| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 Pediatric Cardiology
2 Pathology and NYU Cancer Institute
3 Pharmacology and Physiology & Neurosciences, New York University School of Medicine, New York, New York
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
ion channels; mRNA expression; large-scale real-time quantitative reverse transcriptase polymerase chain reaction
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
There is a very large diversity in terms of the molecular components of ion channels; many of these give rise to ion channels with similar, yet subtly different features. Furthermore, association of ion channel subunits with accessory subunits often leads to altered expression levels, function, regulation, and pharmacological properties of these channels. Despite the relatively large body of data describing changes in the activity of specific ion channels during development, much less is known regarding the expression of molecular components of these ion channel subunits, which is even more diverse than the limited number of functional ion channel systems described in the heart (2, 7, 8, 10). It is imperative to understand both the nature of the ion channel transcriptome in the heart and how it changes during development before one can realistically start dealing with important clinical issues related to neonatal cardiology, including congenital abnormalities, cardiomyopathies, heart failure, arrhythmias, and cardiac drug therapy (5).
The recent development of efficient tools for large-scale analysis of gene expression has provided new insights into the involvement of gene networks and regulatory pathways in various processes. These methods include oligonucleotide and DNA microarrays, which can be used to analyze the expression of thousands of genes at a time. These methods, however, suffer from disadvantages (low sensitivity and unfavorable signal/noise ratio) when examining genes with low expression levels (e.g., ion channel genes) (13). An alternative is to use the real-time RT-PCR assay, which is more accurate and quantitative but is typically used to examine the expression of a smaller number of candidate genes. We developed a large-scale real-time RT-PCR screen to perform a comprehensive analysis of ion channel expression in the developing heart (late fetal and neonatal) and compared the results to expression in adult heart. Our data, together with publicly available microarray data sets, suggest a surprisingly rich complement of ion channels to be present in the immature heart and point to several fundamental age-related differences.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
Large-scale real-time RT-PCR.
We used real-time quantitative RT-PCR (qRT-PCR) to quantify the mRNA expression of a large number of ion channel transcripts during ontogeny. SYBR Green was used to detect dye-amplicon complex formation, and reactions were characterized by the fractional PCR cycle at which amplification of the PCR product reached a defined threshold (within the exponential phase of amplification). The parameter Ct (threshold cycle) is defined as the fractional cycle number at which this fluorescence level is reached. Thus, smaller Ct values are associated with larger starting quantities of the target mRNA species.
Primer design.
Primers were designed specifically against mouse transcripts, using Primer3 (45). As far as possible, primers were designed to span an intron to avoid PCR amplification of genomic sequences. Other design considerations included a melting temperature of 
60°C and a PCR amplicon of 90–120 bp. We conducted searches in dbEST, htgs, and nr databases at National Center for Biotechnology Information (1) to confirm the gene specificity of the primer nucleotide sequences and the absence of single nucleotide polymorphisms. No-template control and no-reverse-transcriptase control assays produced negligible signals, suggesting that primer-dimer formation and genomic DNA contamination effects were small. PCR specificity was verified by melting-curve analysis and by agarose gel electrophoresis (if doubts existed).
RNA extraction.
We used Swiss-Webster mice for these studies. All animal procedures were performed under National Institutes of Health guidelines with institutional approval. Adult male mice were anesthetized (120 mg/kg pentobarbital ip), and their hearts were removed. Ventricular tissue was used for RNA isolation. Ventricular tissue was obtained from the fetuses [18 days postcoitus (dpc)] of timed-pregnant Swiss-Webster females and from neonatal mice 1 day after birth. Tissue was minced (<1 mm) and stored in 10 volumes of RNAlater (Ambion) at 4°C until use (within 2 wk). Total RNA was extracted using the acid-phenol guanidinium method (TriReagent, Sigma). The quality of the RNA samples was determined by agarose electrophoresis as assessed by intact 18S and 28S RNA bands. RNA concentrations were determined spectrophotometrically (absorbance at 260 and 280 nm).
qRT-PCR study design.
In a first series of experiments, we isolated RNA from nine experimental groups. They consisted of three adult samples (5–8 wk; 1 heart in each), three neonatal samples (1 day, 10–14 pooled hearts per sample), and three fetal samples [embryonic day (E) 17.5; 4–14 pooled hearts for each]. We used only ventricular tissue (atria were separated before RNA isolation). Data generated from this study are presented in 
Figs. 2–4. In a second series of experiments, we pooled the hearts from several mice for RNA isolation. In the latter studies, we used three male adult hearts, 40 fetal hearts, and 40 newborn whole hearts. Data generated from this study are presented in Fig. 5.
|
|
|
|
|
PCR amplification.
PCR was performed in 384-well format. Plates were prepared by loading individual wells with 5 pmol (0.5 µM of each primer in 10 µl volume) of premixed primer pairs. Loading primers into the plates was facilitated using a BioMek 2000 robotics system (Beckman). The plates were lyophilized, hermetically sealed and stored at –20°C until use (within 2–4 wk). The layout of the plate was as follows: Each plate was divided into two identical sections, which allowed simultaneous measurements of two separate conditions that can be compared with each other (e.g., comparison of fetal and adult RNA on the same plate). All PCR reactions were performed in triplicate, and data are expressed as an average of the triplicates. A set of nine control genes was included (see later), which allowed us to examine 55 target genes per 384-well plate for each of the conditions.
All PCR reactions were performed using an ABI Prism 7900HT Sequence Detection System (Perkin-Elmer Applied Biosystems) available through the New York University (NYU) Cancer Institute Genomics Facility. On the experimental day, PCR master-mixes were prepared by mixing cDNAs with Power PCR SYBR Green PCR Master Mix (Perkin-Elmer Applied Biosystems). The RT-reaction product was diluted twofold and then mixed with the PCR master-mix (at a ratio of 1 µl of RT reaction to 9 µl of the master-mix). Thus, each PCR well contained cDNA that was generated from 60 ng of total RNA.
The thermal cycling conditions comprised of an initial denaturing step at 95°C for 10 min, and 40 cycles at 95°C for 5 s, 60°C for 15 s, and 72°C for 15 s. Following this cycling condition, a melting curve analysis was performed by cooling the PCR product to 60°C for 15s before gradual heating (ramp rate of 1°C/min) to 95°C. During this phase, double-stranded DNA separate and the fluorescence decreases. The "shoulder" in the melting point analysis is observed as a sharp peak in the first derivative of the fluorescence and is used as an indicator of PCR specificity.
Ion channel genes targeted in this large-scale assay.
We designed primers specific to transcripts of 9 Ca2+ channel subunits, 12 Ca2+ channel regulatory subunits, 9 voltage-activated Na+ channel subunits, 1 Na+ channel beta subunit (Navß1), 3 nonvoltage-activated Na+ channel subunits [epithelial sodium channel (ENaC) family], 6 transmitter-gated channel subunits (PTX family), 15 Cl– channel subunits, 33 voltage-activated K+ (Kv) channel subunits, 13 Kir channel subunits, 5 Ca2+-activated K+ (KCa) channel subunits, 10 two-pore K+ (K2P) channel subunits, 4 pacemaker channel subunits, 16 K+ channel regulatory subunits, and subunits of 5 proteins involved in Ca2+ homeostasis. Please refer to Table S1 (online supplement) for the full list of transcripts, gene names, and database entry locations. (The online version of this article contains supplemental material.) Of the classes of channels represented in the screen, the following transcripts were not included: ENaC
, 3 of the Na+ channel ß-subunits, Cav-4, Cav-
3, Cav-7, Cav-8, slo-ß3, NCS-1, DPP6, DPP10, Kir2.2, Kir7.1, Kv1.8, Kv6.1, Kv6.2, Kv6.3, Kv6.4, Kv10.2, Kv12.3, KCa4.1, KCa4.2, KCa5.1, K2P9.1, K2P10.1, K2P16.1, and K2P17.1. Classes of ion channels not represented in the screen include cyclic nucleotide gated cation channels (CNGs), transient receptor potential channels (TRPs), IP3 receptors, ryanodine receptors, neurotransmitter-operated channels, connexins, exchangers, or pumps. For a full list, please refer to Supplementary Table 1.
Data analysis.
The PCR Ct values were measured within the exponential phase of the PCR reaction using SDS software (version 2.2, Applied Biosystems). A correction was performed using a passive reference dye (Rox), which is present in the PCR master mix. Reactions with a Ct value >36 or with any evidence of nonspecificity (low melting temperatures or multiple peaks in melting point analysis) were excluded from the analysis.
The widely used –
Ct method (31), where the Ct data of a target transcript are expressed as relative changes between the two experimental conditions, does not provide information on absolute gene expression, which is normally possible only when comparing data to calibration curves obtained with known amounts of input cDNA. This is impractical at the scale of our assay. We therefore calculated target gene expression relative to that of a subset of HKGs. The selection of stable HKGs (i.e., that did not change their expression during development) was made using the Bestkeeper algorithm (41). A normalized Ct value (CtA) was obtained by calculating the geometric mean of the Ct values for each of the selected HKGs. The normalized expression of the selected HKGs was then calculated as:
 | (1) |
 | (2) |
Statistical analysis.
The mRNA levels in each subgroup of samples were characterized by their median values and ranges, standard error of the mean of individual reactions, and their coefficients of variation. Groups were compared by ANOVA, followed by a Dunnett's t-test with the fetal group as the control. Differences between two samples were judged significant at confidence levels >95% (P < 0.05).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
12%) were designated to have marginal expression or deemed to be present (calls of P or M) in at least one of the age groups. Clustering of these transcripts (Fig. 1) indicates three distinct groups; the expression of the majority is upregulated with development and a few others are either down regulated or unchanged. Many of the ion channel transcripts known to be expressed in the heart (e.g., Kirs, Kv4s, Kv1s, etc.) are not present in these microarray data. Although useful, these microarrays data are not sufficiently comprehensive to describe the alterations of ion channel gene expression. We therefore developed a large-scale real-time RT-PCR assay to examine ion channel expression during perinatal development.
Reproducibility of the qRT-PCR assay.
We performed an initial characterization of the qRT-PCR assay in terms of its reproducibility and sensitivity. There was a high degree of reproducibility within plates, irrespective of the developmental stage. The median value of the coefficient of variation of Ct values for triplicate reactions was 0.77% (range: 0.04–14.4%). Between individual plates, there was also an excellent reproducibility. As an example, Fig. 2 depicts a direct comparison of the Ct values obtained from 115 transcripts studied (over three separate plates) in two separate samples (different RT reactions from same adult male heart RNA). The resulting Ct values are plotted against each other (Fig. 2A). There was an excellent correlation (r = 0.97, P < 0.001) between these Ct values (the dotted lines represent 95% confidence intervals). A similar high correlation between Ct values was observed when using RNA prepared from different hearts (Fig. 2B), demonstrating the high reproducibility of the assay.
Normalization to control genes.
For semiquantitative real-time RT-PCR, the most commonly used normalization strategy involves standardization of gene expression to a constitutively expressed control (HKG) gene. However, the expression of the reference gene may itself be altered during experimental conditions. In recent years, it has become increasingly common to express data relative to a set of control genes that is statistically demonstrated not to be altered by the experimental intervention (41, 52). We included the following control (HKG) genes in each of our assays: 18S RNA, GAPDH, ß-actin, cyclophilin B, ribosomal protein S6, histone H3, HMBS, dihydrofolate reductase (DHFR), glucose-6-phosphate dehydrogenase 2 (G6PD), and hypoxanthine guanine phosphoribosyl transferase 1 (HGPRTase). Using the Bestkeeper algorithm (41), the HKGs were found to be well correlated with each other, with the exception of G6PD and HGPRTase. For the analysis of ion channel expression levels, we excluded those HKG with high expression levels. We therefore depicted target gene expression relative to averaged expression of GAPDH, cyclophilin B, and HMBS.
Using the qRT-PCR assay to determine alterations in ion channel gene expression during perinatal development.
We performed an initial analysis with primers specific to 50 transcripts of ion channel primary or accessory subunits. Each assay was performed in triplicate (i.e., we performed three separate assays for each of the following developmental time points: E17.5, 1 day neonate, and adult). The data are shown in Figs. 3 and 4. Given the high biological and technical reproducibility of the assay, we performed a subsequent assay without biological replicates, but with a larger primer set to 144 ion channel transcripts. These transcripts were chosen since none of them were present on the Affymetrix GeneChip (apart from Clca1 and HCN2, which served as internal controls) The extended RT-PCR data are shown as a heat map (Fig. 5), and raw data are available as Supplementary Table S2.
Developmental changes in K+ channel transcripts.
Of the Kir channel transcripts studied, the most predominantly expressed transcripts were Kir2.1, Kir3.1, Kir3.4, Kir6.1, and Kir6.2 (Figs. 1, 3, and 5). Most of the other Kir transcripts were expressed at relatively low levels. In general, the Kir channel transcripts were upregulated with perinatal development.
Among the Kv channel subfamily, the transcripts expressed most abundantly were Kv1.1 (Fig. 5), Kv2.1, and Kv4.2 (Fig. 3). Some were expressed at moderate levels (including Kv1.3, Kv1.4, Kv1.5, Kv1.6, and Kv7.1) and others at low levels. With the exception of Kv1.4, which was expressed at higher levels in immature myocardium relative to the adult, there was mostly an upregulation of Kv channel gene expression with development.
Transcripts for KCa channels were expressed at low levels in mouse heart (Figs. 3 and 5). Their expression did not appear to be regulated during perinatal development.
Of the K2P K+ channel transcripts studied, K2P3.1 and K2P6.1 were expressed at highest levels and their expression levels were higher in the adult than in the immature heart (Figs. 1, 3, and 5). K2P2.1 was expressed at lower levels and was also developmentally upregulated (Fig. 5). In contrast, K2P5.1 and K2P13.1, which were also expressed at relatively low levels, were developmentally downregulated (Fig. 5)
We also examined expression of several transcripts of K+ channel auxiliary subunits. The major transcripts expressed were minK, KChIP2, SUR1, and SUR2 (Figs. 1, 3, and 5). The minK transcript was expressed at high levels in the immature heart and at much lower levels in adult heart. KChiP2, SUR1, and SUR2A expression was each upregulated during perinatal development. The RT-PCR data also identified other K+ channel auxiliary to be expressed in the mouse heart (albeit at relatively low levels), including MiRP1, MiRP2, MiRP3, and Kvß3 (Figs. 3 and 5).
Developmental changes in Ca2+ channel transcripts.
Of the Ca2+ channel transcripts studied, we found the highest expression of Cav1.2, Cav3.1, and Cav3.2 in the mouse heart, while Cav2.2 and Cav2.3 were detected with low expression levels (Figs. 1, 4, and 5). Cav1.2 was markedly upregulated during perinatal maturation. In contrast, both of the Cav3 subfamily members were expressed at relatively higher levels in immature heart. Of the Ca2+ channel auxiliary subunits examined by qRT-PCR, Cavß2 and Cav1 were expressed higher than the others (Figs. 4 and 5). Both were upregulated during perinatal development. Other Ca2+ channel regulatory subunits were expressed at lower levels, including Cavß1, Cavß3, Cavß4, Cav2, Cav3, Cav1, Cav4, Cav6 (Fig. 5); expression of these transcripts was generally higher in the immature compared with the mature heart.
Although expressed at relatively low levels, some of the Ca2+ channel -subunits (including Cav1, Cav4, Cav6) were transcriptionally expressed at higher levels in the immature heart compared with the adult heart.
Developmental changes in Na+ channel transcripts.
We found the highest expression levels of Nav1.5, Nav2.3, and Nav1.4 in the adult mouse heart (Figs. 4 and 5). Expression of each of these transcripts was markedly upregulated during perinatal maturation. Other Na+ channel principal subunit mRNA (including Nav1.1, Nav1.3, Nav1.6, Nav1.8, Nav1.9; Fig. 5) were expressed at lower levels. The major Na+ channel regulatory subunit appears to be Navß1, which was markedly upregulated during perinatal development (Fig. 4). We also found Navß3 to be expressed at low levels, without any developmental regulation. We examined three of the four nonvoltage-gated ENaC Na+ channels; none of which were expressed at significant levels in the mouse heart (Fig. 5).
Developmental changes in pacemaker channels transcripts.
We examined expression of four pacemaker (HCN) channel transcripts. In the adult heart, we found highest expression of HCN2, HCN1, and HCN4 transcripts (Figs. 4 and 5). Expression levels of each of these were lower in the neonatal and fetal hearts.
Developmental changes in Cl– channels transcripts.
Cl– channel transcripts were remarkably well represented in the mouse heart, with almost all of the examined transcripts expressed at some developmental age. In the adult, the most predominant Cl– channel transcripts were CLC-4, CLC-3, CLC-7, CLC-2, CLC-6, and CLC-1 (Figs. 4 and 5). With the exception of CLC-5 (Figs. 1 and 5), the adult heart expressed higher levels of the CLC family of Cl– channel transcripts relative to the immature hearts (Figs. 1, 4, and 5). In contrast, the Clca transcripts (Clca1, Clca2, Clca3, and Clca4) were expressed at higher levels in the immature heart (particularly Clca1) and in some cases (e.g., Clca3 and Clca4) were not detectable in the adult heart. CFTR was expressed at low levels relative to some of the other Cl– channel transcripts and interestingly was more abundant in fetal heart compared with neonatal or adult hearts.
Developmental changes in P2X channels.
P2X4 and P2XM were well represented in the adult heart (Figs. 1, 3, and 5). Both were developmentally upregulated. Although expressed at relatively low levels compared with the other transcripts in this family, P2X1 and P2X7 were expressed at higher levels in fetal heart relative to neonatal or adult hearts (Figs. 1 and 5).
Subunits involved in intracellular Ca2+ homeostasis.
We examined the expression of a few selected subunits of channels involved in intracellular Ca2+ homeostasis. The mRNA levels (relative to the HKGs used) of these were in general much higher than the other ion channels we examined (Fig. 4). The expression of sarcoplasmic reticulum (SR) Ca2+ pump (SERCA2), SR Ca2+ release channel (RYR2), and phospholamban were all upregulated during perinatal development. Similarly, the Na+-K+ ATPase-
2 subunit expression was also higher in adult compared with neonatal of fetal ventricle. Expression levels of the Na+-Ca2+ exchanger (NCX1) decreased after birth. The microarray data additionally demonstrate expression of SERCA3, Na+-K+ ATPase-ß3 and sorcin in cardiac tissue (Fig. 1), as well as a number of proteins involved in intracellular pH regulation.
Ion channel transcripts not expressed in the mouse heart.
In preliminary experiments (not shown), we examined the ability of each primer pair to amplify a specific PCR product when using a mixture of RNA isolated from mouse brain and heart. Of all the primer pairs used in this study, primers to the following six transcripts were unable to produce a specific PCR product: Kv3.3, K2P18.1, slo-ß1, slo-ß2, slo-ß3, and Nav1.7. Possible reasons may include inadequate primer performance (despite several redesign attempts) or the fact that these transcripts are expressed at levels too low for detection by this assay. A subset of 16 primer pairs were able to amplify a specific product using brain/heart cDNA mixtures as template but did not amplify a specific product when using fetal, neonatal or adult mouse heart as template. These include Kir3.3, Kv11.2, Kv12.1, Kv3.1, Kv7.2, Kv7.3, Kv9.1, Kv9.3, K2P1.1, K2P4.1, K2P12.1, Cav-2, Cav-5, Cav1.1, Cav1.4, and Nav1.6. The latter transcripts are therefore unlikely to be expressed in the mouse heart, exhibit highly regional expression patterns in localized areas or cells, or are expressed at undetectably low levels using this assay.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
125 ng of RNA per reaction, but we may be able to decrease the input RNA by at least two orders of magnitude and still be within the sensitivity of the assay (particularly for the higher expressing transcripts). Using this assay, we analyzed transcriptional expression of a large number of ion channel genes in the mouse heart and describe how ion channel gene expression is altered during perinatal development.
Kir channels.
The "resting" membrane potential (or maximum diastolic potential) of cardiac cells is set by the inward rectifying K+ current, IK1 (25). Members of the Kir2 subfamily (particularly Kir2.1, Kir2.3, and Kir2.4) are generally thought of as molecular candidates of IK1 channels due to their strong rectification properties and pronounced Ba2+ block when expressed in heterologous expression systems (10). Furthermore, Kir2.1 antisense or dominant negative Kir2 subunits strongly reduce the cardiac IK1 (38), as does genetic ablation of the genes coding for the Kir2.1 and Kir2.2 subunit transcripts in mice (59). In keeping with this published data, we found high expression of Kir2.1 mRNA levels and low expression of Kir2.3 and Kir2.4 in the mouse heart (Kir2.2 was not included in the screen). There is important upregulation of IK1 before and directly after birth in the heart. Consequently, cardiac myocytes isolated from fetuses and neonates have a less negative maximum diastolic potential compared with adult (48). This increasing resting potential during fetal cardiac development must have important functional electrophysiological consequences in the developing embryonic, fetal, and neonatal heart, such as during the generation of spontaneous beating of the heart in the presence of an underdeveloped nodal system. This is best exemplified by the linear relationship that exists between the beating rate of fetal myocytes and the maximum diastolic potential (50). The less negative membrane potential in fetal cells may be partly due to a lower intracellular potassium concentration (16, 42). However, there is ample evidence that the whole cell IK1 density, which is the primary current responsible for the negative resting potential of adult myocytes, is low in the fetus and that it increases during pre- and postnatal development (9, 23, 25, 35). Our data presented here demonstrate upregulation of Kir2.1 mRNA levels, which we previously also observed using RNase protection assays and competitive RT-PCR (39). There may be other subunits contributing to the IK1 during early development, since mouse cardiac myocytes from Kir2.1(–/–) mice have a fairly negative maximum diastolic membrane potential (59). It is possible that this redundancy may be fulfilled by other Kir members (such as the Kir4 subfamily) or K2P channels, which demonstrated interesting developmental changes. Future experiments should be directed at determining the roles of these subunits in generating membrane potential during early development.
Other Kir channels in the heart includes the G protein-activated IK,Ach channels (10), which opens when acetylcholine binds to M2 muscarinic receptors and which is responsible for negative chronotropic and inotropic effects upon vagal stimulation. These channels are generally thought of as being comprised of heteromeric Kir3.1 and Kir3.4 subunits (10). Expression of Kir3.1 mRNA paralleled the known developmental upregulation of IK,Ach channels (49). Interestingly, expression level of Kir3.4 did not show a similar profile and was biphasically regulated after birth. The significance of this finding is not entirely clear, since functional IK,Ach channels can only be formed when both of these subunits are present. Future investigation should be directed at resolving this apparent discrepancy.
Another inward rectifier that is thought to open mainly under pathophysiological or stress conditions is the adenosine 5'-triphosphate-sensitive potassium (KATP) channel. Activation of these channels contributes to action potential shortening and cardioprotection during ischemia and ischemic preconditioning. The molecular candidates for the KATP channel are the pore-forming Kir6.1 or Kir6.2 subunits that heteromultimerize with regulatory SUR1 or SUR2 subunits (10). We found expression of each of these subunits in adult heart and clear upregulation during development. This result is in keeping with earlier observations of increased KATP channel expression with development (56) and developmental upregulation of these KATP channel subunits (37).
Repolarizing (voltage-activated) K+ currents.
The cardiac action potential repolarizes in part due to the voltage-dependent activation of outward K+ currents. Under voltage-clamp conditions these currents either are transient (i.e., they exhibit rapid spontaneous inactivation during a sustained depolarization) or are of the types referred to as delayed rectifiers (i.e., with little or no inactivation). In the mouse heart, there are two distinct transient outward K+ currents, referred to as Ito,f and Ito,s, with respective molecular correlates thought to be Kv4.2 and Kv1.4 subunits(17). There are also two distinct noninactivating (delayed rectifier type) K+ currents named IK,slow and Iss (57), which are characterized less extensively at the molecular level but may be due to channels assembled by Kv1.5, Kv1.2 and/or Kv2.1 subunits (28, 32). Our data are generally in agreement with this view of Kv channels in that we found relatively high mRNA expression levels of Kv4.2, Kv2.1, and Kv1.5 in the adult mouse heart. The high expression levels found for Kv1.1 mRNA are interesting and may point to additional complexity of molecular components of the mouse repolarizing K+ currents. Furthermore, low expression levels are seen for the following subunits: Kv1.3, Kv1.4, Kv3.2, and Kv4.3 (which give rise to rapidly repolarizing K+ currents) and Kv1.6 (which produces slowly inactivating delayed rectifier K+ currents). These subunits should also be considered as minor candidates for the complexity of the channel repertoire responsible for mouse repolarizing K+ currents. In general, repolarizing K+ channel gene expression was significantly upregulated in the mature heart compared with the immature heart (see also Ref. 43). This is in keeping with the known increase in repolarizing K+ currents (and shortened action potential) that occurs during development (12, 20, 23, 26, 53, 54). In many cases, there is an interesting "dip" in the levels of gene expression just after birth (Fig. 5). Additionally, Kv1.4 mRNA levels are elevated in immature heart relative to the adult (particularly in the neonate). To our knowledge there are no known functional correlates for this behavior in gene expression and more studies are needed to clarify the significance of this interesting finding.
Voltage-activated Ca2+ channels.
Voltage-activated Ca2+ channels open in response to membrane depolarization (during an action potential) and contribute to pacemaker activity and the relatively long-lasting cardiac action potential. The resulting Ca2+ influx serves as a trigger for graded Ca2+ release from the sarcoplasmic reticulum, thus regulating contractile strength. An alphabetical nomenclature has evolved for the distinct functional classes of Ca2+ currents recorded in different cell types (8). L-type Ca2+ currents require a strong depolarization for activation, are long-lasting, and are blocked by specific antagonists, such as dihydropyridines. The Ca2+ currents activated by weaker depolarization and that are resistant to these organic antagonists are referred to as T-type Ca2+ currents. Cardiac L-type Ca2+ currents are thought to be composed of the main, pore-forming Cav1.2 (1C) subunit, and several regulatory components, including ß2, 2, and -subunits (2). Disruption of the Cacna1c (Cav1.2) gene leads to embryonic lethality at 14 dpc (47), underscoring the importance of this channel after this developmental age. Consistent with these data, we found highest expression levels of Cav1.2 in mouse heart. Other Cav subunits are also important for heart function, as suggested by the finding that arrhythmias occur in prenatal hearts after ablation of the Cav2.3 (1E) subunit (33). The low expression level of Cav2.3 found in our studies, however, is suggestive of localized or regional expression patterns (such as the conduction system), or that expression occurs predominantly at earlier developmental time points (e.g., 10–12 dpc when arrhythmias were observed in the latter study). The Cavß subunits modulates channel activity and ensures correct plasma membrane targeting of functional Cav1.2 complexes (14, 18). Of all the ß-subunits, we found highest expression of Cavß2 in the mouse heart. Amongst the subunits, the Cav21 subunit mRNA was best represented in the mouse heart; this subunit enhances trafficking of 1 subunits (and unlike 22 subunits) also accelerates the rate at which the channel activates or inactivates (18). The complement of Ca2+ channel pore-forming and auxiliary subunits was roughly similar in the developing heart, with the difference that these subunits were expressed at lower levels in immature heart (see also Ref. 30). Some of the minor transcripts were expressed at higher levels in the immature heart (such as Cav4), but the physiological significance of this finding is unclear. The findings are generally consistent with the known increase of the L-type Ca2+ current that occurs during perinatal development (40). Alterations in inactivation kinetics (44) may in part be due to alterations in channel composition and alterations by auxiliary subunits.
Channels responsible for the cardiac T-type Ca2+ current are thought to be composed of Cav3.1 (1G) and/or Cav3.2 (1H) subunits (58). We saw expression of both of these in mouse heart. Interestingly, expression was higher in the neonatal heart compared with the fetal or adult heart, suggestive of an important role for these channels in electrophysiology of the immature heart. This is in keeping with the previously observed decreased T-type Ca2+ current density that occurs shortly after birth (15).
Na+ channels.
Na+ channels contribute significantly to the initiation, rapid repolarization, and the rapid propagation of the cardiac action potential. Of the 10 Na+ channel pore-forming () subunits, cardiac Na+ channels are thought to comprise mainly Nav1.5 subunits (7). Recent data also suggest a role for "neuronal" Na+ channel subunits in the heart (34). We found expression of Nav1.5, Nav1.4, and Nav1.8 in the adult mouse heart. We also found high expression levels of Nav2.3, which has a weak sequence similarity to the Nav1 subfamily and has previously been detected in mouse heart (27). During perinatal development, important alterations occur, including changes in Na+ current density, kinetics, and pharmacological sensitivity (36, 51). Our data demonstrate fairly large changes in Na+ channel gene expression with Nav1.5, Nav1.4, and Nav2.3 being expressed at higher levels in adult heart compared with immature heart. Although expressed at lower levels, Nav1.8 mRNA was present at higher levels in the immature heart compared with adult. Na+ channels are heterotrimers and also contain two auxiliary ß-subunits in addition to the pore-forming -subunits. These ß-subunits have multiple functions and affect channel expression level and kinetics but also have other diverse roles such as cytoskeletal interactions, regulation of cell migration, etc. (24). The strong developmental alterations in some of the Na+ channel ß-subunits may contribute to alterations in macroscopic currents changes during development.
Our assay also included three of the four nonvoltage-activated ENaC. As expected, none of these were found to be expressed at significant levels in the mouse heart at any of the developmental periods studied.
Cl– channels.
Cardiac tissue expresses a range of anion channels, including cAMP-activated Cl– channels, PKC-activated Cl– channels, volume-regulated anion channels, Ca2+-activated Cl– channels, Cl– channels activated by purinergic receptors, and voltage-activated Cl– channels (22). We saw expression of a large range of Cl– channel genes including the CLC family, where CLC-4, CLC-3, CLC-7, CLC-2, and CLC-1 were expressed at the highest levels in adult myocardium. Low expression levels of CFTR and the kidney-specific subfamily members (CLC-Ka and CLC-Kb) were found. In general, the CLC family members were all upregulated with perinatal development (with the exception of CLC-5, which was expressed at slightly higher levels in immature heart). This is in contrast to the Clca family, which was generally expressed at higher levels in immature myocardium than in the adult heart. Although each of the Clca family members was found to be expressed in immature heart, the predominant members are Clca1 and Clca2. Since the Clca family members express Ca2+-activated Cl– channels in heterologous expression systems, these data suggest a strong link between Ca2+ homeostasis and action potential maintenance in the immature heart.
Other channels.
P2X receptors are membrane ion channels activated by the binding of extracellular ATP. Although not well characterized in the heart, they are thought to mediate the effects of extracellular ATP on contractility in the heart (21). We found P2X4 and P2XM to be expressed at highest levels in the heart. Each of these two transcripts was upregulated during perinatal development. In contrast, P2X1 (although expressed overall at low levels) was expressed at higher levels in fetal heart compared with neonatal or adult heart. The significance of these findings is unclear without a defined functional role for P2X channels in cardiac function.
Pacemaker channels are predominantly functionally active in the pacemaker and conducting tissue of the heart. HCN subunits, homologous to the cyclic nucleotide-gated family of channels, are thought to be components of pacemaker channels (6). HCN1, HCN2, and HCN4 are all expressed in the heart, with HCN4 being the major component in the pacemaker region (where they may heteromultimerize with HCN1 subunits). Although pacemaker channels are present in the ventricle, they are not normally thought to operate as pacemaker channels in this tissue. HCN2 transcript expression is comparable in conducting and ventricular tissue (19). The predominant expression of HCN2 in our studies therefore most likely reflects the larger mass of ventricular tissue relative to conduction/pacemaker tissue. Upregulation of HCN transcripts during development suggests less emphasis on these channels for the generation of pacemaker activity in the immature myocardium (see also Ref. 46).
Ca2+ regulatory subunits.
Consistent with previously published reports regarding the altered Ca2+ handling mechanisms in the immature heart (5), we found developmental regulation of several prominent subunits involved in the process. For example, there was pronounced upregulation of subunits of proteins involved in SR Ca2+ uptake/release (such as SERCA2, RYR2, etc), whereas the Na+-Ca2+ exchanger (NCX1) was downregulated after birth. Our data are consistent with the notion of a developing role for the sarcoplasmic reticulum in regulating intracellular Ca2+ homeostasis after birth (5).
Limitations.
In this study, we examined alterations in mRNA expression levels of various ion channel transcripts during perinatal development. No mRNA-based assay (including real-time RT-PCR or microarrays) provides information on mRNA stability, posttranslational processes, or subunit assembly. Protein levels may therefore not correspond directly to mRNA transcript levels. Even if information existed on relative expression of ion channel proteins, the picture would not be complete. There are many other processes that may affect ion channel activity, including the specific subunit composition of multisubunit channel complexes, posttranslational modification of ion channel subunits (phosphorylation, myristoylation, palmitoylation, etc.), alterations in surface trafficking, and protein stability. Furthermore, protein levels may not necessarily correlate with channel activity. Thus, to put our data into true physiological context will require extensive experimentation using biochemical and electrophysiological techniques. Despite these reservations, in most cases there was a good correlation between our data and previously published transcript expression levels (at least in the adult mouse heart) and to known alterations in membrane currents during development. Given these limitations, our data have the real potential of highlighting alterations in ion channel composition and function during perinatal development. Another limitation with this assay (or any other assay that uses tissue obtained from whole organs) is that it does not provide information regarding region-specific transcript expression. The heart contains specialized regions with specialized functions (e.g., sino-atrial node, conducting system, atria, etc.) and ion channels are known to be regionally expressed in the heart (29). Therefore, other techniques (or isolation methods) will be needed to examine region-specific alterations in ion channel expression or in cardiac compartments (such as the conduction system, coronary vascular smooth muscle or endothelium). Since the ventricle dominates (at least by mass and cell number), our data most likely has more relevance to ion channel expression levels in the ventricles.
Perspectives and future directions.
The immature heart undergoes many changes in its membrane potential, action potential configuration, conduction, and contraction. In the past years several studies have demonstrated age-specific alterations in ion channel expression in the developing mouse heart. Our study provides a comprehensive overview of ion channel transcript expression during perinatal development. It is interesting to note that repolarizing K+ channel transcripts are generally expressed at low levels in immature heart compared with adult heart. In contrast, Cl– channel transcripts are expressed at higher levels, particularly those that are activated by intracellular Ca2+. This observation suggests an important link between Ca2+ homeostasis and action potential repolarization, which is particularly interesting when considering that the neonatal heart may rely mostly on Ca2+ entry from extracellular sources (via the Na+-Ca2+ exchanger) for contraction (3). Our future assays will be designed to include additional targets involved in Ca2+ homeostasis (including other Na+-Ca2+ exchanger subunits, Na+-K+ pump subunits, Ca2+ pumps, ryanodine receptors, IP3 receptors, etc.) better to address this issue. Our screen also identified unexpected age-dependent alterations for some transcripts. For example, the K+ channel regulatory subunit, minK, is expressed at very high levels in immature heart. Since K+ channels are generally expressed at low levels at this developmental stage, this result is suggestive of additional roles for this protein. Future experiments should be targeted to investigate disparities such as these. Our data provide a framework with which to integrate decades of electrophysiology studies with the more recent identification of genes responsible for the subunits responsible or these currents.
|
GRANTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
Authors' contributions: M. D. Harrell designed and tested the primers sets. Real-time RT-PCR was carried out by M. D. Harrell and S. Harbi, and S. Harbi performed the initial analysis. J. Zavadil was involved in methodological design and data acquisition. W. A. Coetzee further analyzed and interpreted the results and performed bioinformatics and statistical analyses. S. Harbi and J. F. Hoffman managed the limiting factors. This study has been supervised by W. A. Coetzee.
|
FOOTNOTES |
|---|
Article published online before print. See web site for date of publication (http://physiolgenomics.physiology.org).
* M. D. Harrell and S. Harbi contributed equally to this work. 
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  J. D. O'Brien, S. E. Howlett, H. J. Burton, S. B. O'Blenes, D. S. Litz, and C. L. H. Friesen Pediatric cardioplegia strategy results in enhanced calcium metabolism and lower serum troponin T. Ann. Thorac. Surg., May 1, 2009; 87(5): 1517 - 1523. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. P. Flagg, H. T. Kurata, R. Masia, G. Caputa, M. A. Magnuson, D. J. Lefer, W. A. Coetzee, and C. G. Nichols Differential Structure of Atrial and Ventricular KATP: Atrial KATP Channels Require SUR1 Circ. Res., December 5, 2008; 103(12): 1458 - 1465. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. E. Briggs, M. Takeda, A. E. Cuadra, H. Wakimoto, M. H. Marks, A. J. Walker, T. Seki, S. P. Oh, J. T. Lu, C. Sumners, et al. Perinatal Loss of Nkx2-5 Results in Rapid Conduction and Contraction Defects Circ. Res., September 12, 2008; 103(6): 580 - 590. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. W. Elrod, M. Harrell, T. P. Flagg, S. Gundewar, M. A. Magnuson, C. G. Nichols, W. A. Coetzee, and D. J. Lefer Role of Sulfonylurea Receptor Type 1 Subunits of ATP-Sensitive Potassium Channels in Myocardial Ischemia/Reperfusion Injury Circulation, March 18, 2008; 117(11): 1405 - 1413. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. J. Sampson, C. Terrenoire, D. O. Cervantes, R. A. Kaba, N. S. Peters, and R. S. Kass Adrenergic regulation of a key cardiac potassium channel can contribute to atrial fibrillation: evidence from an IKs transgenic mouse J. Physiol., January 15, 2008; 586(2): 627 - 637. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
