ABSTRACT
In the human, there is a circadian rhythm in the resting heart rate and it is higher during the day in preparation for physical activity. Conversely, slow heart rhythms (bradyarrhythmias) occur primarily at night. Although the lower heart rate at night is widely assumed to be neural in origin (the result of high vagal tone), the objective of the study was to test whether there is an intrinsic change in heart rate driven by a local circadian clock. In the mouse, there was a circadian rhythm in the heart rate in vivo in the conscious telemetrized animal, but there was also a circadian rhythm in the intrinsic heart rate in denervated preparations: the Langendorff-perfused heart and isolated sinus node. In the sinus node, experiments (qPCR and bioluminescence recordings in mice with a Per1 luciferase reporter) revealed functioning canonical clock genes, e.g. Bmal1 and Per1. We identified a circadian rhythm in the expression of key ion channels, notably the pacemaker channel Hcn4 (mRNA and protein) and the corresponding ionic current (funny current, measured by whole cell patch clamp in isolated sinus node cells). Block of funny current in the isolated sinus node abolished the circadian rhythm in the intrinsic heart rate. Incapacitating the local clock (by cardiac-specific knockout of Bmal1) abolished the normal circadian rhythm of Hcn4, funny current and the intrinsic heart rate. Chromatin immunoprecipitation demonstrated that Hcn4 is a transcriptional target of BMAL1 establishing a pathway by which the local clock can regulate heart rate. In conclusion, there is a circadian rhythm in the intrinsic heart rate as a result of a local circadian clock in the sinus node that drives rhythmic expression of Hcn4. The data reveal a novel regulator of heart rate and mechanistic insight into the occurrence of bradyarrhythmias at night.
Living things including humans are attuned to the 24 h day-night cycle and many biological processes exhibit an intrinsic ~24 h (i.e. circadian) rhythm. In the human, the resting heart rate (in the absence of physical activity) shows a circadian rhythm and is higher during the day when we are awake1,2. The heart is therefore primed, anticipating the increase in demand during the awake period. Conversely, bradyarrhythmias primarily occur at night1,3. Previously, the circadian rhythm in heart rate in vivo has been attributed to the autonomic nervous system: to high sympathetic nerve activity accompanying physical activity during the awake period and high vagal tone during the sleep period4. This is partly based on heart rate variability as a surrogate measure of autonomic tone5–7; however, we have shown that heart rate variability cannot be used in any simple way as a measure of autonomic tone8. Therefore, the mechanism underlying the circadian rhythm in heart rate is still unknown. We asked the question whether there is a circadian rhythm in the intrinsic heart rate set by the pacemaker of the heart, the sinus node.
In a cohort of nine mice, ECG telemetry in vivo showed a circadian rhythm in mean heart rate and other electrophysiological parameters (PR interval, QRS duration and QT interval) during a 12 h light:12 h dark lighting regime, and circadian rhythms in these parameters were sustained when the mice were placed in constant dark conditions (Figure 1). A circadian rhythm in heart rate, PR interval, QRS duration and QT interval are also observed in the human7,9–11. Mice are nocturnal and as expected they rested more and were less active from Zeitgeber time (ZT) 0 to ZT 12 (lights-on) and were more active from ZT 12 to ZT 0 (lights-off); this activity pattern continued in constant darkness (Figure 1)12,13. The heart rate was highest at ~ZT 13 and it varied by 76±4 beats/min (as determined by the least-squares best fit sine wave) over the course of 24 h (Figure 1; Table S1 in the Supplementary Information).
Circadian rhythm in intrinsic sinus node pacemaking and ion channel expression
Experiments on the isolated, denervated, Langendorff-perfused heart (Figure 2A) dissected at projected ZT 0 or ZT 12 demonstrated that there was a circadian rhythm in intrinsic sinus node pacemaking: the intrinsic heart rate was still higher at ZT 12 than ZT 0 – by 107 beats/min (Table S1). In the Langendorff-perfused heart, a suitable pacing protocol (Figure S1 in the Supplementary Information) showed the corrected sinus node recovery time (cSNRT), a commonly used indicator of sinus node function, to be significantly different at ZT 0 and ZT 12 (Figure 2B). It is concluded that there is a circadian rhythm in the intrinsic heart rate. What is responsible for the intrinsic circadian rhythm? Molecular circadian clock machinery is ultimately responsible for all circadian rhythms. The master clock is in the suprachiasmatic nucleus of the brain, but there are peripheral clocks in other organs14. To determine if there is circadian clock machinery in the sinus node, qPCR was used – qPCR showed that mRNA for a wide range of molecules known to be involved in the circadian clock is present in the sinus node and the expression varied in the expected manner from ZT 0 to ZT 12 (Figure 2C). Two key transcripts, Bmal1 and Clock, were measured in the sinus node at four time points (BMAL1 and CLOCK form a heterodimer); Figure 2D,E shows that they varied in a circadian manner and were at a maximum at ~ZT 0 (Table S1). The presence of an intrinsic circadian clock in the sinus node was confirmed by measuring the bioluminescence in the isolated sinus node from a transgenic mouse (Per1∷LUC) carrying a luciferase reporter gene reporting the activity of Per1, a key circadian clock component (Figure 2F). Per1-driven bioluminescence fluctuated in the expected circadian manner and this periodicity was lost in Cry1−/− Cry2−/− mice lacking the Cryptochrome genes and consequently lacking a functional clock15 (Figure 2F). Pacemaking is the result of the concerted action of ion channels and Ca2+-handling proteins and it is likely that the local circadian clock in the sinus node controls the heart rate by controlling their expression. Expression of mRNA for many of these molecules (and also some key regulatory transcription factors) was measured by qPCR and some transcripts, for example the pacemaking ion channel Hcn4, demonstrated significant daily rhythms (Figure 2G; Table S2).
Circadian rhythm in HCN4 and If
Of the potential mechanisms controlling the intrinsic heart rate, we principally focused on Hcn4, because this is known to play a central role in pacemaking in the sinus node16. Hcn4 mRNA was measured at six time points and was at a maximum at ~ZT 0 (Figure 3A; Table S1). Expression of HCN4 protein in the sinus node was measured using western blot at ZT 0 and ZT 12 – the western blot is shown in Figure 3B and the mean HCN4 band intensities are shown in Figure 3C. HCN4 protein expression was higher at ZT 12 than ZT 0 (Figure 3C). Expression of HCN4 protein in the sinus node was also measured using immunohistochemistry (Figure 3D,E). Figure 3D shows examples of immunolabelling of HCN4 in tissue sections through the sinus node from mice culled at ZT 0 and ZT 12); consistent with the western blot, the labelling was brighter, indicating higher expression, at ZT 12. This is confirmed by the mean data in Figure 3E. In summary, the data show a daily rhythm in HCN4 protein, but it is out of phase with the circadian rhythm in mRNA. Western blot and immunohistochemistry were used to measure HCN4 at four time points and this showed that HCN4 protein peaked at between ZT 6 and 12 (Figure S2; Table S1) – HCN4 protein, therefore, lags behind mRNA - this is expected. If there are changes in HCN4 protein, there should be changes in the corresponding ionic current, If. Figure 4A shows patch clamp recordings of If from isolated sinus node cells – the cells were isolated at ZT 0 and ZT 10 and the recordings were made at ZT 2 and ZT 12 (the earliest time points at which recordings could be made). The current density was approximately double at ZT12 than at ZT2 and this is confirmed by the mean current-voltage relationships for If at the two time points in Figure 4B. Once cells were isolated at ZT 0 and ZT 10, recordings from different cells could continue to be made for ~3 h. In the case of cells isolated at ZT 0 we observed the current density to increase the later the recording, but in the case of cells isolated at ZT 10 we observed the current density to decrease the later the recording. In Figure 4C we have plotted the density of If in a total of 367 sinus node cells against the time of the recording (this includes cells which were isolated at ZT 3, 6, 9, 12 and 15). Figure 4C shows a projected daily rhythm in the density of If – clearly the circadian clock does not stop ticking on isolation of single sinus node cells. If density reached a maximum at ~ZT 10, approximately at the same time as the peak in HCN4 protein as assessed by western blot and immunohistochemistry (Figure S2; Table S1). Current density determines the rate of change of membrane potential and is, therefore, the physiologically important variable. It is calculated by dividing the current amplitude by the cell capacitance, Cm, and these variables are shown in Figure 4D,E. As anticipated there is a projected daily rhythm in If amplitude, which peaks at the same time as If density, but the fractional change in amplitude is less than the fractional change of density (Figure 4D,E). This is because there is a circadian rhythm in Cm (Figure 4E); in other words, a circadian rhythm in Cm contributes to the circadian rhythm in If density. Cm is determined by cell size and other experiments confirmed that there is a circadian rhythm in cell size (Figure S3). A circadian rhythm in cell size may be surprising. However, a circadian rhythm in cell size has been reported in other tissues: in mouse liver, cell size at ZT 12 is less than at ZT 217; axonal volume shows a circadian rhythm in s-LNv clock neurons in Drosophila18; glial cells and neurons in the housefly’s visual system show rhythmic size changes19.
Role of If and other ionic mechanisms in circadian rhythm in intrinsic heart rate
To test whether the changes in If density could account for the daily rhythm in the intrinsic heart rate, the observed changes in If density (Figure 4C) were incorporated into a biophysical model of the spontaneous action potential of the mouse sinus node (Figure 4F). The model predicted a circadian rhythm in heart rate of 101±11 beats/min with a maximum heart rate at ~ZT 10 (Figure 4G; Table S1), roughly consistent with experimental data for the intrinsic heart rate (Figure 2A). This suggests that HCN4 and If participate in the circadian rhythm in the intrinsic heart rate. This was confirmed in the isolated, denervated sinus node dissected at four different time points. Block of If by 2 mM Cs+ (selective blocker of If20) decreased the heart rate as expected (Fig. 5A, top), but the effect of Cs+ on heart rate was greater at ZT 12 than ZT 0 (Figure 5A, bottom). Furthermore, in the presence of Cs+, the circadian rhythm in the intrinsic heart rate was abolished (Figure 5A, top). Analogous results were obtained in vivo. In the conscious mouse, the heart rate at ZT 0 and ZT 12 was measured using an ECGenie (a platform with embedded ECG electrodes). An intraperitoneal injection of 6 mg/kg ivabradine (selective blocker of If21) was given to block If. Once again, block of If decreased the heart rate as expected (Figure 5C) and the effect of ivabradine on heart rate was greater at ZT 12 than ZT 0 (Figure 5D). Furthermore, in the presence of ivabradine, the circadian rhythm in the heart rate in vivo was abolished (Figure 5C). It is concluded that HCN4 and If participate in the circadian rhythm in the intrinsic heart rate as well as the heart rate in vivo.
The so-called membrane and Ca2+ clocks are known to be responsible for pacemaking in the sinus node22. The membrane clock comprises ionic currents carried by a variety of ion channels. Although If carried by HCN channels is considered to be the most important, other ion channels also play a significant role. Various K+ channels showed a circadian rhythm (Figure 2G) and their potential role is considered in the Discussion. The Ca2+ clock involves spontaneous Ca2+ releases (Ca2+ sparks) from the sarcoplasmic reticulum during diastole. This drives Ca2+ extrusion via the electrogenic Na+-Ca2+ exchanger (Slc8a1; NCX1). The resulting inward current generated by NCX1 helps drive pacemaking. Although a circadian rhythm was not detected in any of the principal components of the Ca2+ clock (apart from Ca2+/calmodulin-dependent protein kinase IIδ – Camk2d), the potential role of the Ca2+ clock in the circadian rhythm in the intrinsic heart rate was investigated by recording Ca2+ sparks from isolated sinus node cells and measuring the effect of incapacitating the Ca2+ clock by 2 μM ryanodine on the intrinsic heart rate measured in the isolated sinus node (Figures S4-S6). Measurement of Ca2+ sparks showed that at ZT 12 as compared to ZT 0 there was an increase in spark duration and a non-significant increase in spark amplitude and consequently there was an increase in spark mass (calculated as amplitude×1.206×spark full width at half maximum amplitude3)23 (P=0.056), although there was a decrease in spark frequency and diastolic Ca2+ (Figure S5). The increase in spark mass at ZT 12 could increase the impact of the Ca2+ clock on pacemaking (although the effect could be mitigated by the reduction of Ca2+ spark frequency). In the isolated sinus node, the effect on the intrinsic heart rate of incapacitating the Ca2+ clock by ryanodine was greater at ZT 12 than ZT 0 (Figure S6A). In the presence of ryanodine, the circadian rhythm in the intrinsic heart rate was abolished (Figure S6A, top) and it is concluded that the Ca2+ clock also participates in the circadian rhythm in the intrinsic heart rate.
Link between local circadian clock in sinus node and membrane and Ca2+ clocks
To investigate a possible link between the local circadian clock in the sinus node and the circadian rhythm in the intrinsic heart rate, experiments were conducted on a transgenic mouse in which the Bmal1 gene had been knocked out in the heart only (driven by the α myosin heavy chain promoter)24. Knockout of Bmal1 is known to incapacitate the circadian clock25. Figure 6A confirms that Bmal1 mRNA was effectively knocked out in the sinus node at both ZT 0 and ZT 12 and Figure 6B shows that this disrupted the circadian rhythm in the expression of Clock mRNA – evidence that the circadian clock in the sinus node had been disrupted as expected. In the cardiac-specific Bmal1 knockout mouse, from ZT 0 to ZT 12, there was no longer a significant variation in expression of Hcn4 mRNA (Figure 6C) and HCN4 protein (Figure 3D,E). Consistent with this, the variation in If density from ZT 0 to ZT 12 was blunted (Figure 6D,E); this is best shown by Figure 6F, which compares the variation in If density from ZT 0 to ZT 12 in wild-type and cardiac-specific Bmal1 knockout mice. Finally, Figure 5B (top) shows the intrinsic heart rate as measured in the isolated sinus node: whereas there was a circadian rhythm in the wild-type mice with the intrinsic heart rate peaking at ~ZT 12 (Figure 5A, top), in cardiac-specific Bmal1 knockout mice this normal pattern was lost (Figure 5B, top). Furthermore, whereas there was a circadian rhythm in the reduction of the intrinsic heart rate on block of If by Cs+ in wild-type mice (with the effect peaking at ~ZT 12; Figure 5A, bottom), in cardiac-specific Bmal1 knockout mice, once again, this pattern was lost (Figure 5B, bottom). In contrast, cardiac-specific Bmal1 knockout had little effect on the circadian rhythm in the Ca2+ clock-control of the intrinsic heart rate (Figure S6B). It is concluded that the local circadian clock in the sinus node is controlling HCN4 and If (but perhaps not the Ca2+ clock) and thereby the intrinsic heart rate.
Evidence that clock protein, BMAL1, controls Hcn4 transcription
The CLOCK:BMAL1 heterodimer acts as a transcriptional activator or enhancer by binding to E-box binding sites in the promoter, intron or exon of a gene26,27. Using RVISTA (https://rvista.dcode.org) we found eight canonical E-box binding sites in the Hcn4 gene and 20 kb of its 5′ flanking region (Figure 6H). In vitro ChIP was used to test whether BMAL1 specifically binds to these sites: 3T3-L1 cells were UV cross-linked following transfection with His-tagged Bmal1. His-tagged BMAL1 bound to its DNA targets was then immunoprecipitated using antibodies directed against the His-tag. DNA pulled down by ChIP was analysed by qPCR using primers mapping to the various E-box binding sites (Figure 6G). Figure 6G shows E-box binding sites D and G (within introns of the Hcn4 gene; Figure 6H) were significantly more abundant than background signal (red dashed line).
Role for the autonomic nervous system in circadian rhythm in heart rate in vivo?
This study highlights the role of intrinsic factors in the circadian rhythm in intrinsic heart rate. Nevertheless, a role for the autonomic nervous system must be considered. In this study, in vivo, the heart rate was high during the awake period when the mice were physically active (Figure 1). If the level of physical activity is high, it would be expected to influence the heart rate via the autonomic nervous system (via an increase in sympathetic activity and possibly a decrease in vagal activity). However, in caged-housed mice with no access to a running wheel, activity levels will be low. A light pulse when mice are active is known to cause the mice to freeze and be inactive12. Figure 7A shows that a 1 h light pulse from ZT13 to ZT14 caused the physical activity of the mice to fall to baseline values, whereas the heart rate remained relatively high. In contrast, a 1 h light pulse from ZT1 to ZT2 was again associated with a baseline level of physical activity, but the heart rate was relatively low. Therefore, in this experiment, the heart rate was primarily influenced by the time of day rather than the activity level. Figure S7 shows that there is no discernible relationship between heart rate and physical level over 24 h (Figure 7A) and before, during and after the light pulses (Figure 7B). In vivo, the heart rate in the absence of physical activity can be obtained by recording the ECG from the anaesthetised mouse (although anaesthetic is known to depress the intrinsic pacemaker activity of the sinus node28 as well as cardiac vagal and sympathetic tone29). The heart rate in the anaesthetised mouse also varied in a daily manner – the heart rate was highest at ~ZT 12 and it varied by 51±15 beats/min over 24 h (Figure 7B; Table S1). It is concluded that in this study the effect of activity on the heart rate of the mice was slight and not discernible. Next, the effect of vagotomy was investigated. There are right and left vagal nerves and sectioning of both is lethal. However, vagal nerves to the sinus node are primarily, but not exclusively, from the right vagus30. The right vagus was sectioned in a group of rats (rat was studied, because we have experience of vagotomy in the rat31). Figure 7C shows that following the vagotomy the circadian rhythm in heart rate persisted.
We further tested the involvement of the autonomic nervous system by blocking cardiac muscarinic and β receptors by atropine and propranolol. Experiments were conducted on anaesthetised mice, and intraperitoneal injections of 1 mg/kg atropine and 1 mg/kg propranolol were given, similar doses to those used by others32–34 and by us in an earlier study of exercise training-induced bradycardia in the mouse35. After autonomic blockade, the circadian rhythm in heart rate persisted, although it was reduced in amplitude (Figure 7D). In our earlier study, similar concentrations of atropine and propranolol also failed to abolish the exercise training-induced bradycardia35; we concluded that the autonomic nervous system is not responsible for the exercise training-induced bradycardia35 and based on Figure 7D we could conclude that it is also not solely responsible for the circadian rhythm in heart rate in vivo. However, following the publication of our study of exercise training-induced bradycardia, Aschar-Sobbi et al.36 used intraperitoneal injections of higher concentrations of atropine and propranolol (2 mg/kg atropine and 10 mg/kg propranolol) and stated that the exercise training-induced bradycardia in the mouse is abolished on autonomic blockade. Therefore, we repeated the experiment in Figure 7D, but we used intraperitoneal injections of 2 mg/kg atropine and 10 mg/kg propranolol as did Aschar-Sobbi et al.36. This time, after autonomic blockade, the circadian rhythm in heart rate was lost (Figure 7E). Although this suggests that the autonomic nervous system is involved in the circadian rhythm in heart rate in vivo, the effect of the higher concentrations of atropine and propranolol must be taken into consideration. Propranolol blocks β-receptors with an IC50 of 12 nM (http://www.selleckchem.com/products/propranolol-hcl.html). However, work on heterologously expressed ion channels in cell lines has shown that propranolol blocks the cardiac Na+ channel, Nav1.5 (Scn5a) with an IC50 of 2.7 μM37 and HCN4 with an IC50 of 50.5 μM38. It was confirmed that propranolol also blocks If in native heart cells: in mouse sinus node cells, 68.4 μM propranolol blocked If by 54% (Figure 7F); typical current traces and current-voltage relationships are shown in Figure S8. Assuming that only 10% of propranolol is free39, an injection of 10 mg/kg propranolol is estimated to give a propranolol concentration of 6 μM if it partitions in all body water (~60% of body mass in the mouse40) and 66 μM if it only partitions into the blood (58.5 ml/kg in the mouse; https://www.nc3rs.org.uk/mouse-decision-tree-blood-sampling). Therefore, based on these estimates, it is possible that 10 mg/kg propranolol may block, to some extent, Nav1.5 and HCN4, both of which vary in a circadian manner (Figure 2G). If propranolol did block HCN4 to a significant extent, it is not surprising that propranolol abolished the circadian rhythm in heart rate (Figure 7E), because ivabradine also did (Figure 5C). Propranolol at a dose of 1 mg/kg as used for Figure 7D is expected to cause less block of Nav1.5 and little or no block of HCN4 and yet complete block of β-receptors (the expected plasma concentration is still many times greater than the IC50 for block of β-receptors). As an aside, calculation shows that atropine at a dose of 1 or 2 mg/kg is sufficient for complete block of M2 muscarinic receptors. Based on Figure 7D, the autonomic nervous system cannot be solely responsible for the circadian rhythm in heart rate in vivo.
DISCUSSION
For the first time, we show here that: (i) there is a circadian or daily rhythm in the intrinsic heart rate set by the pacemaker of the heart, the sinus node, (ii) there is a functioning circadian clock in the sinus node, (iii) there is a circadian rhythm in the expression of a variety of cardiac ion channels, including Hcn4 and its corresponding ionic current, If, in the sinus node, (iv) Hcn4 transcription is directly controlled by the clock transcription factor, BMAL1, (v) the circadian rhythm in Hcn4 plays an important role in the circadian rhythm in the intrinsic heart rate.
Local circadian clock in heart
Previously, it has been demonstrated that there a functioning circadian clock in the hearte.g.41. This is the first report of a functioning circadian clock in the sinus node: qPCR showed the presence of transcripts for many key circadian clock components and many showed the expected circadian rhythm and expected phase relationship; for example, Bmal1 was downregulated, but Cry2, Per1 and Per2 were upregulated at ZT 12 compared to ZT 0 (Figure 2C). We also showed a circadian rhythm in Per1 using a bioluminescent reporter gene (Figure 2F).
Circadian rhythm in cardiac ion channel expression is common
Here we show a circadian rhythm in Hcn4, Nav1.5 (Scn5a), Kv1.4 (Kcna4), Kv4.2 (Kcnd2), Kv1.5 (Kcna5), ERG (Kv11.1; Kcnh2), Kir6.1 (Kcnj8), Kir6.2 (Kcnj11), SUR1 (Abcc8), SUR2 (Abcc9), TASK-1 (K2p3.1; Kcnk3) and Cl− voltage-gated channel 2 (Clcn2) in the sinus node (Figure 2G). Further experiments have shown that Hcn1 also shows a circadian rhythm in the sinus node (Yanwen Wang, unpublished data). Curiously, Hcn2 has been reported to show a circadian rhythm in liver42. A circadian rhythm in cardiac ion channel expression is not only seen in the sinus node – it is likely to be occurring throughout the heart. In atrial muscle, a circadian rhythm has been reported in Kv4.2 (Kcnd2)41,43,44, KChIP2 (Kcnip2)41,44, Kv1.5 (Kcna5)41,43, TASK-1 (K2p3.1; Kcnk3)41, Cx40 (Gja5)45 and Cx43 (Gja1)45. In ventricular muscle, a circadian rhythm has been reported in Nav1.5 (Scn5a)46, Cav1.2 (Cacna1c)47, Kv4.2 (Kcnd2)41,43,48, KChIP2 (Kcnip2)41,44, Kv1.5 (Kcna5)41,43, ERG (Kv11.1; Kcnh2)48, TASK-1 (K2p3.1; Kcnk3)41, Cx40 (Gja5)45 and Cx43 (Gja1)45. Some circadian varying ion channels are common to the sinus node and the atrial or ventricular muscle: Nav1.5 (Scn5a), Kv4.2 (Kcnd2), Kv1.5 (Kcna5), ERG (Kv11.1; Kcnh2) and TASK-1 (K2p3.1; Kcnk3). Of these, all were significantly more highly expressed at ZT 12 than ZT 0 in the sinus node (Figure 2G); in atrial and ventricular muscle, there was a qualitatively similar circadian rhythm in Nav1.5 (Scn5a)46, Kv4.2 (Kcnd2)41,43,48, Kv1.5 (Kcna5; in ventricular muscle, but not atrial muscle)41,43 and ERG (Kv11.1; Kcnh2)48. However, the opposite circadian rhythm in TASK-1 (K2p3.1; Kcnk3) has been reported in atrial and ventricular muscle41.
This study has demonstrated that Hcn4 is likely to be under the control of BMAL1 generated by the local clock (Figure 6). In the ventricle, it has been reported that: Nav1.5 (Scn5a) and ERG (Kv11.1; Kcnh2) are under the control of CLOCK:BMAL1 generated by the local clock46,48; and Kv4.2 (Kcnd2), Kv1.5 (Kcna5), TASK-1 (K2p3.1; Kcnk3), Cx40 (Gja5) and Cx43 (Gja1) are under the control of the suprachiasmatic nucleus (not the local clock) via the autonomic nervous system41. Confusingly, KChIP2 (Kcnip2) has been reported to be under the control of a BMAL1-dependent (i.e. local clock-dependent) oscillator, krüppel-like factor 15 (Klf15)44, and yet under the control of the suprachiasmatic nucleus (not the local clock) via the autonomic nervous system41.
Circadian rhythm in intrinsic heart rate
Cardiac-specific knockout of Bmal1 abolished the normal circadian rhythm in the intrinsic heart rate (Figure 5A,B), proving that it is under the control of the local circadian clock (however, it is interesting that some type of rhythm, albeit abnormal, remained after cardiac-specific knockout of Bmal1 – Fig. 5B). We measured a circadian rhythm in HCN4 protein and If (Figures 3B-E, 4 and S2) roughly in phase with the circadian rhythm in the intrinsic heart rate (in contrast Hcn4 mRNA was out of phase – Figure 3A; Table S1). It is likely that the circadian rhythm in If plays an important role in the intrinsic heart rate: computer modelling suggested that the changes in If are sufficient to explain the changes in intrinsic heart rate (Figure 4G); block of HCN4 and If by Cs+ abolished the circadian rhythm in the intrinsic heart rate (Figure 5A); and cardiac-specific knockout of Bmal1 abolished the normal circadian rhythm in Hcn4 mRNA and protein, If and the effect of Cs+, as well the intrinsic heart rate (Figures 3E, 5B and 6C,E,F). However, it is likely that If is not the only mechanism involved: there was a circadian rhythm in CaMKIIδ (Camk2d) and Ca2+ sparks (Figures 2G and S5), and incapacitating the Ca2+ clock with ryanodine eliminated the circadian rhythm in the intrinsic heart rate (Figure S6A). In addition, it is possible that the circadian rhythm detected in various K+ channels, particularly ERG (Kv11.1; Kcnh2) (Figure 2G), also plays a role; in the sinus node, ERG (Kv11.1; Kcnh2), responsible for the rapid delayed rectifier K+ current, IK,r, sets the maximum diastolic potential and blocking IK,r abolishes pacemaking49.
Circadian rhythm in heart rate in vivo – is it multifactorial?
This study has demonstrated a circadian rhythm in the intrinsic heart rate. Implications of this for the circadian rhythm in heart rate in vivo can be speculated on. Previously, the circadian rhythm in heart rate in vivo has been attributed to the autonomic nervous system and in particular to high vagal tone during the sleep period4. This primarily has been based on heart rate variability as a surrogate measure of autonomic tone5; however, as discussed above, this cannot be used in any simple way as a measure of autonomic tone8. We report that autonomic blockade utilising high concentrations of atropine and propranolol abolished the circadian rhythm in heart rate in vivo (Figure 7E). Abolition of the circadian rhythm in heart rate in vivo on autonomic blockade has also been reported by Tong et al.41 However, Oosting et al.50 have reported that autonomic blockade does not abolish the circadian rhythm in heart rate. Furthermore, caution must be used in interpreting the effect of autonomic blockade involving atropine and propranolol, because propranolol is non-specific and also blocks Nav1.5 (Scn5a) and INa and HCN4 and If (Figures 7F and S8)37,38. Because If at least is greater during the awake period, propranolol will be expected to have a greater depressing effect on heart rate during the awake period. This is an alternative explanation of why the circadian rhythm in heart rate in vivo was abolished by autonomic blockade in this study when utilising high concentrations of atropine and propranolol (Figure 7E). Nevertheless, if it is assumed that the autonomic nervous system is involved, the right vagus nerve (primarily, but not exclusively, responsible for the innervation of the sinus node30) cannot be solely responsible, because sectioning the right vagus did not impact the circadian rhythm in heart rate in vivo in the rat (Figure 7C). This suggests that a circadian rhythm in sympathetic nerve activity or plasma catecholamine levels is more important (if it is assumed that the autonomic nervous system is involved). However, it has been reported that transgenic knockout of the M2 receptor or β1, β2 and β3 adrenoceptors has no effect on the circadian rhythm in heart rate in vivo51. In addition, cardiac transplant patients with autonomic denervation have a preserved nocturnal bradycardia 7-36 months after transplantation52,53. This work does not support a role for the autonomic nervous system in the circadian rhythm in heart rate in vivo.
This study has shown that there is a circadian rhythm in the intrinsic heart rate of the appropriate amplitude (72-107 beats/min) and phase (peaking at ~ZT 12; Figures 2A and 5A; Table S1) to be able to explain the circadian rhythm in heart rate in vivo (Figure 1). Furthermore, block of HCN4 and If by ivabradine had a bigger effect on heart rate during the awake period and abolished the circadian rhythm in heart rate in vivo in the mouse (Figure 5C). Data consistent with this have been obtained by Ptaszynski et al.54 and Grigoryan et al.55 who studied the effect of ivabradine on the heart rate of patients either with inappropriate sinus node tachycardia or ischaemic heart disease and heart failure; whereas ivabradine caused a large decrease in heart rate during the day, it caused little decrease at night. In other words, the effect of ivabradine on heart rate showed a circadian rhythm and was greater when the subjects were awake. This is consistent with the work on the mice (Figure 5C,D)54. However, it is known that cardiac-specific knockout of Bmal146 or cardiac-specific expression of a dominant negative Clock mutant56 does not abolish the circadian rhythm in heart rate in vivo (although it does reduce it). Furthermore, we show here that cardiac-specific knockout of Bmal1 abolishes the normal circadian rhythm in Hcn4 (mRNA and protein), If and intrinsic heart rate (Figures 3E, 5 and 6C,E,F). Therefore, HCN4 and If cannot be the only mechanism controlling the circadian rhythm in heart rate in vivo – perhaps the reduction of the circadian rhythm in heart rate in vivo46 on cardiac-specific knockout of Bmal1 represents the contribution of the circadian rhythm in the intrinsic heart rate to the circadian rhythm in heart rate in vivo. What is responsible for the circadian rhythm in heart rate in vivo that remains after cardiac-specific Bmal1 knockout? This could be due to the autonomic nervous system. Consistent with this, Figure 7D shows that the circadian rhythm in heart rate in vivo was reduced in amplitude after autonomic blockade (achieved with the lower doses of atropine and propranolol). However, the contribution may not be in the way originally conceived (acute control of heart rate via changes in ionic conductance): Tong et al.41,45 have shown that the circadian rhythm of Kv4.2 (Kcnd2), KChIP2 (Kcnip2), Kv1.5 (Kcna5), TASK-1 (K2p3.1; Kcnk3), Cx40 (Gja5) and Cx43 (Gja1) in the atria and ventricles is lost after autonomic blockade suggesting transcriptional regulation mediated by the autonomic nervous system. It is interesting that β-agonists affect both Bmal1 and Per2 in heart57,58.
Summary
In summary, this study has shown that there is a circadian rhythm in the intrinsic heart rate and this is likely to contribute to the circadian rhythm in heart rate in vivo. Our findings provide new mechanistic insight into the fundamental question of why the heart rate of a mammal is lower when asleep and also explains the nocturnal occurrence of bradyarrhythmias1,3,59–63.
METHODS
Extended methods are given in the Supplementary Information.
Animals
The telemetry study was approved by the Norwegian Council for Animal Research, in accordance with the Guide for the Care and Use of Laboratory Animals from the European Commission Directive 86/609/EEC. Ethical approval for vagotomy in rats was from University College London, in accordance with the UK Animals (Scientific Procedures) Act 1986. All remaining procedures were approved by the University of Manchester and were in accordance with the UK Animals (Scientific Procedures) Act 1986. Experiments were conducted on adult male mice: C57BL/6J mice obtained from Harlan Laboratories; Per1∷LUC mice (in which luciferase expression is driven by the mouse Per1 promoter and 5’-UTR elements64) on a Cry1+/+Cry2+/+ or Cry1−/−Cry2−/− background; and cardiac-specific Bmal1 knockout mice.
Electrophysiology
ECGs or ECG-like electrograms were recorded for measurement of heart rate: from conscious unrestrained mice using either subcutaneously implanted radio telemetry transmitters or an ECGenie; from isofluorane-anaesthetised mice using a conventional three-lead system; from isolated Langendorff-perfused hearts; and from isolated right atrial preparations encompassing the sinus node. In the Langendorff-perfused hearts, a suitable pacing protocol was also used to determine the corrected sinus node recovery time. Strips of sinus node tissue were dissociated into single cells by a standard enzymatic and mechanical procedure and If was investigated using the patch clamp technique in the whole-cell mode. A previously developed biophysically-detailed mathematical model of the mouse sinus node cell action potential65 was used to assess the effect of changes in the density of If on the spontaneous action potential.
mRNA and protein expression
RNA was isolated from frozen 1 mm sinus node punch biopsies and gene expression measured using quantitative PCR (qPCR) either using medium throughput custom-designed Taqman Low Density Array Cards or individual SYBR green assays. HCN4 protein was investigated using either sinus node homogenates and western blotting or tissue cryosections and immunohistochemistry. Per1 expression was also measured using the Per1∷LUC mouse: total bioluminescence was recorded for 96 h from freshly dissected intact sinus node preparations from Per1∷LUC mice using a photomultiplier tube assembly.
In vitro chromatin immunoprecipitation (ChiP)
ChIP was performed on 3T3-L1 cells using the SimpleChIP chromatin immunoprecipitation kit according to the manufacturer’s instructions. Prior to cross-linking, 3T3-L1 cultures at a density of 104 cells/cm2 were transfected with plasmid containing His-tagged Bmal1 for 48 h. Antibody directed against the His-tag was used for immunoprecipitation. DNA obtained from ChIP was analysed by qPCR using primers mapping to canonical E-box binding sites, i.e., consensus binding sites for the CLOCK∷BMAL1 heterodimer. E-box binding sites on the HCN4 gene and in 20 kb of the 5′ flanking region were obtained using the RVISTA function within ECR browser (http://ecrbrowser.dcode.org/). Quantification of Hcn4 E-box binding sites was conducted by, first, normalisation to the housekeeping genes Gapdh and L7 and, secondly, normalisation to data from 3T3-L1 cells subjected to transfection treatment without plasmid
AUTHOR CONTRIBUTIONS
Mark R. Boyett, Alicia D’Souza and George Hart conceived the project. Mark R. Boyett, Alicia D’Souza, Hugh D. Piggins, Halina Dobrzynski and Elizabeth Cartwright obtained funding for the project. Anne Berit Johnsen, Alicia D’Souza, Ulrik Wisloff, Eleanor Gill, Charlotte Cox, Yanwen Wang and Nick Ashton were involved in biotelemetry recordings. Yanwen Wang carried out electrophysiological experiments with assistance from Dario DiFrancesco and Annalisa Bucchi in patch clamp recordings. Alicia D’Souza carried out transcriptional profiling, bioluminescence recording and autonomic block experiments with assistance from Cali Anderson, Sven Wegner and Sunil Logantha. Alicia D’Souza, Yu Zhang and Yanwen Wang carried out western blotting experiments. Yanwen Wang and Nicholas Black performed immunofluorescent labelling studies. Haibo Ni and Henggui Zhang performed computer modelling studies using experimental data collected by Yanwen Wang. Yanwen Wang was responsible for the generation, breeding and maintenance of transgenic animals and all experiments on these animals (with assistance from Cheryl Petit, Hugh D. Piggins and Alicia D’Souza). Serve Olieslagers, Alicia D’Souza and Paula da Costa Martins were responsible for chromatin immunoprecipitation experiments. Svetlana Mastitskaya performed vagotomy experiments. Mark R. Boyett and Alicia D’Souza produced the first version of the manuscript, although subsequently all authors contributed.
SOURCES OF FUNDING
This work was supported by the British Heart Foundation (RG/11/18/29257; PG/15/16/31330) and the CARIPLO Foundation (ACROSS 2014-0728).
COMPETING INTERESTS
None declared.
ACKNOWLEDGEMENTS
We thank Ms Rayna Samuels for technical assistance.
Footnotes
↵† Joint senior authors