Differential Expression of Ion Channel Transcripts in Atrial Muscle and Sinoatrial Node in Rabbit
The aim of the study was to identify ion channel transcripts expressed in the sinoatrial node (SAN), the pacemaker of the heart. Functionally, the SAN can be divided into central and peripheral regions (center is adapted for pacemaking only, whereas periphery is adapted to protect center and drive atrial muscle as well as pacemaking) and the aim was to study expression in both regions. In rabbit tissue, the abundance of 30 transcripts (including transcripts for connexin, Na+, Ca2+, hyperpolarization-activated cation and K+ channels, and related Ca2+ handling proteins) was measured using quantitative PCR and the distribution of selected transcripts was visualized using in situ hybridization. Quantification of individual transcripts (quantitative PCR) showed that there are significant differences in the abundance of 63% of the transcripts studied between the SAN and atrial muscle, and cluster analysis showed that the transcript profile of the SAN is significantly different from that of atrial muscle. There are apparent isoform switches on moving from atrial muscle to the SAN center: RYR2 to RYR3, Nav1.5 to Nav1.1, Cav1.2 to Cav1.3 and Kv1.4 to Kv4.2. The transcript profile of the SAN periphery is intermediate between that of the SAN center and atrial muscle. For example, Nav1.5 messenger RNA is expressed in the SAN periphery (as it is in atrial muscle), but not in the SAN center, and this is probably related to the need of the SAN periphery to drive the surrounding atrial muscle.
This is the centenary of the discovery of the sinoatrial node (SAN), the pacemaker of the heart, by Keith and Flack.1 Early intracellular recordings of pacemaker and action potentials in the SAN were made by de Carvalho et al in 19592 and, in the ≈50 years since then, a wealth of data has been accumulated concerning the pacemaker and action potentials in the SAN and the underlying ionic currents.3,4 However, little is known about the molecular basis of ionic currents in the SAN and the aim of this study was to measure the abundance of messenger RNAs (mRNAs) coding for ion channels and related proteins in the SAN. The study was performed on rabbit, because of the existence of extensive functional data from this species: the early study of de Carvalho et al,2 as well as the majority of the studies on SAN since,3,4 have been performed on rabbit.
The SAN is a complex and heterogeneous tissue.5 The action potential is first initiated in the center of the SAN.4 It then propagates from the leading pacemaker site in the center to the periphery of the SAN (where SAN connects to atrial muscle) and then onto the atrial muscle of the crista terminalis and right atrial free wall.4 The SAN center is adapted for pacemaking: it has poor electrical coupling to protect it from the inhibitory hyperpolarizing influence of surrounding atrial muscle and it has a complement of ionic currents appropriate for pacemaking.3,4,6 The SAN periphery, although still capable of pacemaking, is adapted to drive the surrounding atrial muscle: it has (1) a large inward TTX-sensitive Na+ current (consequently, an action potential with a rapid upstroke) to generate sufficient inward (depolarizing) current to drive atrial muscle; and (2) strong electrical coupling to deliver the current to the atrial muscle.4 In this study, the abundance of mRNAs was measured in the periphery as well as the center of the rabbit SAN.
Materials and Methods
SAN tissue was sampled from male New Zealand White rabbits (1.5 to 2.5 kg; supplied by University of Leeds) for both quantitative PCR (qPCR) and in situ hybridization (ISH). Total RNA was extracted using a modified Qiagen protocol and 150 ng of total RNA from each sample was reverse transcribed; qPCR was performed using either Taqman probe or SYBR green chemistries. For ISH, digoxigenin labeled complementary RNA (cRNA) probes were used. For further details please see the online data supplement available at http://circres.ahajournals.org.
Figure 1D shows a section through the rabbit SAN ≈at the level of the leading pacemaker site. The section was cut perpendicular to the crista terminalis and includes part of the right atrial free wall, the crista terminalis (thick bundle of atrial muscle) and the intercaval region, where the SAN center is located. Typically, the leading pacemaker site is in the SAN center, ≈0.5 to 2 mm from the crista terminalis.4 From the leading pacemaker site, the action potential propagates through the SAN periphery (on endocardial surface of crista terminalis) to the right branch of the sinoatrial ring bundle (RSARB; vestige of embryonic venous valve) and then onto the atrial muscle of the crista terminalis etc.2,4 The action potentials in these regions are different.4 Tissue isolated from the SAN center shows spontaneous activity and the action potential has a low take-off potential, slow upstroke (≈2 V/s), small overshoot, small amplitude, long duration, low maximum diastolic potential (MDP), and a pacemaker potential (Figure 1C).2,4 Tissue isolated from the RSARB also shows spontaneous activity, which is paradoxically faster than that of the SAN center (Figure 1B).4 The action potential in the RSARB has a higher take-off potential, faster upstroke (≈50 V/s), larger overshoot, larger amplitude, shorter duration, higher MDP, and a steeper pacemaker potential than the action potential in the SAN center (Figure 1B).2,4 Similar to the RSARB, the action potential in atrial muscle has a fast upstroke (>100 V/s), large overshoot, large amplitude, short duration and high MDP, but it does not have a pacemaker potential (Figure 1A). From 7 rabbits, tissue samples were taken from the center (in intercaval region) and periphery (on endocardial surface of crista terminalis) of the SAN, as well as the atrial muscle in the right atrial free wall. From these samples, total RNA was reverse transcribed to generate complementary DNA (cDNA). The relative abundance of cDNA from 30 different transcripts was measured with qPCR. For an assessment of the accuracy of the tissue sampling for qPCR see the online data supplement available at http://circres.ahajournals.org. The distribution of selected transcripts in sections equivalent to that in Figure 1D was visualized using ISH.
Housekeeping Genes and SAN Markers
In the case of qPCR, abundance of transcripts is given relative to a housekeeping gene to correct for variations in input RNA. We assessed 3 housekeeping genes: 28S, GAPDH, and Na+-K+ pump (α1 isoform). The abundance of all 3 was constant in the 3 tissues studied (supplemental Table IV and supplemental Figure I in the online data supplement); 28S was chosen as the housekeeping gene, but similar data were obtained using the other housekeeping genes.
To characterize the tissue studied (Figure 1D), 2 markers of SAN tissue,5 atrial natriuretic peptide (ANP; hormone) and middle neurofilament (cytoskeletal protein), were investigated. Figure 1 (E and F) shows abundance of mRNA for ANP and neurofilament; abundance is expressed as a percentage of abundance in atrial muscle. Figure 2 (A and B) shows localization of ANP and neurofilament mRNA in sections. ANP mRNA was abundant throughout the atrial muscle of the crista terminalis and also in the RSARB, but it was not expressed in the SAN center (Figure 2A). In the SAN periphery, many cells did not express ANP mRNA, although some cells did (examples highlighted by arrows in Figure 2A). Neurofilament mRNA was absent in the atrial muscle of the crista terminalis, but was abundant in the SAN center (Figure 2B). Neurofilament mRNA was present in many, but not all, cells in the RSARB and SAN periphery (examples of cells not expressing neurofilament mRNA highlighted by arrows in Figure 2B). Interestingly, 2 tracts of neurofilament mRNA-expressing cells penetrated into the crista terminalis (Figure 2B, *).
Three connexin (Cx) transcripts (responsible for electrical coupling) were investigated. Expression of Cx40 mRNA (responsible for 200 pS gap junction channels) was significantly lower in the SAN than in atrial muscle (Figure 1G). ISH detected a low level of expression of Cx40 mRNA in atrial muscle (supplemental Figure II, red arrows) and no detectable expression in the SAN (supplemental Figure II). Cx40 mRNA was abundant in the walls of blood vessels (supplemental Figure II, yellow arrow). Cx43 mRNA (responsible for 60 to 100 pS gap junction channels) was significantly less abundant in the SAN than in atrial muscle (Figure 1H; similar results, not shown, were also obtained using a second, independent Cx43 qPCR assay). ISH showed abundant Cx43 mRNA in the atrial muscle of the crista terminalis as well as in the RSARB, but none in the SAN center (Figure 2C). In the SAN periphery, the majority of cells did not express Cx43 mRNA, but some cells did (examples highlighted by arrows in Figure 2C). Cx45 mRNA (responsible for 20 to 40 pS gap junction channels) did not vary significantly between tissues (Figure 1I).
Ca2+ Handling Proteins
Intracellular Ca2+ has been suggested to play an important role in pacemaking in the rabbit SAN;7 mRNAs for four Ca2+ handling proteins were investigated. RYR2 mRNA was significantly less abundant in the SAN compared with atrial muscle (Figure 1J). In contrast, RYR3 mRNA was significantly more abundant in the SAN center compared with atrial muscle. Therefore, from atrial muscle to SAN center, there is an apparent isoform switch from RYR2 to RYR3. The abundance of NCX1 (Na+-Ca2+ exchanger) mRNA was significantly lower in the SAN periphery compared with other tissues (Figure 1L). mRNA levels for SERCA2a (sarcoplasmic reticulum Ca2+ pump) did not vary between tissues (supplemental Table IV).
Na+, Ca2+ and HCN channels
The voltage-gated Na+ current, INa, is responsible for the rapid action potential upstroke in working myocardium (including atrial muscle) and in the RSARB (Figure 1, A and B).4 There is no measurable INa in the SAN center and the L-type Ca2+ current, ICa,L, is responsible for the slow action potential upstroke in this region (Figure 1C).4 Nav1.5 is the cardiac Na+ channel responsible for the majority of INa. As expected, the abundance of Nav1.5 mRNA was significantly lower in the SAN center than in atrial muscle; in the SAN periphery, it was higher, but still significantly lower, than in atrial muscle (Figure 3B). ISH showed abundant Nav1.5 mRNA in atrial muscle, but no signal in the SAN center (Figure 4C). Nav1.5 mRNA was present in the RSARB (Figure 4B) and, although Nav1.5 mRNA was not present in many cells in the SAN periphery, it was present in some cells (examples highlighted by arrows in Figure 4C). For all ion channels studied with ISH, the labeling of mRNA appeared as spots at low power and intracellular rings at high power (Figure 4, A and C), which probably correspond to the rough endoplasmic reticulum (in which mRNA is translated) in the perinuclear region. Recently, neuronal Na+ channels have been shown to be expressed in cardiac myocytes.8 Figure 3A shows data for Nav1.1 mRNA. In contrast to Nav1.5 mRNA, the abundance of Nav1.1 mRNA was significantly greater in the SAN than in atrial muscle (Figure 3A). Therefore, from atrial muscle to the SAN center, there is an apparent isoform switch from Nav1.5 to Nav1.1.
Transcripts for 2 L-type Ca2+ channel isoforms were investigated. Cav1.2 mRNA was significantly less abundant in the SAN than in atrial muscle; the reverse was true of Cav1.3 mRNA (Figure 3, C and D). In addition, Cav1.3 mRNA was significantly more abundant in the center than in the periphery of the SAN. ISH showed Cav1.2 mRNA to be abundant in the atrial muscle of the crista terminalis, but absent from the SAN center (Figure 5A). In the RSARB and SAN periphery (at foot of crista terminalis in this example), there was little or no Cav1.2 mRNA (Figure 5A). In contrast, ISH showed Cav1.3 mRNA was abundant throughout the SAN (in center and periphery and also in RSARB), with only weak labeling in atrial muscle (Figure 5B). Therefore, once again, between atrial muscle and the SAN center, there is an apparent isoform switch (from Cav1.2 to Cav1.3).
HCN channels are responsible for the hyperpolarization-activated current, If, that plays a key role in the pacemaker potential. The expression patterns of mRNA for 2 HCN isoforms, HCN1 and HCN4, were similar (Figure 3, E and F). HCN1 and HCN4 mRNA was significantly more abundant in the SAN center than in both the SAN periphery and atrial muscle (Figure 3, E and F). In the case of HCN4, but not HCN1, the mRNA was also significantly more abundant in the SAN periphery than in atrial muscle. ISH for HCN4 confirmed the qPCR data: it showed that HCN4 mRNA was abundant in the SAN center, but absent in the atrial muscle of the crista terminalis (Figure 6). It also showed that HCN4 mRNA was absent from the RSARB and, although present in some cells in the SAN periphery (examples highlighted by arrows in Figure 6B), it was absent from many (Figure 6B). This has important implications for pacemaking in the RSARB (Figure 1B).
In the SAN, as in atrial muscle, there is a transient outward current, Ito. In the SAN periphery, Ito can result in an action potential with a spike and dome profile (Figure 7A).4 Three α subunits are responsible for Ito in the heart, Kv1.4, Kv4.2, and Kv4.3. Once again there was evidence of an apparent isoform switch: Kv1.4 mRNA was significantly less abundant in the SAN than in atrial muscle (Figure 7B), whereas the reverse was true of Kv4.2 mRNA (Figure 7C). The abundance of Kv4.3 mRNA tended to decline from atrial muscle to the SAN center, but not significantly so (Figure 7D). KChIP2 is a β-subunit for Kv4.2 and Kv4.3.9 The abundance of KChIP2 mRNA was significantly lower in the SAN periphery than in atrial muscle (Figure 7E). ISH showed a decrease in KChIP2 mRNA in the SAN periphery: Figure 8A shows that KChIP2 mRNA was abundant in the atrial muscle of the crista terminalis and in the SAN center, but was largely absent from the SAN periphery (see expanded view in Figure 8A).
IK,ur, IK,r and IK,s are the ultrarapid, rapid, and slow delayed rectifying K+ currents, respectively. There is a 4-AP-sensitive sustained outward current in rabbit SAN possibly corresponding to IK,ur.10 IK,r and IK,s are present in the rabbit SAN: activation during the action potential triggers repolarization and sets the MDP and slow deactivation after the action potential plays an important facilitatory role in the pacemaker potential.3 We investigated transcripts for Kv1.5, responsible for IK,ur, ERG (Kv11.1) responsible for IK,r, KvLQT1 (Kv7.1) responsible for IK,s, and minK, a β subunit for KvLQT1. Whereas the abundance of Kv1.5 mRNA was similar in all tissues (Figure 7F), the abundance of KvLQT1 mRNA tended to be greatest in the SAN center (Figure 7H) and ERG mRNA was significantly more abundant in the SAN center than in atrial muscle as measured using a paired t-test (Figure 7G). Consistent with this, ISH showed that ERG mRNA was abundant in the SAN center, whereas labeling of ERG mRNA was substantially weaker in the atrial muscle of the crista terminalis (Figure 8B). No significant variation in the abundance of minK mRNA was detected (supplemental Table IV).
The background inward rectifier K+ current, IK,1, is generated by Kir2 channels. The ACh-activated K+ current, IK,ACh, is generated by a heterotetramer of Kir3.1 and Kir3.4. There were no significant differences in abundance of mRNAs for Kir2.1, Kir2.2 and Kir3.1 between tissues (supplemental Table IV). The ATP-sensitive K+ current, IK,ATP, is generated by the α subunit, Kir6.2: and the β subunit, SUR2A; mRNA for both was significantly less abundant in the SAN than in atrial muscle (Figure 7, I and J).
Of the 30 transcripts studied by qPCR, there were significant differences between tissues in the case of 19 (63%). When grouping transcripts with similar expression profiles, cluster analysis (please see the online data supplement available at http://circres.ahajournals.org) identified 2 significantly different clusters: transcripts that tended to increase (neurofilament, Cx45, Nav1.1, Cav1.3, HCN1, HCN4, Kv4.2, ERG, KvLQT1, Kir2.2, Kir3.1) and those that tended to decrease (ANP, Cx40, Cx43, RYR2, NCX1, SERCA2a, Nav1.5, Cav1.2, Kv1.4, Kv4.3, KChIP2, Kv1.5, minK, Kir2.1, Kir6.2, SUR2a) from atrial muscle to the SAN center. When grouping tissues based on transcript abundance, cluster analysis showed that the expression profiles of atrial muscle and the SAN (center and periphery) were significantly different (supplemental Figure III). Although overall transcript expression profiles of the center and periphery of the SAN were similar (supplemental Figure III), there were significant differences for individual transcripts between the center and periphery of the SAN (RYR3, Cav1.3, HCN1 and HCN4; Figure 1K and Figure 3, D through F). Furthermore, ISH shows that there are differences between the center and periphery of the SAN; Figure 8C summarizes data from ISH. Whereas atrial muscle and the SAN center showed distinct transcript profiles, the SAN periphery showed an intermediate profile (Figure 8C). The RSARB marks the most peripheral part of the SAN (Figure 1D) and it had the most atrial-like transcript profile (Figure 8C).
This study shows for the first time a complex variation in expression of many ion channel transcripts from atrial muscle to the periphery and center of the SAN.
Cx43 forms medium conductance gap junction channels and is the principal connexin isoform in the working myocardium. qPCR showed that abundance of Cx43 mRNA was lower in the SAN than atrial muscle (Figure 1H). ISH confirmed this, but, in addition, showed that Cx43 mRNA was present in some cells in the SAN periphery and, furthermore, Cx43 mRNA was abundant in the RSARB (Figure 2C). In rabbit, the distribution of Cx43 protein is identical.5 Cx45 forms small conductance gap junction channels. qPCR detected Cx45 mRNA and showed no significant differences between atrial muscle and the SAN (Figure 1I). Although Cx45 protein has been detected in rabbit atrial muscle by some11 but not all12 investigators, it is agreed that Cx45 protein is present in the rabbit SAN.12 Cx40 forms large conductance gap channel channels. qPCR and ISH showed that Cx40 mRNA was present in atrial muscle, and less abundant or absent in the SAN (Figure 1G and supplemental Figure II). Consistent with these findings, Cx40 protein has been reported in rabbit atrial muscle11 and blood vessels.12 Low levels of Cx40 protein have been reported in the rabbit SAN.12 Cx30.2 forms small conductance homomeric (≈9 pS) and, with Cx43 or Cx40, heteromeric (15 to 18 pS) gap junction channels.13 Cx30.2 has been recently reported in the conduction system of the mouse heart13 and, in the rat, using qPCR we have measured Cx30.2 mRNA to be significantly more abundant in the SAN than in atrial muscle (data not shown). Electrical coupling should be (1) weak in the SAN center to protect the SAN from the hyperpolarizing influence of atrial muscle; and (2) stronger in the SAN periphery to allow the SAN periphery to drive atrial muscle.4,6 Our findings are consistent with theoretical expectations: Cx40 mRNA was either weakly expressed or not expressed in all tissues; Cx43 mRNA was abundant in atrial muscle, present in the RSARB and SAN periphery, but absent in the SAN center; in the SAN center, Cx45 (and Cx30.2) mRNA was present.
Intracellular Ca2+ Handling Proteins
There was an apparent isoform switch from RYR2 to RYR3 from atrial muscle to the SAN (Figure 1, J and K). Consistent with this, we have previously shown that there is a decrease in RYR2 protein in the center of the rabbit SAN14 and, in mouse and rat heart, RYR3 mRNA has been shown to be expressed primarily in the cardiac conduction system.15 Intracellular Ca2+ release has been suggested to be involved in pacemaking in the rabbit SAN.7 The apparent isoform switch from RYR2 to RYR3 may, therefore, be functionally important. Differences in intracellular Ca2+ handling between the center and periphery of the rabbit SAN have been observed16 and this may again be related to the isoform switch. The abundance of NCX1 mRNA was significantly lower in the SAN periphery compared with the other tissue samples. This again could underlie differences in intracellular Ca2+ handling between the center and periphery of the rabbit SAN.16 There were no significant differences between atrial muscle and the SAN in abundance of mRNA for the Ca2+ handling protein, SERCA2a (supplemental Table IV).
qPCR and ISH showed that mRNA for the cardiac Na+ channel isoform, Nav1.5, was abundant in atrial muscle, but absent in the SAN center (Figure 3B and Figure 4). We have shown a similar distribution of Nav1.5 protein in mouse SAN.8 ISH also showed that Nav1.5 mRNA was abundant in the RSARB and, in the SAN periphery, there were some cells with Nav1.5 mRNA (Figure 4). Functional studies are consistent with these findings: in rabbit, there is a progressive decrease in the upstroke velocity of the action potential (from ≈50 to ≈2 V/s) from the RSARB to the SAN center.4 TTX block of INa greatly slows the action potential upstroke in the RSARB of the rabbit to ≈5 V/s; it also shortens the action potential and slows pacemaking.4 In contrast, it has no effect on the action potential in the center of the rabbit SAN.4 Furthermore, INa is recorded in putative peripheral, but not putative central, SAN cells.4 It is suggested that Nav1.5 in the SAN periphery is the important source of depolarizing current for the SAN to drive atrial muscle (see introduction).
qPCR showed Nav1.1 mRNA to be present in both atrial muscle and the SAN (Figure 3A). ISH failed to generate labeling of Nav1.1 (data not shown). This suggests that the amount of Nav1.1 mRNA present is relatively low. In mouse, both Nav1.1 mRNA and protein have been shown to be present in SAN and block of Nav1.1 by nM concentrations of TTX (which have little effect on Nav1.5) slows pacemaking.8,17 However, in the adult rabbit, TTX (even high concentrations that block Nav1.5 as well as Nav1.1) has no effect on the SAN center and central SAN cells do not have INa.4 Therefore, in the adult rabbit, although Nav1.1 mRNA can be detected in the SAN, it is unlikely to be functionally important. In contrast, Nav1.1 is present and functionally important in the SAN of newborn rabbit: nM concentrations of TTX decrease the spontaneous activity in the newborn, but not in the adult.18 The mouse and newborn rabbit have faster heart rates than the adult rabbit and perhaps this is why functionally important amounts of Nav1.1 are present in the SAN of the mouse and newborn rabbit, but not the adult rabbit.
There was an apparent isoform switch from Cav1.2 to Cav1.3 from atrial muscle to the SAN (Figure 3, C and D, and Figure 5). It is possible that there is a similar isoform switch in mouse.17 Furthermore, Cav1.3 knockout mice have sinus bradycardia, and SAN cells from these mice have a decreased rate of diastolic depolarization and slower spontaneous activity and develop spontaneous arrhythmias.19 ICa,L and not INa is responsible for the action potential upstroke in the SAN center4 and, therefore, it is important for pacemaking in the SAN center. Cav1.3 activates at more negative potentials than Cav1.2 19 and this is ideal for pacemaking.20
Both HCN1 and HCN4 mRNA was abundant in the SAN center as expected (Figure 3, E and F and Figure 6). However, unexpectedly, both were less abundant or perhaps absent from the SAN periphery and the RSARB (Figure 3 E and F and Figure 6). Immunocytochemistry showed that HCN4 protein is also less abundant in the periphery than in the center of the rabbit SAN (data not shown). This finding is unexpected, because the intrinsic pacemaker activity of the SAN periphery and the RSARB is faster than that of the SAN center (Figure 1 B and C).4 It is, therefore, possible that the ionic basis of pacemaking is different in the center and periphery of the SAN - perhaps intracellular Ca2+ is of particular importance in the RSARB.
Kv1.4 mRNA was more abundant in atrial muscle than in the SAN, whereas the reverse was true of Kv4.2 mRNA (Figure 7, B and C). Although Kv1.4 protein expression does not necessarily follow Kv1.4 mRNA expression,21 in this case the findings are consistent with findings of functional studies: an antisense oligodeoxynucleotide directed against Kv1.4 reduced Ito in rabbit atrial cells.22 In rabbit, recovery of Ito from inactivation is slower in atrial cells (time constants, 1.5 and 6.7 s) than in SAN cells (time constant, 45 ms).23 This is consistent with the primary importance of Kv1.4 in atrial muscle and Kv4 channels in the SAN, because recovery of Kv1.4 is markedly slower than that of Kv4 channels.24 Furthermore, the Kv4.2 channel has been cloned from rabbit SAN and shows similar properties to those of Ito in rabbit SAN cells.25 Kv4.3 mRNA did not vary between tissues (Figure 7D). This is not unexpected, because antisense oligodeoxynucleotide directed against Kv4.3 reduced Ito in rabbit atrial cells22 and recovery of Kv4.3 from inactivation is fast24 (as is recovery from inactivation in SAN23). KChIP2 mRNA was present in the SAN center and in atrial muscle, but not in the SAN periphery (Figure 8A). KChIP2 closely associates with and is important for expression of Kv4.2 and Kv4.3, but not Kv1.4.9 The presence of KChIP2 in the SAN center and atrial muscle may be important for the expression of Kv4.2 and Kv4.3 in the SAN center and Kv4.3 in atrial muscle. The significance of the absence of KChIP2 mRNA from the SAN periphery is unclear. The action potential is longest at the leading pacemaker site (SAN center) and the action potential duration decreases down the conduction pathway (ie, action potential duration: SAN center>SAN periphery>atrial muscle).2,26 This effect can be seen in Figure 1, A through C, and it is possibly the result of Ito, because, in rabbit, the density of Ito in atrial muscle is approximately double that in the SAN.23
Delayed Rectifier K+ Currents
Kv1.5 mRNA was detected in atrial muscle and the SAN (Figure 7F). Previously, we have shown, in guinea-pig, the presence of Kv1.5 protein in the SAN and surrounding atrial muscle27 and, in rabbit, the presence of 4-AP sensitive sustained outward current (possibly IK,ur for which Kv1.5 is responsible).4 Consistent with the presence of IK,r and IK,s in rabbit SAN,28 ERG, KvLQT1, and minK mRNA was detected in the SAN (Figure 7, G and H, and 8⇑B; supplemental Table IV). The abundance of both ERG and KvLQT1 mRNA tended to be higher in the SAN center than in atrial muscle (Figure 7, G and H, and 8⇑B). Perhaps this is because, in atrial muscle, IK,1 generates the resting potential and IK,r and IK,s control APD, whereas in the SAN IK,r and IK,s alone must generate the MDP as well as control the APD. Electrophysiological experiments suggest that the density of IK,r and IK,s is greater in the periphery than in the center of the rabbit SAN.4,28 In the present study, no evidence of such a difference was observed, but a difference in current density could be the result of control at the translational level.
Inward Rectifier K+ Currents
IK,1 is responsible for the resting potential in the working myocardium. IK,1 is absent from the SAN;3 this facilitates pacemaking and explains why the MDP in the SAN center is more positive than in atrial muscle (Figure 1, A and C). Therefore, surprisingly, the abundance of Kir2.1 and Kir2.2 mRNA was not significantly different between atrial muscle and the SAN (supplemental Table IV). This could be because Kir2.3 is a major IK,1 isoform in atria,29 or IK,1 is controlled at the translational level. Kir3.1 mRNA was present in atrial muscle and the SAN and there were no significant differences between tissues (supplemental Table IV). This is expected, because ACh via IK,ACh affects both atrial muscle and the rabbit SAN.3 We have previously shown Kir3.1 protein in both tissues in guinea-pig and rat.30 The abundance of both Kir6.2 and SUR2A mRNA was significantly lower in the SAN than in the surrounding atrial muscle (Figure 7, I and J). Again the presence of the 2 transcripts in both tissues is expected, because metabolic inhibition results in the activation of IK,ATP in both atrial muscle and the rabbit SAN.31 If the lower abundance of the transcripts in the SAN is translated into a lower density of IK,ATP, it is possible that the density of IK,ATP does not need to be as high as in atrial muscle to exert a physiological effect, because of the high input resistance of SAN cells.
For further discussion of the mechanistic implications of the results from this study see the online data supplement available at http://circres.ahajournals.org.
We thank Professor G. Smith (University of Glasgow) for supplying the SERCA2a qPCR primers.
Sources of Funding
This work was supported by a British Heart Foundation programme grant (RG/2001009) to M.R.B.
Original received May 4, 2006; revision received October 18, 2006; accepted October 19, 2006.
Keith A, Flack MW. The form and nature of the muscular connections between the primary divisions of the vertebrate heart. Journal of Anatomy and Physiology. 1907; 41: 172–189.
De Carvalho AP, De Mello WC, Hoffman BF. Electrophysiological evidence for specialized fiber types in rabbit atrium. Am J Physiol. 1959; 196: 483–488.
Irisawa H, Brown HF, Giles W. Cardiac pacemaking in the sino-atrial node. Physiol Rev. 1993; 73: 197–227.
Boyett MR, Honjo H, Kodama I. The sinoatrial node, a heterogeneous pacemaker structure. Cardiovasc Res. 2000; 47: 658–687.
Dobrzynski H, Li J, Tellez J, Greener ID, Nikolski VP, Wright SE, Parson SH, Jones SA, Lancaster MK, Yamamoto M, Honjo H, Takagishi Y, Kodama I, Efimov IR, Billeter R, Boyett MR. Computer three-dimensional reconstruction of the sinoatrial node. Circulation. 2005; 111: 846–854.
Bogdanov KY, Vinogradova TM, Lakatta EG. Sinoatrial nodal cell ryanodine receptor and Na+-Ca2+ exchanger: molecular partners in pacemaker regulation. Circ Res. 2001; 88: 1254–1258.
Maier SKG, Westenbroek RE, Yamanushi TT, Dobrzynski H, Boyett MR, Catterall WA, Scheuer T. An unexpected requirement for brain-type sodium channels for control of heart rate in the mouse sinoatrial node. Proc Natl Acad Sci U S A. 2003; 100: 3507–3512.
Lei M, Honjo H, Kodama I, Boyett MR. Characterisation of the transient outward K+ current in rabbit sinoatrial node cells. Cardiovasc Res. 2000; 46: 433–441.
Verheule S, van Kempen MJ, te Welscher PH, Kwak BR, Jongsma HJ. Characterization of gap junction channels in adult rabbit atrial and ventricular myocardium. Circ Res. 1997; 80: 673–681.
Coppen SR, Kodama I, Boyett MR, Dobrzynski H, Takagishi Y, Honjo H, Yeh H-I, Severs NJ. Connexin45, a major connexin of the rabbit sinoatrial node, is co-expressed with connexin43 in a restricted zone at the nodal-crista terminalis border. J Histochem Cytochem. 1999; 47: 907–918.
Kreuzberg MM, Sohl G, Kim JS, Verselis VK, Willecke K, Bukauskas FF. Functional properties of mouse connexin30.2 expressed in the conduction system of the heart. Circ Res. 2005; 96: 1169–1177.
Musa H, Lei M, Honjo H, Jones SA, Dobrzynski H, Lancaster MK, Takagishi Y, Henderson Z, Kodama I, Boyett MR. Heterogeneous expression of Ca2+ handling proteins in sinoatrial node. J Histochem Cytochem. 2002; 50: 311–324.
Mangoni ME, Couette B, Bourinet E, Platzer J, Reimer D, Striessnig J, Nargeot J. Functional role of L-type Cav1.3 Ca2+ channels in cardiac pacemaker activity. Proc Natl Acad Sci U S A. 2003; 100: 5543–5548.
Inada S, Mitsui K, Honjo H, Boyett MR. Why is Cav1.3 expressed in the sinoatrial node. Biophys J. 2005.
Brahmajothi MV, Campbell DL, Rasmusson RL, Morales MJ, Trimmer JS, Nerbonne JM, Strauss HC. Distinct transient outward potassium current (Ito) phenotypes and distribution of fast-inactivating potassium channel alpha subunits in ferret left ventricular myocytes. J Gen Physiol. 1999; 113: 581–600.
Wang Z, Feng J, Shi H, Pond A, Nerbonne JM, Nattel S. Potential molecular basis of different physiological properties of the transient outward K+ current in rabbit and human atrial myocytes. Circ Res. 1999; 84: 551–561.
Conley E, O’Beirne H, Owen JM, Hopkins PM, Boyett MR, Hancox JC Cloning and expression of a Kv4.2 channel from the rabbit sinoatrial node. J Physiol. 1998; 511: 149P–150P.
Dobrzynski H, Rothery SM, Marples DDR, Coppen SR, Takagishi Y, Honjo H, Tamkun MM, Henderson Z, Kodama I, Severs NJ, Boyett MR. Presence of the Kv1.5 K+ channel in the sinoatrial node. J Histochem Cytochem. 2000; 48: 769–780.
Melnyk P, Zhang L, Shrier A, Nattel S. Differential distribution of Kir2.1 and Kir2.3 subunits in canine atrium and ventricle. Am J Physiol. 2002; 283: H1123–H1133.
Dobrzynski H, Marples DDR, Musa H, Yamanushi TT, Henderson Z, Takagishi Y, Honjo H, Kodama I, Boyett MR. Distribution of the muscarinic K+ channel proteins, Kir3.1 and Kir3.4, in the ventricle, atrium and sinoatrial node of heart. J Histochem Cytochem. 2001; 49: 1221–1234.