Ca²⁺ Influx Through T- and L-Type Ca²⁺ Channels Have Different Effects on Myocyte Contractility and Induce Unique Cardiac Phenotypes

Abstract

T-type Ca²⁺ channels (TTCCs) are expressed in the developing heart, are not present in the adult ventricle, and are reexpressed in cardiac diseases involving cardiac dysfunction and premature, arrhythmogenic death. The goal of this study was to determine the functional role of increased Ca²⁺ influx through reexpressed TTCCs in the adult heart. A mouse line with cardiac-specific, conditional expression of the α1G-TTCC was used to increase Ca²⁺ influx through TTCCs. α1G hearts had mild increases in contractility but no cardiac histopathology or premature death. This contrasts with the pathological phenotype of a previously studied mouse with increased Ca²⁺ influx through the L-type Ca²⁺ channel (LTCC) secondary to overexpression of its β2a subunit. Although α1G and β2a myocytes had similar increases in Ca²⁺ influx, α1G myocytes had smaller increases in contraction magnitude, and, unlike β2a myocytes, there were no increases in sarcoplasmic reticulum Ca²⁺ loading. Ca²⁺ influx through TTCCs also did not induce normal sarcoplasmic reticulum Ca²⁺ release. α1G myocytes had changes in LTCC, SERCA2a, and phospholamban abundance, which appear to be adaptations that help maintain Ca²⁺ homeostasis. Immunostaining suggested that the majority of α1G-TTCCs were on the surface membrane. Osmotic shock, which selectively eliminates T-tubules, induced a greater reduction in L-versus TTCC currents. These studies suggest that T- and LTCCs are in different portions of the sarcolemma (surface membrane versus T-tubules) and that Ca²⁺ influx through these channels induce different effects on myocyte contractility and lead to distinct cardiac phenotypes.

Ca²⁺ influx through voltage-regulated ion channels is essential for initiating and regulating cardiac function.^1,2 Changes in cytosolic [Ca²⁺] also influence a variety of signaling pathways,^3,4 regulate normal cardiac metabolism,⁵ and cause both physiological and pathological hypertrophy.^6-9 The primary pathway for Ca²⁺ influx in the adult heart is via the L-type Ca²⁺ channel (LTCC). This Ca²⁺ influx pathway is essential for triggering sarcoplasmic reticulum (SR) Ca²⁺ release and is the major source of Ca²⁺ to load the SR.¹⁰ Excess Ca²⁺ influx through LTCCs has been linked to necrotic myocyte death, cardiac dysfunction, and premature death.¹¹

Ca²⁺ can also enter cardiac myocytes through the voltage-operated T-type Ca²⁺ channel (TTCC). These channels are expressed throughout cardiac development until the end of the fetal/neonatal period.^12-14 TTCC expression decreases soon after birth, with little or no expression in normal adult ventricular myocytes (VMs). TTCCs are reexpressed when the heart is subjected to pathological stressors that induce cardiac hypertrophy and failure^15-19 and are associated with depressed cardiac function and sudden death.²⁰ However, the functional significance of TTCC reexpression in adult VMs has not been established and is the focus of this study. Our major goal was to determine whether Ca²⁺ influx through T-and L-type Ca²⁺ channels induce different cardiac phenotypes.

Three TTCC genes have been identified, but only 2, Ca_v3.1 (α1G) and Ca_v3.2 (α1H), have been found in the heart.^12,21,22 All TTCCs have similar biophysical properties, including activation at more negative voltages and faster inactivation than LTCCs. Based on studies in Purkinje cells and guinea pig VMs, TTCC current (I_Ca,T) appears to be less efficient than LTCC current (I_Ca,L) in excitation–contraction (EC) coupling.^23,24 Electrophysiological assessment of I_Ca,T is technically difficult because of its small amplitude. Additionally, it is difficult to extrapolate findings in Purkinje cells to VMs because of the limited T-tubule and ryanodine receptor (RyR) organization of Purkinje cells.²⁵

α1G and α 1H are reexpressed in the adult ventricle under pathological stress,^15-19 but the involvement of these channels in cardiac dysfunction, pathological remodeling and arrhythmias is not known. To understand the function of these channels in the myocardium, we generated mice with inducible, cardiac-specific expression of the α1G-TTCC. Increased Ca²⁺ influx through TTCCs should augment myocyte contractility and activate Ca²⁺ regulated signaling pathways. Recently, we showed that increasing Ca²⁺ influx through LTCCs, by overexpressing its β2a subunit, increased myocyte contractility but eventually induced SR Ca²⁺ overload and Ca²⁺-mediated myocyte necrosis, leading to heart failure and premature death.¹¹ In the present study, we asked whether excess Ca²⁺ influx via TTCCs also causes alterations in contractility, cardiac dysfunction, and premature death.

Our experiments show that mice with cardiac-specific α1G-TTCC expression have increased Ca²⁺ influx through functional TTCCs. Hearts from α1G animals were mildly hypercontractile, but there was no cardiac pathology or premature death. Myocytes isolated from α1G hearts had increased Ca²⁺ influx, contractions, and systolic [Ca²⁺]_i transients, but the SR Ca²⁺ load was not increased. The duration of the [Ca²⁺]_i transient and contractions were shortened, secondary to what appears to be adaptive changes in Ca²⁺ regulatory proteins. Experiments with electrophysiological and immunostaining approaches suggest that α1G-TTCCs are more highly concentrated in the surface membrane than in T-tubules and their associated junctional SR. These findings explain why I_Ca,T is an inefficient trigger for SR Ca²⁺ release and an inefficient source of Ca²⁺ to load the SR. These studies also show that unlike LTCCs, Ca²⁺ influx through TTCCs does not induce Ca²⁺-mediated cardiac pathologies or arrhythmogenic sudden death. Taken together, our new findings suggest that the source of the Ca²⁺ influx pathway can be a critical determinant of the induced cardiac phenotype.

Materials and Methods

All experiments involving animals were approved by the Institutional Animal Care and Use Committee of Temple University.

Mice

Transgenic mice with conditional and cardiac-specific expression of mouse α1G subunit or rat β2a subunit were generated using a modified murine α-myosin heavy chain (α-MHC) promoter expression vector.^11,26 Animals between 10 to 16 weeks of age were used for experiments.

Echocardiography

Whole heart morphology and function were assessed via echocardiography. Hearts were viewed in short-axis and analyzed in M-mode.

Histology

Hearts from animals (9 months old) were fixed in formalin, and longitudinal sections were evaluated for gross morphology and fibrosis.

Western Blot Analysis

Abundance/phosphorylation status of Ca²⁺ regulatory proteins in ventricular tissue was assessed as described previously.²⁷

Electrophysiology

[Ca²⁺]_i transients (Fluo-3 epifluorescence), fractional shortening, and SR Ca²⁺ content of isolated VMs were assessed at 0.5 Hz, 35°C.²⁸ Thapsigargin (1 μmol/L) was used to inhibit SR function and assess net Ca²⁺ influx. Action potentials (APs), Ca²⁺ currents, and Na⁺/Ca²⁺ exchange (NCX) currents were measured via a Axopatch 2B voltage-clamp amplifier as described previously.^11,29

Immunostaining

Isolated VMs were fixed, permeabilized, and stained to assess membrane localization of α1G, α1C, or RyR2.

Detubulation

Isolated VMs were incubated in a formamide solution (1.5 mol/L) for 15 minutes. Cells were quickly transferred to a formamide-free Tyrode’s solution to induce detubulation.

An expanded Materials and Methods section can be found in the online data supplement at http://circres.ahajournals.org.

Results

Phenotype of Adult Mice With α1G-TTCC Overexpression

Previously¹¹ we showed that inducible, cardiac-specific overexpression of the β2a subunit of the LTCC increases Ca²⁺ influx and myocyte contractility but eventually leads to depressed pump function, fibrosis, myocyte necrosis, cardiac hypertrophy, and premature death.¹¹ In the present study, we asked whether increases in Ca²⁺ influx of a similar magnitude through TTCCs, a class of Ca²⁺ channels reexpressed during cardiac disease, produced a similar excess Ca²⁺ phenotype. Mice with cardiac-specific, inducible α1G expression were generated to define the functional role of these channels in the myocardium. A bitransgenic system was used in which a standard α-MHC promoter-driven expression of tTA (regulated by doxycycline) is coupled to a modified α-MHC promoter containing the tet-operon for regulated expression of the α1G cDNA (Figure 1A). α1G mice, unlike the β2a mice,¹¹ had no premature death during the first 12 months of life. Similarly, there were no signs of histopathology (Figure 1B), cardiac dysfunction (Figure 1C), ventricular remodeling, or increases in myocyte death. For comparison, β2a hearts showed robust histopathology by 5 months of age (Figure 1B), as well as premature lethality.¹¹ Provocatively, not only were α1G mice without disease, they also showed enhanced ventricular performance (fractional shortening) compared with control animals (Figure 1D). These data show that expression of α1G-TTCCs induces a fundamentally different cardiac phenotype from overexpression of LTCC-β2a subunit.¹¹

Figure 1. Properties of mice with α1G-TTCC overexpression.

A, Schematic of the bitransgenic system for cardiac-specific α1G overexpression. tTA is the tetracycline-controlled transactivator system. B, Histology of α1G mice was not different than control. C, Representative M-mode echo recording in control and α1G mice. D, Fractional shortening (FS) and ejection fraction (EF) were greater in α1G vs control mice at comparable heart rates. *P<0.05.

Ca²⁺ Influx via T- and LTCCs in α1G Mice

One possible explanation for the pathological phenotype of β2a versus α1G mice is that Ca²⁺ influx is increased to a greater extent in β2a than α1G myocytes, resulting in Ca²⁺-mediated myocyte death. To address this possibility, we measured total Ca²⁺ influx in myocytes from α1G and β2a myocytes. I_Ca,T and I_Ca,L were measured (Figures 2 and 3) using standard biophysical conditions. Control VMs had no I_Ca,T, as expected, but I_Ca,T was present in every α1G VM. Peak I_Ca,T density (Figure 2B) was similar to peak I_Ca,L density in the β2a mice (Figure 3B). The voltage dependence of I_Ca,T was similar to published values observed in wild-type channels, with a peak near −30 mV (Figure 2B). I_Ca,T also had normal recovery from inactivation, with 80% of TTCCs fully recovered within 100 ms (Figure 2C). Therefore, at the normal heart rates of α1G mice, TTCCs will be fully available for activation. I_Ca,T was less sensitive to Ca²⁺ than I_Ca,L and was blocked by Ni²⁺ (Figure 2D), consistent with the known properties of wild-type α1G-TTCCs. These results show that the expressed α1G has functional and pharmacological properties like native I_Ca,T.

Figure 2. Ca²⁺ influx in α1G and β2α myocytes.

A, Representative I_Ca,T (I_Ca,L from −50 mV holding potential was subtracted from total Ca²⁺ current recorded from −90 mV holding potential) in control and α1G VMs. B, Voltage–current relationship of I_Ca,T (subtraction as in A) in control (n=10) and α1G (n=15) myocytes. C, Time course of I_Ca,T recovery from inactivation (n=7) and representative example of I_Ca,T recovery from inactivation (inset). D, Representative example of total Ca²⁺ currents under AP voltage clamp in control, β2a, and α1G VMs at 0.5 Hz. E, Integral of total Ca²⁺ current (without drugs) at different pacing frequencies in control (n=6), β2a (n=8), and α1G VMs (n=6). F, Representative example of [Ca²⁺]_i transients at baseline and after thapsigargin application (1 μmol/L) at 1 mmol/L and 3 mmol/L bath Ca²⁺. G, Peak [Ca²⁺]_i transient and rise rate of the [Ca²⁺]_i transient after thapsigargin treatment (in 3 mmol/L Ca) in control (n=20), β2a (n=20), and α1G VMs (n=25).

Figure 3. I_Ca,L density in control, α1G, and β2a VMs.

A, Representative examples of I_Ca,L in control, α1G, and β2a VMs. B and C, Voltage–current relationships (B) and peak I_Ca,L (C) in control (n=11), α1G (n=15), and β2a VMs (n=10). D and E, Voltage dependence of I_Ca,L activation (D) and average half-activation potential (V_0.5) (E) in control (n=11), α1G (n=15), and β2a VMs (n=8). *P<0.05, **P<0.01.

AP voltage clamp was used to measure and compare the total Ca²⁺ influx through both T- and LTCCs during a mouse AP (Figure 2D). Net Ca²⁺ influx during an AP was similar in α1G and β2a VMs and 3-fold greater than in control VMs (Figure 2E). These studies show that the amount of excess Ca²⁺ entry via TTCCs in α1G mice is approximately equal to the excess Ca²⁺ entry via LTCCs in the β2a mice.

Differences in Ca²⁺ influx in control, β2a, and α1G VMs were also compared by measuring the field-stimulated [Ca²⁺]_i transient after inhibition of SR function with thapsigargin (1 μmol/L) (Figure 2F and 2G). Peak [Ca²⁺]_i transients after thapsigargin treatment were similar in α1G and β2a VMs and significantly greater than control VMs. Peak rising rate of the [Ca²⁺]_i transient, which is an index of Ca²⁺ flux, was also similar in α1G and β2a VMs and greater than control VMs. Collectively, these results show that Ca²⁺ entry is increased to a similar extent in α1G and β2a mice. Therefore, differences in Ca²⁺ influx magnitude do not explain the different phenotypes of α1G and β2a mice.

I_Ca,L density was significantly smaller in α1G versus control myocytes (Figure 3A through 3C), and its voltage dependence of activation was shifted in the depolarized direction (Figure 3D and 3E). These changes may be a compensation (reduced Ca²⁺ influx via the normal influx pathway) for the increased Ca²⁺ influx through α1G-TTCCs.

AP, Contractions, and [Ca²⁺]_i Transients in α1G Myocytes

We have previously shown that excess Ca²⁺ influx through the LTCC increases myocyte contractions and [Ca²⁺]_i transients.¹¹ α1G VMs had longer AP durations at 50%, 70%, and 90% repolarization and a more prominent plateau phase than control VMs (Figure 4A and 4B). Peak systolic [Ca²⁺]_i transients were greater in α1G versus control myocytes, and the rate of [Ca²⁺]_i transient decay was accelerated (Figure 4C and 4D). Contraction magnitude of α1G myocytes was significantly greater than control myocytes. The duration of contraction was significantly shorter, and relengthening kinetics were accelerated in α1G versus control VMs (Figure 4E and 4F). These results show that increases in Ca²⁺ influx through α1G induces an increase in [Ca²⁺]_i transient and myocyte contraction but also suggest substantial changes in fundamental aspects of Ca²⁺ handling in α1G myocytes. The increased [Ca²⁺]_i transient and contractility in α1G myocytes was significantly smaller than in β2a myocytes (Figure 4).¹¹ One possible explanation for these results is that increased Ca²⁺ entry through TTCCs induces adaptations in myocyte Ca²⁺ handling such as reducing Ca²⁺ influx through the LTCCs (see above).

Figure 4. APs, contractions, and cytosolic [Ca²⁺]_i transients are altered in α1G VMs.

A, Representative example of APs in control and α1G VMs. B, Average AP durations (APDs) at 50%, 70%, and 90% repolarization in control (n=9) and α1G (n=11) VMs. C and E, Representative example of [Ca²⁺]_i transients (C) and contractions (E) from control, β2a, and α1G VMs. D, Peak [Ca²⁺]_i transients and ô (Tau) of control (n=10), β2a (n=19), and α1G VMs (n=17). F, Average fractional shortening (FS) in control (n=10), β2a (n=19), and α1G VMs (n=17). *P<0.05, **P<0.01. See text for discussion.

SR Calcium Load and I_NCX

Our data suggest that excess Ca²⁺ entry through TTCCs causes a much more modest increase in contractility than a similar increase in Ca²⁺ entry through LTCCs. One possible explanation is that the 2 sources of Ca²⁺ entry cause different degrees of SR Ca²⁺ loading. To test this idea, we measured SR Ca²⁺ content, defined as the caffeine-induced [Ca²⁺]_i transient. SR Ca²⁺ content was not different in α1G versus control VMs (Figure 5A and 5B), whereas SR Ca²⁺ content was increased significantly in β2a myocytes (Figure 5A and 5B). The decay phase of the caffeine-induced [Ca²⁺]_i transient primarily results from NCX-mediated Ca²⁺ efflux and was not different between control and α1G VMs. I_NCX measured at +60 mV and −80 mV (Figure 5C and 5D) was also similar in α1G and control myocytes. These results suggest that there are no significant changes in the density or [Ca²⁺]-dependent activity of the NCX in α1G myocytes, unlike what we have observed in β2a myocytes.¹¹ These results suggest that increased Ca²⁺ influx through T- and LTCCs are not equivalent in their respective abilities to load the SR.

Figure 5. SR Ca²⁺ content and NCX current are increased in β2a but not in α1G VMs.

A, Representative example of caffeine-induced [Ca²⁺]_i transients in control, β2a, and α1G VMs. B, Average data of peak caffeine-induced [Ca²⁺]_i transients and ô of decay in control (n=15), β2a (n=13), and α1G VMs (n=17). C, Representative example of I_NCX from control and α1G VMs. D, Peak I_NCX at +60 mV and −80 mV in control (n=16) and α1G VMs (n=17).

Ca²⁺ Regulatory Proteins in α1G Hearts

The rate of decay of the systolic [Ca²⁺]_i transient and its duration were significantly shorter in α1G versus control VMs, suggesting significant alterations in myocyte Ca²⁺ handling. To address this possibility, we measured Ca²⁺ regulatory protein abundance and phosphorylation state in control and α1G hearts. Western blot analysis of ventricular tissue confirmed expression of α1G-TTCCs in α1G hearts and found no detectable α1G protein in control tissue (Figure 6A). α1G expression was associated with reduced α1C protein abundance, consistent with reduced I_Ca,L amplitude (see Figure 3).

Figure 6. Western blot analysis of Ca²⁺ regulatory proteins in control and α1G hearts.

A, Representative Western blots of Ca²⁺ regulatory proteins from control (n=7) and α1G (n=6) hearts. B, Analysis of Ca²⁺ regulatory protein abundance normalized to GAPDH. RyR and PLB phosphorylation levels normalized to total RyR and PLB, respectively. R.U. indicates relative units. *P<0.05, **P<0.01.

SERCA2a abundance was significantly greater, and phospholamban (PLB) abundance was reduced with increased phosphorylation of PLB-Ser16 and PLB-Thr17 in α1G versus control ventricles (Figure 6A and 6B). These changes in SR protein abundance and phosphorylation can account for the enhanced rate of relaxation and more rapid decay of the [Ca²⁺]_i transient in α1G myocytes. No differences in RyR abundance or RyR-Ser2808 phosphorylation were observed in α1G hearts. NCX abundance was slightly but significantly greater in α1G versus control ventricles.

I_Ca,T and EC Coupling

We next studied whether Ca²⁺ influx through both T- and LTCCs were equally effective triggers of SR Ca²⁺ release. [Ca²⁺]_i transients and contractions induced by I_Ca,L and I_Ca,T in α1G VMs were measured in Na⁺- and K⁺-free conditions to eliminate Na⁺ and NCX currents and to minimize loss of voltage control. Voltage steps from −90 to −40 mV resulted in an I_Ca,T with a 3-fold larger amplitude than the I_Ca,L induced by voltage steps from −50 to +10 mV (Figure 7). However, I_Ca,T induced significantly smaller [Ca²⁺]_i transients than I_Ca,L (Figure 7C), and these transients had a slower rate of rise than those induced by I_Ca,L (Figure 7F). Contractions induced by I_Ca,T were smaller than those induced by I_Ca,L, and the kinetics of contractions were slower (Figure 7D and 7G). EC coupling gain (the ratio of peak [Ca²⁺]_i transient to peak Ca²⁺ current) was nearly 3-fold smaller for I_Ca,T versus I_Ca,L (Figure 7E). Similar experiments in control VMs showed no I_Ca,T, [Ca²⁺]_i transients, or contractions with voltage steps from −90 to −40 mV, whereas I_Ca,L induced [Ca²⁺]_i transients and contractions from −50 to +10 mV (data not shown). These findings show that Ca²⁺ influx through α1G-TTCCs is not an effective trigger of SR Ca²⁺ release, at least under our conditions.

Figure 7. I_Ca,T is less effective than I_Ca,L in inducing SR Ca²⁺ release (n=10).

A, Representative example of peak I_Ca,L and I_Ca,T and their corresponding [Ca²⁺]_i transients and contractions. B, Average I_Ca,L (at + 10 mV) and I_Ca,T (at −40 mV), as measured in 1 mmol/L bath Ca²⁺. C and D, Peak [Ca²⁺]_i transients (C) and peak contractions (D) triggered by I_Ca,L and I_Ca,T. E, EC coupling gain, determined as the ratio of peak [Ca²⁺]_i transients/peak current for I_Ca,L and I_Ca,T. F, Maximum rate of rise of the [Ca²⁺]_i transient induced by I_Ca,L and I_Ca,T. G, Maximum rate of contraction attributable to I_Ca,L and I_Ca,T. **P<0.01.

TTCC Membrane Localization

One reason why α1G-TTCCs may be ineffective triggers of SR Ca²⁺ release is that these channels are not localized to regions of the sarcolemma near the junctional SR. Localization of α1G on the VM membrane was determined using an α1G-specific antibody validated in previous studies.¹² This antibody labeled α1G-TTCCs that were primarily on the surface sarcolemma of VMs, with less staining within the T-tubules (Figure 8A). By comparison, the staining pattern of α1C, the pore-forming subunit of the LTCC, was primarily within the T-tubules, producing a similar pattern of coincident localization as RyR2 (Figure 8C and 8D). Membrane localization of T- and LTCCs was performed using confocal imaging at the level of the nucleus to ensure that similar intracellular and membrane surface regions were examined in every myocyte. This staining pattern supports the idea that I_Ca,T is an ineffective trigger of SR Ca²⁺ release because TTCCs are concentrated in the surface membrane away from the Ca²⁺ release channels (RyR2).

Figure 8. TTCC localization to surface sarcolemma away from the junctional SR.

A, TTCC staining pattern (red) in α1G VMs at the level of the nucleus (blue) using an α1G-specific antibody. B, Background α1G staining pattern (red) in control VMs. C, LTCC staining pattern (red) using an α1C-specific antibody. D, RyR staining pattern using a RyR2-specific antibody. E and F, Membrane staining with di-8-ANEPPS (green) showing extensive T-tubular system in normal α1G VMs (E) and lack of T-tubules in detubulated α1G VMs (F) at the level of the nucleus (blue). G, Representative example of peak I_Ca,L and I_Ca,T in detubulated α1G VMs. H, Comparison of peak I_Ca,L and I_Ca,T before (n=15) and after (n=7) detubulation in α1G VMs. **P<0.01.

Using antibodies to define membrane localization of ion channels can produce variable results.³¹ Therefore, we used another independent technique to confirm that TTCCs are more concentrated in the surface membrane than in the T-tubules of α1G VMs. Formamide-induced osmotic shock produces rapid changes in cellular volume that causes T-tubules to detach from the surface membrane.³⁰ This detubulation technique has been used to confirm the principal localization of LTCCs within the T-tubule system.³⁰ We used this approach to examine the idea that TTCCs are not in high density within the T-tubules. Membrane staining with di-8-ANEPPS confirmed the presence and loss of T-tubules in normal and detubulated α1G VMs, respectively (Figure 8E and 8F). Voltage-clamp techniques were then used to measure I_Ca,L and I_Ca,T in α1G VMs with and without detubulation (Figure 8G). Detubulation in α1G myocytes was associated with a 72% reduction in I_Ca,L and a significantly smaller, 32% decline in I_Ca,T, consistent with the antibody studies (Figure 8H). Collectively, these data support the idea that α1G-TTCCs are primarily on the surface membrane of VMs.

Discussion

The objective of this study was to define the functional role of TTCCs in adult VMs. TTCCs are expressed early in the developing heart,^12-14 where they are thought to contribute to pacemaking and the induction of contraction in immature myocytes. TTCC expression decreases after birth and few, if any, of these channels exist in adult VMs.¹⁴ Cardiovascular diseases that increase systolic wall stress in the ventricle, such as hypertension and myocardial infarction, induce reexpression of TTCCs,^15-19 which was hypothesized to contribute to cardiac dysfunction, arrhythmias, and sudden death.²⁰ However, a direct link between Ca²⁺ influx through TTCCs and these processes has not been causally established. To explore the role of TTCCs in the adult heart, we generated a mouse model with inducible expression of α1G-TTCCs.

α1G mice had functional TTCCs that caused an increased Ca²⁺ influx (Figure 2). Previously, we showed that increasing Ca²⁺ influx through LTCCs by expressing its β2a subunit increases myocyte contractility but over time causes cardiac dysfunction by inducing myocyte death from SR Ca²⁺ overload–induced necrosis.¹¹ These findings strongly support the idea that persistent increases in Ca²⁺ influx eventually cause cardiomyopathy by reducing the number of functional cardiac myocytes, consistent with the known benefit of LTCC blockers in select forms of heart disease.³²

The α1G mice used in the present experiments had increases in Ca²⁺ influx that were similar to or greater than in β2a mice, but, surprisingly, they did not develop a pathological cardiac phenotype. These hearts and their resident myocytes were hypercontractile, but we could not detect evidence for cardiac histopathology or sudden death, at least during the first 12 months of life. These new studies show that increasing Ca²⁺ influx through α1G-TTCCs, by itself, is not sufficient to induce cardiac dysfunction or arrhythmias. The reasons why increasing Ca²⁺ influx through T- and LTCCs induces such fundamentally different phenotypes in the adult heart is not clear and will be a major topic for future studies. Our current studies suggest that the location of the Ca²⁺ influx pathway appears to be the critical determinant of the induced cardiac phenotype.

Expression of α1G-TTCC Induces Alterations in Myocyte Ca²⁺ Handling

Ca²⁺ influx was ≈3-fold greater than normal in both β2a and α1G VMs. In the absence of compensatory changes in other Ca²⁺ regulatory processes, this increase in Ca²⁺ entry should cause similar increases in myocyte contractility and SR Ca²⁺ loading in myocytes from both transgenic mice. However, myocyte contractions and [Ca²⁺]_i transients were smaller in α1G versus β2a myocytes (Figure 4 and elsewhere¹¹). This appears to result from the fact that the SR Ca²⁺ load was not increased in α1G VMs (Figure 5), suggesting that T- and L-type Ca²⁺ fluxes are handled differently by cardiac myocytes. In this regard, we observed a significant decrease in the duration of contraction and the [Ca²⁺]_i transient in α1G myocytes (Figure 4C and 4E), suggesting that α1G-TTCC expression and the associated increase in Ca²⁺ influx induced remodeling of other aspects of myocyte Ca²⁺ handling. We explored these compensatory changes using cellular biophysical techniques and Western blot analysis.

Western blot analysis of ventricular tissue showed that the abundance and phosphorylation state of certain Ca²⁺-handling proteins were altered in α1G hearts. The protein abundance of the LTCC-α1C pore-forming subunit was significantly smaller in α1G versus control hearts. I_Ca,L density was also smaller in α1G versus control VMs, consistent with the Western blot analysis. These results suggest that the increased Ca²⁺ influx through the α1G-TTCCs induces a reduction in LTCC density. This compensatory change could reduce the likelihood of SR Ca²⁺ overload and associated pathologies in α1G hearts. Western blot analysis also showed an increased SERCA2a expression level and a reduced PLN abundance with increased phosphorylation at PLB-Ser16 and PLN-Thr17 in α1G versus control hearts. These changes will promote Ca²⁺ uptake by the SR and can explain the shortened [Ca²⁺]_i transient duration that we observed in α1G mice. Western blot analysis showed an increase in NCX abundance, but we did not find an increase in Ca²⁺-mediated NCX activity in α1G VMs in response to caffeine-induced SR Ca²⁺ release.

Peak systolic Ca²⁺ was increased in α1G versus control myocytes, but their SR Ca²⁺ loads were not greater than in controls. We speculate that the increased systolic Ca²⁺ in α1G myocytes results from an increase in Ca²⁺ influx through TTCCs, together with a similar (to that in control VMs) Ca²⁺ release from the SR (consistent with our finding that the SR Ca²⁺ load is normal). Why an increase in α1G-mediated sarcolemmal Ca²⁺ influx does not result in an increase in SR Ca²⁺ loading is not clear, because increasing Ca²⁺ influx via the LTCCs caused a significant increase in SR Ca²⁺ loading (Figure 5).

TTCCs Are on the Surface Sarcolemma, and LTCCs Are in T-Tubules

Our experiments demonstrate that Ca²⁺ influx through T- and LTCCs have a different capacity to load the SR (Figure 5) and to induce SR Ca²⁺ release (Figure 7). I_Ca,T induced [Ca²⁺]_i transients with smaller amplitudes and slower rates of rise than those caused by I_Ca,L, suggesting that TTCCs are not proximal to the RyRs to induce efficient SR Ca²⁺ release. Our results also suggest that Ca²⁺ entry via TTCCs is not taken up by the SR to the same extent as the Ca²⁺ influx via LTCCs. We therefore went on to test the idea that T- and LTCCs are localized to different regions of the surface membrane.

Immunostaining showed that α1G-TTCCs are localized to the surface membrane and are not significantly localized within the T-tubules, the principal site of LTCC localization (Figure 8). The T-tubule system is composed of extensive invaginations of the cell membrane that allow rapid transmission of electric signals into the interior of the VM for spatial uniformity of SR Ca²⁺ release for myocyte contraction. The fraction of cell membrane comprised by the T-tubular system is very species-dependent and has been reported to be between 50% and 64% for murine VMs.^33-37 LTCCs are localized to regions of the T-tubules (Figure 8) that are in close apposition to the junctional SR.³⁰ To confirm that LTCCs are primarily within the T-tubule system and that TTCCs are primarily on the surface membrane, we used osmotic shock, which on the swelling phase of the protocol, causes T-tubules to separate from the surface membrane. Our imaging studies confirm that the primary membrane lost following osmotic shock is from the T-tubules and that this membrane contains more L- than TTCCs (Figure 8). We did not perform electrophysiological measurements in the same cells before and after osmotic shock, so we could not determine the percentage reduction in membrane capacitance in osmotically shocked myocytes. Collectively, our results show that Ca²⁺ influx through T- and LTCCs has fundamentally different effects on EC coupling and SR Ca²⁺ loading. We conclude that Ca²⁺ entering VMs through T- and LTCCs occupies different submembrane compartments, where it has different effects on myocyte physiology.

Our findings in α1G VMs confirm previous reports that TTCCs are ineffective triggers of EC coupling.^23,24 Indeed, Sipido et al demonstrated that TTCCs are much less efficient at triggering Ca²⁺ release than LTCCs, and the authors attributed these differences to differential localization of T-and LTCCs.²³

We speculate that the differential membrane localization of T- and LTCCs in α1G VMs is responsible for the different phenotypes of β2a and α1G mice. The increased Ca²⁺ entry through TTCCs at the surface membrane could be rapidly transported out of the cell, so that only a small fraction of this Ca²⁺ enters the SR. This, coupled with a compensatory reduction in Ca²⁺ entry through LTCCs, could explain the modest increase in contractility, with no significant change in SR Ca²⁺ loading. Similar increases in total Ca²⁺ entry through the LTCCs in β2a mice caused increased SR Ca²⁺ loading, [Ca²⁺]_i transients, and contractility. Our most important finding is that increased Ca²⁺ entry through TTCCs did not induce cardiac dysfunction, histopathology, or premature death. These results show that the Ca²⁺ entry pathway, the location of the Ca²⁺ channels within the membrane, as well as the Ca²⁺ influx magnitude, determines the resultant cardiac phenotype. Additional studies are needed to determine the specific role of increased Ca²⁺ influx via TTCCs under pathological conditions.