulating reactive signaling and cellular proliferation, and more interestingly, these channels are associated with specialized membrane structures such as lipid rafts (14). A third possible source of reactive Ca2+ in myocytes is associated with members of the transient receptor potential channel that are thought to mediate store-operated Ca2+ entry and are expressed in the heart (15) (Figure 1, iii). Indeed, previous work has suggested that neonatal and/or adult rat cardiac myocytes are capable of store-operated Ca2+ entry once depleted with PLC-dependent agonists (16–18). Moreover, Hunton and colleagues showed that general inhibitors of store-operated Ca2+ entry could reduce calcineurin activation in cardiac myocytes (17). Indeed, in nonmyocytes, calcineurin signaling is prominently regulated by store-operated Ca2+ entry (19). In general, store-operated Ca2+ entry is associated or coupled with PLC-dependent InsP3R Ca2+ release. Thus, the fourth potential source of Ca2+ for reactive signaling is InsP3R dependent (Figure 1, iv). In adult ventricular cardiac myocytes, InsP3R2 is expressed and primarily localized to the nuclear envelope (5). Interestingly, InsP3R2 is physically associated with CaMKII and thus could serve a more specialized role in providing a local pool of Ca2+ to activate this kinase in or around the nucleus (Figure 1, iv).
In this issue of the JCI, Wu et al. provide convincing evidence that Ca2+ from an InsP3R-dependent store regulates activation of HDAC5 nuclear export through CaMKII in adult ventricular cardiomyocytes (2). HDAC4/5 activity and nuclear occupancy are directly regulated by CaMK-, PKC-, and protein kinase D–mediated (PKD-mediated) phosphorylation (20, 21). That this regulatory relationship is central to the cardiac hypertrophic response is consistent with the observation that Hdac9- and Hdac5-null mice each develop exaggerated hypertrophy following pressure overload or when crossed with the calcineurin transgene (22, 23). Here, the authors demonstrate that the G protein–coupled agonist (PLC-activating) endothelin-1 induced export of HDAC5 from adult cardiac myocytes without affecting total cytosolic Ca2+ concentration and that this export was blocked with an InsP3R inhibitor or stimulated directly with an InsP3R agonist (2). More convincingly, HDAC5 nuclear export after endothelin-1 stimulation is completely absent in adult cardiac myocytes from Ip3r2 gene–targeted mice. These results unequivocally demonstrate that HDAC5 regulation depends in part on InsP3R signaling. Assessment of Ca2+ levels within the cell using Fluo-5N suggested that an InsP3 signal induces Ca2+ mobilization in the perinuclear area where the InsP3R2 is localized in the nuclear envelope. Thus, HDAC5 nuclear export, which permits gene activation and induction of the hypertrophic response, can be directly regulated by an InsP3-dependent pathway through a region-specific pool of Ca2+ within or near the nuclear envelope. This movement of HDAC5 in association with InsP3R-dependent Ca2+ mobilization was partially mediated by CaMK activation since pharmacologic inhibition blocked approximately half of the HDAC5 nuclear export. Thus, the current work by Wu et al. establishes at least 1 paradigm whereby a specialized pool of Ca2+ can regulate a hypertrophic signaling circuit within a ventricular cardiac myocyte independent of contractile Ca2+.
As with most landmark studies, Wu et al. raises a number of important issues. First, it is interesting that pacing of adult cardiac myocytes up to 2 Hz did not cause HDAC5 nuclear export, suggesting that ECC-mediated Ca2+ fluxing has little effect on this regulatory pathway (2). In contrast, pacing of skeletal muscle myotubes enhanced Ca2+ transients enough to induce nuclear export of HDAC4 as well as nuclear factor of activated T cells (NFAT), a sensor of calcineurin signaling (24, 25). These later observations suggest that contractile Ca2+ can set in motion 2 separate Ca2+-dependent signaling pathways in skeletal muscle myotubes although this paradigm does not seem to apply to adult cardiac myocytes. Indeed, calcineurin does not even regulate hypertrophy in skeletal muscle as it does in heart muscle, demonstrating profound differences in signaling between these tissues (26). Hence, cardiac myocytes appear to be unique in their ability to compartmentalize or control Ca2+ such that the ECC-based pool is distinct from the pool that regulates reactive signaling.
One such Ca2+ microenvironment in the adult ventricular cardiac myocytes appears to be dependent on the InsP3R localized to the perinuclear region. Indeed, CaMKII is directly complexed with the InsP3R in cardiac myocytes, suggesting a mechanism whereby CaMKII might only respond to exceedingly high Ca2+ levels in the microenvironment of the InsP3R and hence be insensitive to total intracellular Ca2+ levels that cannot reach such a putative threshold. For example, such a microenvironment exists at the junction between the sarcolemma and SR, where local Ca2+ regulates the independent activation of each ryanodine receptor (27). While Wu et al. only evaluated HDAC5 nuclear export as regulated by CaMKII (2), this paradigm should be evaluated for other signaling circuits that are controlled by Ca2+, such as other class II HDACs, calcineurin, and/or PKC isoforms. Similarly, it would also be interesting to determine if CaMKII and HDAC5 can be regulated by other potential microdomains of Ca2+ within the cardiac myocyte so that a hierarchy or specialization of Ca2+ microdomains might be established. Thus, the results of Wu et al. have not only established 1 Ca2+-dependent regulatory paradigm, but they have also provided the conceptual framework for further parsing contractile versus reactive signaling Ca2+ in the heart.
Nonstandard abbreviations used: CaMKII, Ca2+-calmodulin–dependent protein kinase II; ECC, excitation-contraction coupling; HDAC, histone deacetylase; InsP3, inositol 1,4,5-trisphosphate; InsP3R, InsP3 receptor; PLC, phospholipase C; SR, sarcoplasmic reticulum.
Conflict of interest: The author has declared that no conflict of interest exists.
Reference information: J. Clin. Invest.116:623–626 (2006). doi:10.1172/JCI27824.
See the related article beginning on page 675.