Enhanced Expression of L-type Cav1.3 Calcium Channels in Murine Embryonic Hearts from Cav1.2-deficient Mice*

Voltage-gated calcium (Ca2+) channels play a key role in the control of heart contraction and are essential for normal heart development. The Cav1.2 L-type calcium channel is the predominant isoform in cardiomyocytes and is essential for excitation-contraction coupling. Although the inactivation of the Cav1.2 gene caused embryonic lethality before embryonic day E14.5, hearts were contracting before E14 depending on a dihydropyridine-sensitive calcium influx. We analyzed the consequences of the deletion of the Cav1.2 channel on the expression level of other voltage-gated calcium channels in the embryonic mouse heart and isolated cardiomyocytes. A strong compensatory up-regulation of the Cav1.3 calcium channel was observed on the mRNA as well as on the protein level. Reverse transcriptase PCR indicated that the recently identified new Cav1.3(1b) isoform was strongly up-regulated, whereas a more moderate increase was found for the Cav1.3(1a) variant. Heterologous expression of Cav1.3(1b) in HEK293 cells induced Ba2+ currents with properties similar to those found in Cav1.2 (–/–) cardiomyocytes, suggesting that this isoform constitutes a major component of the residual L-type calcium current in Cav1.2 (–/–) cardiomyocytes. In summary, our results imply that calcium channel expression is dynamically regulated during heart development and that the Cav1.3 channel may substitute for Cav1.2 during early embryogenesis.

Two T-type Ca 2ϩ channels have been detected in vertebrate heart, namely Ca v 3.1 (␣ 1G ) and Ca v 3.2 (␣ 1H ) (12,13). T-type Ca 2ϩ channels have been mainly studied in cardiac pacemaker cells and are thought to act together with the hyperpolarization-activated, cyclic nucleotide-gated (HCN) channels (14) and the L-type Ca 2ϩ channels in the generation of pacemaker potentials. There is evidence that in heart the T-type current also contributes to pathological processes such as ventricular hypertrophy, post-myocardial infarction, and arrhythmogenesis that occur during atrial fibrillation (15,16). In contrast to these observations, Ca v 3.1 knockout mice do not display a cardiac phenotype but reveal a lack of the burst firing of thalamocortical relay neurons and resistance to absence seizures (9).
Ca 2ϩ channels have been extensively investigated by molecular biology, biochemical, and electrophysiological methods in heart cells (17)(18)(19). In contrast, there is only limited knowledge on the transcriptional levels of L-and T-type channels during heart development. We have shown previously that an intact Ca v 1.2 gene is required for normal embryonic development after embryonal day E14 (5). Deletion of the cardiac calcium channel ␤2 gene results in embryonal death already after day E9.5 (20). In view of the relative late death of the Ca v 1.2 (Ϫ/Ϫ) embryos, the early death of the ␤2 knockout mice was surprising. Furthermore, the analysis of Ca v 1.2 (Ϫ/Ϫ) cardiomyocytes at day E12.5 revealed that rhythmic contractility depended on an L-type-like Ca 2ϩ current with a low affinity for dihydropyridines (5). These results suggested that an unidentified L-type like calcium channel replaced the lacking Ca v 1.2 channel during early embryogenesis.
In the present study, we compared the expression of six potential surrogate calcium channels by RT-PCR 1 in murine cardiomyocytes derived from Ca v 1.2-deficient embryos with that from wild type cells. We found major differences in the expression level of other calcium channel genes with a predominant up-regulation of a Ca v 1.3 splice form that has the electrophysiological and pharmacological properties of the L-type current found in Ca v 1.2 (Ϫ/Ϫ) cardiomyocytes.

MATERIALS AND METHODS
Animals and Cardiomyocytes Isolation-Wild type mice and Ca v 1.2 (Ϯ) mice (5) were housed at 20°C with a 12-h light/dark cycle. Individual embryos were obtained after breeding of wild type or heterozygous Ca v 1.2 (Ϯ) mice. The hearts were dissected at embryonic day E9.5, E12.5, or E15.5 and were either used for cell isolation or quickly frozen and stored at Ϫ80°C for RNA isolation. Embryonic cardiomyocytes were isolated as described (5). The dispersed cells were plated on 60-mm dishes for 48 h and then harvested for RNA isolation.
RNA Isolation, First Strand cDNA Synthesis, and RT-PCR-Frozen heart tissue or cultured cardiomyocytes were homogenized in Trizol LS * This work was supported by grants from the Deutsche Forschungsgemeinschaft and Fond der Chemie. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
‡ To whom correspondence should be addressed.  (Table I). For each primer pair, amplification of the cDNA was carried out in a reaction mixture containing 2 l of the first strand cDNA preparation, 1ϫ PCR buffer, 25 pmol of each primer, and 2.5 units of Taq polymerase (Promega) in a total volume of 50 l. The PCR cycles were denaturing at 94°C for 1 min, annealing at appropriate temperature (Table I) for 1 min, and extension at 72°C for 1 min during 35ϳ40 cycles. The PCR products were separated on 2% agarose gels and stained with ethidium bromide. The gels were recorded and analyzed by an imaging software (Bio-Rad Gel 2000, Quantity One software version 4). The relative amount of the amplicons were determined and normalized to that of the GAPDH fragment. The identity of the PCR products was verified by DNA sequencing.
The cloning and expression of the two N-terminal Ca v 1.3 splice variants (Ca v 1.3 (1a) and Ca v 1.3 (1b)) has been described (21). The relative expression levels of the Ca v 1.3 (1a) or (1b) mRNA were analyzed by the use of a mixture of the two forward primers Dutr01 and Race20 (1:1 ratio) and the reverse primer Race12 (see Fig. 3). Detection of the two splice variants by this mixture is possible because this PCR reaction yields different sized amplicons (21).
Western Blots-E12.5 embryonic hearts were exposed to one freezethaw cycle and homogenized in lysis buffer (20 mM KH 2 PO 4 , pH 7.2, and 1 mM EDTA) containing protease inhibitors (1 mM benzamidine, 0.1 mM phenylmethylsulfonyl fluoride and protease inhibitor mixture (1:500); Sigma). Lysates (60-g proteins) were loaded on 11% SDS-polyacrylamide gels. After electrophoresis, proteins were transferred to a polyvinylidene difluoride membrane (Millipore). Membranes were blocked in 1ϫ Tris-buffered saline-Tween 20 buffer containing 3% bovine serum albumin and 0.1% Tween 20. For detecting Ca v 1.2 and Ca v 1.3, the membranes were incubated overnight at 4°C with primary antibody (1:200) (Alamone and Chemicon, respectively). For ␤-actin, the lower part of the membranes was incubated overnight at 4°C in a 1:200 primary antibody. All immunoblots were followed by incubation for 1 h at room temperature in 1ϫ Tris-buffered saline-Tween 20 buffer with a 1:10,000 diluted peroxidase-conjugated goat anti-rabbit IgG (Jackson ImmunoResearch) for detecting Ca v 1.2 and Ca v 1.3 or peroxidaseconjugated goat anti-mouse IgG (Jackson ImmunoResearch) for detecting ␤-actin. The immunoblots were visualized by the ECL system (PerkinElmer Life Sciences).
Electrophysiology-Electrophysiological experiments were performed as described earlier (5,21). HEK 293 cells were transiently transfected with the expression vector for the Ca v 1.3 (1a) or (1b) subunit (0.2 g per well) and the expression vectors for the ␤ 3 and the ␣ 2␦ -1 subunits (0.15 g each per well) using LipofectAMINE according to the manufacturer's instructions (Invitrogen). The holding potential (HP) was Ϫ80 mV. Current-voltage relations (I-V) were recorded from Ϫ60 to 50 mV with 10-mV increments and a frequency of 0.2 Hz. Cumulative dose-response curves were measured using 2-3 different nisoldipine concentrations per cell. Trains of test pulses were to 0 or 10 mV for 40 ms with 0.1 Hz for I Ba of native cells or to Ϫ10 or 0 mV for 100 ms with 0.2 Hz for I Ba of expressed channels. IC 50 values were calculated by fitting the averaged dose-response curves to the Hill equation shown in Equation 1, where [nisoldipine] is the concentration of nisoldipine, IC 50 is the halfblocking concentration, H is the Hill coefficient, I is the current measured at any concentration of nisoldipine, and I max is the current measured in the absence of drug. The data for the I Ba of wild type cardiomyocytes were fitted with a two-component Hill equation.

Expression of Calcium Channel ␣ 1 Subunit mRNAs in Murine Fetal
Heart and Cardiomyocytes-To analyze Ca 2ϩ channel diversity in murine embryonic hearts, we determined the relative mRNA expression at day E9.5, a time point when the heart starts regular beating, at day E12.5, a time point when Ca v 1.2-deficient embryos still develop, and at day E15.5, a time point when the heart is fully developed (22). To compare expression levels obtained from independent experiments, we normalized the data by using GAPDH as an internal standard. In agreement with others (1, 23, 24), we obtained amplicons for Ca v 1.1, Ca v 1.2, Ca v 1.3, Ca v 3.1, and Ca v 3.2 channels (Fig. 1A). No specific DNA fragments were detected for the R-type Ca v 2.3 and L-type Ca v 1.4 calcium channels. Transcripts of the L-type channels Ca v 1.1, Ca v 1.2, and Ca v 1.3 were present throughout the fetal heart development. At day E9.5 the expression levels of the three channels were nearly identical. In contrast to the Ca v 1.3 and Ca v 1.1 mRNAs, the Ca v 1.2 subunit was up-regulated ϳ3-fold at day E15.5 (Fig. 1B). Ca v 1.2 becomes the predominant L-type channel isoform in late developmental stages. In isolated embryonic cardiomyocytes, up-regulation of Ca v 1.2 was even slightly more pronounced than in total heart at E15.5 (not shown). In general, the changes in mRNA levels were more pronounced in cardiomyocytes than in total heart tissue, which may be caused by various expression levels in multiple cardiac cell types. Whereas in total heart the Ca v 3.1 T-type Ca 2ϩ channel subunit could be consistently found at all three time points investigated, Ca v 3.2 was barely detectable at all devel-  (5), although hearts of Ca v 1.2 (ϩ/ϩ) and Ca v 1.2 (Ϫ/Ϫ) embryos contracted with the same frequency at day E12.5. This surprising result indicated that Ca v 1.2 is either not required for rhythmic heart contraction or is functionally compensated by another calcium channel during early embryogenesis. Therefore, we determined the expression of calcium channel ␣ 1 subunit mRNAs in cardiomyocytes isolated from days E9.5 and E12.5 of Ca v 1.2 (Ϫ/Ϫ) embryos (Fig. 2). Transcripts of the Ca v 1.2 subunit could be detected in (Ϫ/Ϫ) cardiomyocytes. These transcripts are not translated into a functional calcium channel protein, because their open reading frame is disrupted by the introduction of a premature stop codon in exon 3 (5). Deletion of the Ca v 1.2 gene increased the expression levels of three other channels, Ca v 1.1, Ca v 1.3, and Ca v 3.1 at day E9.5 (Fig. 2B). The mRNA of the Ca v 3.1 channel was not detected in wild type cardiomyocytes but was present in Ca v 1.2 (Ϫ/Ϫ) cells at day E9.5 ( Fig. 2B). At day E12.5, similar levels of mRNA were detected for the Ca v 1.1 and Ca v 3.1 channels. In contrast, the expression of Ca v 1.3 changed dramatically. At days E9.5 and E12.5, the mRNA of this channel was up-regulated 4-fold in Ca v 1.2 (Ϫ/Ϫ) cardiomyocytes (Fig. 2B).  (Fig. 3A). RT-PCR indicated that both splice variants of Ca v 1.3 were slightly increased at day E9.5 in Ca v 1.2 (Ϫ/Ϫ) hearts. Notably, the up-regulation of Ca v 1.3 observed at day E12.5 was even more pronounced. At this time, Ca v 1.3(1b) was clearly the dominant splice variant. (Fig. 3B).

Expression of Ca v 1.3 Splice Variants in Ca v 1.2 Wild Type and Ca v 1.2-deficient Murine Embryonic Hearts-To
These PCR results were confirmed by Western blots with an Ca v 1.3-specific antibody (Fig. 3C). The amount of Ca v 1.3 protein was substantially increased in the heart of Ca v 1.2 (Ϫ/Ϫ) embryos. As expected, Ca v 1.2 was absent in knockout embryos, whereas the expression levels of ␤-actin were not affected by the knockout.
Electrophysiological and Pharmacological Properties of the Ca v 1.3 Splice Variants-It was tempting to speculate that the previously described (5)  and ␤3 subunits has been reported in cardiac muscle (25). In initial experiments Ca v 1.3b was coexpressed with the ␤2a or ␤3 subunit, but in the presence of the ␤2a subunit only current densities below 1.0 pA/pF were obtained. Therefore we investigated the Ca v 1.3 subunit in further experiments in the presence of the ␤3 subunit. The coexpression of ␤3 with Ca v 1.3 will also facilitate comparability of our results with those of others (26). Both isoforms induced L-type barium currents with slow inactivation kinetics in 55% (Ca v 1.3(1a)) and 16% (Ca v 1.3(1b)) of transfected HEK 293 cells (Fig. 4, A and B). Some of the untransfected HEK 293 cells (control cells) showed a small rapid inactivating inward current with a current density at 0 mV of 0.64 Ϯ 0.08 pA/pF (n ϭ 3). A similar current could be detected in pcDNA3 vector-transfected cells. The current was not sensitive to the dihydropyridine nisoldipine and disappeared at an HP of Ϫ40 mV, suggesting that this endogenous current was not an L-type current. To safely separate cells expressing Ca v 1.3(1b) from cells only containing endogenous current, only cells showing a slowly inactivating current and a current density above 1.0 pA/pF were considered for analysis. With 5 mM Ba 2ϩ as charge carrier, the current density of Ca V 1.3(1a)-and Ca V 1.3(1b)-transfected cells was 11.8 Ϯ 2.0 pA/pF (n ϭ 30) and 1.9 Ϯ 0.24 pA/pF (n ϭ 25), respectively. The I-V curves indicated that the current through the Ca V 1.3 splice forms activated negative from Ϫ40 mV and were maximal between Ϫ10 mV and 0 mV (Fig. 4C). This result is in good agreement with the L-type current described in Ca V 1.2 (Ϫ/Ϫ) cardiomyocytes (5).
I Ba of wild type cardiomyocytes was blocked by nisoldipine with IC 50 values of 0.1 and 3.9 M. In contrast, I Ba currents of Ca V 1.2 knock-out cells were blocked with an IC 50 of 1.1 M (Fig.  4E). The IC 50 values for the nisoldipine block of I Ba induced by the heterologously expressed Ca V 1.3 splice variants differed from each other, being 0.1 M (Ca v 1.3(1a)) and 0.41 M (Ca v 1.3(1b)) (Fig. 4D). The latter value is close to that found in Ca V 1.2 (Ϫ/Ϫ) cardiomyocytes, suggesting that Ca v 1.3(1b) channels could be responsible for the L-type like Ca 2ϩ current in Ca V 1.2 knockout cells. DISCUSSION In this study, we investigated for the first time the developmental expression levels of all cardiac calcium channel ␣ 1 sub-units in murine fetal hearts. In agreement with previous electrophysiological data (27), we observed an up-regulation of three subunits underlying L-type calcium currents (Ca v 1.1, Ca v 1.2, and Ca v 1.3). Up-regulation of the Ca v 1.2 subunit was much more pronounced than that of the other two subunits. This finding emphasizes the particular importance of Ca v 1.2 for heart function, especially during late fetal stages and in the adult heart. In contrast, in early cardiac developmental stages, when the expression levels of Ca v 1.2 are low, mechanisms that do not necessarily include the Ca v 1.2 channel may control heart contraction (28,29). Like Ca v 1.2, the Ca v 1.1 and Ca v 1.3 subunits are also significantly up-regulated between days E9.5 and E12.5. In contrast to Ca v 1.2, the expression levels of the two latter subunits increase only slightly after day E12.5. Taken together, these findings imply that Ca v 1.1 and Ca v 1.3 possibly fulfill an unknown role in early heart development that may change at later stages of cardiac development. A function for Ca v 1.1 has not been reported, but for Ca v 1.3 it has been shown that this channel is important for the generation of spontaneous action potentials in sino-atrial node cells of the adult heart (10,11).
The expression levels of the two cardiac T-type calcium channels differ profoundly from each other during heart development. Ca v 3.2 expression levels were very low. In isolated cardiomyocytes and in heart tissue this subunit could not be detected at day E15.5, suggesting that Ca v 3.2 does not play a substantial role in adult heart muscle. However, we cannot exclude that Ca v 3.2 may be expressed only in a subset of heart cells (e.g. cells of the conduction tissue) and, therefore, was not detected in whole heart preparations. The low but clearly measurable expression of Ca v 3.2 at days E9.5 and E12.5 points to a possible role of this subunit in early heart development. In agreement with previous studies (6,13), the Ca v 3.1 mRNA was consistently detected in the developing heart, indicating that Ca v 3.1 is the major cardiac T-type calcium channel isoform.
As we pointed out, mice deficient for Ca v 1.2 die in utero before day E14.5. The results of this study indicate that the loss of this key subunit increases the mRNA levels of Ca v 1.3, Ca v 1.1, and Ca v 3.1. Although the increase of the latter two subunits is only moderate, the mRNA and the protein levels of the Ca v 1.3 subunit are profoundly up-regulated. RT-PCR analysis indicated that the increase in Ca v 1.3 expression is mainly due to the up-regulation of the Ca v 1.3(1b) isoform of this channel. The electrophysiological and pharmacological properties of the Ca v 1.3(1b) subunit heterologously expressed in HEK293 cells are in good agreement with that of the residual L-type calcium current found in Ca v 1.2 (Ϫ/Ϫ) cardiomyocytes. These results strongly suggest that Ca 2ϩ influx through Ca v 1.3(1b) is a major entry pathway in Ca v 1.2 (Ϫ/Ϫ) cardiomyocytes.
The functional relevance of Ca v 1.3 up-regulation is evident. Most likely, the up-regulation of Ca v 1.3 constitutes a compensatory mechanism of the cell to rescue the loss of Ca v 1.2 function, strongly suggesting the importance of Ca v 1.2. However, the compensatory increase in the Ca v 1.3 subunit is not sufficient to allow development of the embryo beyond day E14. As shown in this report, the Ca v 1.2 mRNA level increases at this time, presumably because only this channel provides the necessary calcium ions for cardiac contraction. The importance of the cardiac Ca v 1.2 calcium channel during embryogenesis has also been demonstrated recently by another study showing that the inactivation of the cardiac ␤2 subunit resulted in an early death at E11.5 (20). It is likely that the early death of the ␤2 subunit (Ϫ/Ϫ) mice is caused by an ER-retention of the Ca v 1.2 and the Ca v 1.3 subunit, because it was shown that the ␤2 subunit suppresses an ER-retention signal present in the ␣ 1 subunits (30).
The results of this study add to the growing evidence that the different ␣ 1 subunits of the high voltage-activated calcium channels can substitute for each other in certain, but not all, functions. Our results indicate that there is a transcriptional cross-talk between the Ca v 1.2 and Ca v 1.3 calcium channel genes. A cross-talk between these two calcium channels has also been described in pancreatic ␤ cells of Ca v 1.3-deficient mice (31). Furthermore, a compensatory increase in the expres-sion of calcium channel Ca v 1.2 subunit mRNA in Purkinje cells of the cerebellum was described in tottering mice (32). Tottering mice inherit a recessive mutation of the P/Q-type Ca v 2.1 subunit. In wild type mice, Ca v 1.2 mRNA is expressed at extremely low levels in Purkinje cells; but, in tottering mice Ca v 1.2 mRNA is significantly increased in these cells. The results of these studies support the notion that a compensatory up-regulation of the calcium channel mRNA levels occurs only in specific cell types. Clearly, future studies will be necessary to verify the above hypothesis and to pinpoint the mechanism by which the mutual control of Ca v 1.2, Ca v 1.3, and other Ca v genes is organized.