Spontaneous beating of SAN myocytes is initiated, sustained, and regulated by a complex coupled system of cellular “clocks” that integrates ion channels and transporters on the cell membrane surface or “voltage clock,” with subcellular Ca²⁺ handling machinery, also referred to as an intracellular “Ca²⁺ clock” (8, 36, 37) (Figure 3). The firing of SAN cells is due to diastolic depolarization, a slow depolarizing phase of the membrane potential (V_m), mediated by the concomitant action of both membrane and Ca²⁺ clocks. Since SAN cells lack I_K1 current expression (41), following the minimum, or most hyperpolarized diastolic potential, potassium I_K current (I_Ks, and I_Kr) conductance decreases, allowing the inward hyperpolarization-activated current (I_f) (42, 43) and a low-threshold, voltage-gated T-type Ca²⁺ current (I_Ca,T), which contribute to the early fraction of diastolic depolarization (Figure 3) (44). In addition, L-type Ca_v1.3 Ca²⁺ channels open during diastolic depolarization to generate an inward Ca²⁺ current (45, 46) and enabling the sustained inward Na⁺ current I_st (47). Local Ca²⁺ release (LCR) from the sarcoplasmic reticulum (SR) via subsarcolemmal ryanodine receptors (RyRs) generates small increments in intracellular Ca²⁺ concentration. These LCRs activate Na⁺/Ca²⁺ exchange (NCX) to pump Ca²⁺ out of the cell in exchange for Na⁺ ions at a ratio of 1 Ca²⁺:3 Na⁺, to generate an inward NCX current (I_NCX), and this contributes to both early and late phases of diastolic depolarization (48) and subsequent depolarization of the V_m to the threshold of the next beat (Figure 3). The exact molecular mechanisms responsible for SR Ca²⁺ release during late diastole are not completely understood. While some studies show that such local Ca²⁺ release events are spontaneous, independent of transmembrane potential, and likely include stochastic opening of hyperphosphorylated RyRs (49–51), other evidence suggest that these events might be triggered by Ca²⁺ entry via low-voltage activated T-type Ca²⁺ channels (52) or Ca_v1.3 L-type Ca²⁺ channels (46). Particularly, recent studies indicate that Ca_v1.3 channel activity contributes to generation and synchronization of diastolic LCRs (46) and that Ca_v1.3 is necessary for the Ca²⁺ clock function during SAN firing (53). Overall, the sum of I_f, I_Ca,T, I_NCX and Ca_v1.3-mediated L-type Ca²⁺ current (I_Ca,L) contributes to diastolic depolarization required to ultimately trigger activation of cardiac Ca_v1.2-mediated I_Ca,L that initiates the AP, and global Ca²⁺-induced Ca²⁺ release. In nature, neither clock functions in the absence of the other. Abundant evidence indicates that functional interactions between the two clock components are critical for normal SAN automaticity (8, 36, 37, 46).

There has been significant interest in determining the underlying cellular and molecular mechanisms responsible for intrinsic SAN mechanosensitivity. Mechanical modulation of SAN pacemaking adds another level of complexity to SAN automaticity that has been proposed by Quinn and Kohl as the additional “mechanics-clock” (6, 7) or, more accurately, as a third coupled oscillator. The authors specifically highlighted that in case of fast, beat-to-beat changes in heart rate, the voltage and Ca²⁺ clocks do not inherently account for the rapid response of the SAN to changes in hemodynamic load and that another set of mechanisms must contribute to spontaneous diastolic depolarization of the SAN. The importance of “mechanics-clock” could be further supported by the fact that stretching of quiescent tissue frequently induces spontaneous activity. In particular, arrhythmic isolated hearts of Prosobranch gastropod become rhythmic when the pacemaker tissue is stretched by with internal perfusion and improve in form as pressure is increased (25). It may be more accurate to describe mechanical modulation of SAN pacemaking as an additional coupled oscillator since it could be applied to both fast and slow changes in heart rate through coupling with fast and slow mechanical oscillators, respectively.

Since ion channels are both central for the regulation of SAN automaticity and can sense mechanical stimuli via various mechanisms, they provide plausible molecular candidates for SAN mechanosensitivity. Ion channels may be grouped into two categories with respect to mechanosensitivity: (1) directly mechanoactivated (fast) and (2) indirectly mechanoresponsive (slow) (Figure 1). Although both categories of channels can change their open probability and other biophysical characteristics in response to mechanical stimulation, they differ in how mechanical forces transduce these effects. Fast directly mechanoactivated ion channels (Piezo1-2, TREK-1, TRAAK, and BK channels) are intrinsically sensitive to mechanical forces applied to the protein or to the lipid bilayer in which the channel resides and do not require any other associated proteins or protein complexes to confer mechano-responsiveness (54, 55). Slow indirectly mechanoresponsive ion channels are polymodal ion channels (TRP channels, SWELL1/LRRC8, and ClC) that respond to mechanical forces in cell-type specific contexts but may not themselves be intrinsically mechanosensitive, for example, when reconstituted in a minimal lipid membrane, devoid of other cellular proteins, with some channels displaying faster activation kinetics and some slower.

Piezo1 and Piezo2 ion channels are bona fide mechanoactivated cation channels with established roles in the cardiovascular system (57, 72) and provide another potential mediator of SAN mechanosensitivity. These non-selective cationic channels may be activated by shear stress from nearby blood flow (laminar or turbulent) as well as membrane stretch induced by increased blood pressure, and their activation is highly sensitive to mechanical stimulus variation in frequency and duration (73). In addition, they are non-selective and are therefore permeable to Ca²⁺ and Na⁺, in addition to K⁺, and have a relatively low threshold for mechanoactivation (57, 74). Although there have been no studies directly examining Piezo1 in the SAN, it is expressed in cardiac tissue (54) and appears highly expressed in the murine SAN (Figure 2). Moreover, Piezo1 plays an important role in the regulation of vascular tone (54) and baroreceptor pressure sensing (55). Since Piezo1 channels provide a depolarizing current in response to mechanoactivation, they are good candidates for mechanically activated increases in SAN automaticity and heart rate acceleration (72). Moreover, Piezo1 channels activate rapidly (within milliseconds) and are responsive to phasic, high-frequency mechanical inputs, such as systolic contractions, but may also be modulated by more gradual mechanical inputs (75), such as increases in atrial filling pressures and could therefore govern SAN mechanosensitivity on a beat-to-beat basis and during periods of chronic stretch. Future studies with targeted genetic deletion of Piezo1 from SAN may directly test these hypotheses.

Cardiac cells have two-pore domain potassium currents with little time- or voltage-dependency, also known as background currents, that regulate resting membrane potential and cell excitability (76). The family of cloned mammalian background K⁺ channels includes 14 members encoded by different genes. The members were divided into six subfamilies, TWIK, TREK, TASK, TALK, THIK, and TRESK, on the basis of sequence homology and functional similarities (77). Two-pore domain potassium channels are typically insensitive to conventional K⁺ channel blockers such as 4-AP, TEA, Ba²⁺, Cs⁺, and glibenclamide (76), but they are sensitive to membrane stretch, changes in extracellular or intracellular pH, fatty acids, and inhalation anesthetic agents (e.g., isoflurane) and are regulated by second messenger phosphorylation (76). Structurally, these channels possess four transmembrane domains and two pore domains; each subunit contains two pore-forming domains, so two subunits can form a complete pore of the channels. In murine SAN (Figure 2), two subtypes of two-pore domain potassium channels are mainly expressed: stretch-activated K⁺ channel TREK-1 (78–80) and the acid-sensitive K⁺ channel TASK1 (81–83). Under basal conditions, the activity of the TREK channels is low; however, applying negative pressure to the cell membrane reversibly activates TREK-1 (84). In addition, laminar shear stress stimulates TREK-1, whereas the cell shrinkage induced by extracellular hyperosmolarity reduces the amplitude of TREK-1 (84). Indeed, it has been shown that TREK-1 mechanosensitivity is mediated directly by the lipid membrane perturbations and changes in plasma membrane tension (85). Given its expression in the SAN (Figure 2), it is a candidate contributor to SAN mechanosensitivity on a fast (<1 s) basis.

A number of studies on zebrafish and mice that inhibited plasma membrane trafficking of TREK-1 by inactivating the interacting proteins POPDC1 and POPDC2 revealed exercise- and age-dependent sick sinus syndrome and atrioventricular block (78, 86), suggesting a role for TREK-1 in cardiac automaticity. Similarly, transgenic overexpression of a C-terminal truncation of beta IV spectrin, which also disrupts TREK-1 plasma membrane trafficking, results in sick sinus syndrome (87). These studies provide indirect evidence of TREK-1-mediated effects on SAN automaticity. However, more direct evidence was provided by Unudurthi et al. (69). The authors determined that TREK-1 protein is indeed expressed in both murine and rabbit SAN, and TREK-1-like background currents were reduced in patch-clamped SAN cells isolated from cardiac-specific TREK-1 KO mice (αMHC-Kcnk2^f/f). Also, freely moving, telemetered αMHC-Kcnk2^f/f mice exhibited sinus bradycardia at rest, consistent with studies by Hund et al. (87) where disrupted plasma membrane TREK-1 trafficking induced sick sinus syndrome. Paradoxically, isolated TREK-1 KO SAN cells exhibited increased rather than decreased firing rates as compared with wild-type (WT) SAN. Furthermore, exercise and treatment with epinephrine uncovered stress-induced sinus pauses in αMHC-Kcnk2^f/f mice via unclear mechanisms, or possibly via variation in sympathetic and parasympathetic activity. Finally, intrinsic heart rates measured in telemetered αMHC-Kcnk2^f/f mice with atropine and propranolol treatment exhibited no significant differences, suggesting that neurohumoral inputs are important for TREK-1-dependent regulation of SAN automaticity. These studies illustrate the incomplete understanding of TREK-1 and its contribution to SAN mechanosensitivity, which requires further investigation.

BK (large-conductance Ca²⁺- and voltage-activated K⁺) channels are another promising contributor to SAN mechanosensitivity and automaticity [reviewed in (88)]. These channels are characterized as having large single-channel conductance and selective inhibitors and are regulated by voltage and Ca²⁺ (70). Imlach et al. (89) determined that BK channel inhibition via paxilline caused a reduction in heart rate in isolated mouse and rat hearts but not in hearts from Kcnma1 KO mice. This finding was confirmed at a cellular level when Lai et al. (70) demonstrated that paxilline applied directly to isolated mouse SAN cells reduced AP firing rate in WT mice but not in Kcnma1 KO mice. Lastly, Zhao et al. (90) found that BK channels are mechanosensitive to a small degree, showing an increased activity in chick ventricular myocytes plated on stretched extracellular matrix. Given this and their expression in murine SAN (Figure 2), BK channels are a putative contributor to SAN mechanosensitivity. In this case, SAN membrane depolarization, increases in cytosolic Ca²⁺, and mechanical stimulation from SAN/atrial systole all coincide to activate BK channels after the peak of the SAN AP to augment AP repolarization, re-initiation of diastolic depolarization, and heart rate acceleration (Figure 3). Based on this model, mechanoactivation of BK channels must be relatively rapid to contribute to SAN AP repolarization, as published data suggest; however, it remains unclear if these channels are rapidly or slowly mechanoactivated in SAN.

Putative candidates for stretch-responsive non-selective cation channels include TRP channels expressed in murine SAN: TRPC1, TRPC3, TRPM4, TRPM7, TRPV2, and TRPV4 (Figure 2). A number of TRP channels that we found expressed in murine SAN have been described as mechanoresponsive either directly or indirectly (29, 91). However, thus far, only TRPM4 and TRPM7 have been studied in the context of SAN function. TRPM4 is an intracellular Ca²⁺-activated, non-selective cation channel, which is possibly indirectly mechanoresponsive (29). At negative membrane potentials, TRPM4 activation allows Na⁺ influx, leading to the membrane depolarization, whereas, at the positive membrane potentials, TRPM4 allows K⁺ efflux, leading to membrane repolarization (92, 93) (Figure 3). In rodent SAN, TRPM4 is thought to contribute to diastolic depolarization and a positive chronotropic response in response to stretch (64, 65, 94). TRPM7, an ion channel and protein kinase (chanzyme), permeable to both divalent cations, including Zn²⁺, Mg²⁺, and Ca²⁺, as well as monovalent cations such as Na⁺ and K⁺ (95, 96), is broadly expressed. TRPM7 is highly expressed in murine SAN at the mRNA level (Figure 2) and generates a robust current in both SAN and atrioventricular node cells (66). Both cardiac- and SAN-targeted TRPM7 deletion impaired cardiac automaticity (67); however, the mechanism was proposed to be via regulation of HCN4 and I_f current rather than a direct effect on diastolic depolarization via channel activity (66, 67). While it is clear that none of these TRP channels are intrinsically mechanoactivated (97, 98), it is possible that some of these channels form part of a mechanosensory system (29) and therefore may be mechanoresponsive within specific cellular contexts (99). Testing these hypotheses would require directly measuring these mechanoresponsive currents in isolated SAN, as performed by Kohl et al. (4, 100), in genetic knock-outs of each of these putative mechanoresponsive channels or using specific pharmacologic inhibitors.

Another mechanoresponsive ionic current that has been implicated in the regulation of SAN automaticity is I_Cl,SWELL or the swell-activated chloride current. This ionic current may be carried by VRAC or ClC ion channels, both of which are most commonly activated by cell swelling, which is typically achieved by applying hypotonic or hypo-osmolar solution to cells. However, in a few studies, anion or chloride conductances were demonstrated by application of mechanical forces, as described in further details below. VRACs are activated by cell swelling, ubiquitously expressed in various mammalian cell types and thought to be implicated in many physiological and pathophysiological processes, including fluid secretion, glutamate release, membrane potential regulation, and apoptosis [summarized in the review article (101)]. Although the biophysical properties of VRACs have been well-characterized in multiple cell types over the course of decades, the molecular identity of VRAC remained a mystery until the Patapoutian (58) and Jentsch (102) groups identified leucine-rich repeat containing 8a (LRRC8a, also known as SWELL1) as a required component of a heterohexameric channel complex consisting of various stoichiometries of LRRC8a, and/or LRRC8b,c,d,e. Although the function of the VRAC current has been attributed to cell volume regulation in response to relatively non-physiological osmotic gradients, the broad tissue expression pattern of LRRC8 proteins and presence of VRAC/I_Cl,SWELL current in a multitude of cell types (103–108), including cardiac myocytes (109–113) that are rarely exposed to hypotonic swelling, suggests that the actual physiological role of VRAC and LRRC8 channels remains unknown. Indeed, experiments using magnetic dynabeads bound to monoclonal antibodies for beta1-integrins demonstrated activation of I_Cl,SWELL in cardiac myocytes in response to mechanical force applied via magnetic fields (109, 110), supporting the notion that I_Cl,SWELL is mechanoresponsive in cardiac cells. Therefore, given the high mRNA counts of LRRC8a (SWELL1) and associated subunits LRRC8b,c,d in murine SAN relative to HCN4 (Figure 2), the contribution of SWELL1-mediated I_Cl,SWELL to pacemaker activity and response to stretch warrants further investigation.

Elegant studies by Hagiwara et al. (114) demonstrated that mechanical inflation of isolated rabbit SAN cells using positive pressure via the patch-pipette in whole-cell configuration induces an outwardly rectifying, stretch-activated anion current that is inhibited by chloride channel blockers, 4,4′-diisothiocyano-2,2′-stilbenedisulfonic acid (DIDs) and 9-anthracenecarboxylic acid (9-AC). Also, this current exhibits a shift in reversal potential consistent with a chloride conductance (115) and has a sequence of anion permeability (I⁻ > NO3- > Br⁻ > Cl⁻ > F⁻) similar to VRAC or LRRC8 channels. I_Cl,SWELL activates over the course of minutes (~2 min), which implies responsiveness to tonic changes in membrane tension, as may be expected from gradual atrial stretch-associated increased venous return, but relatively unaffected by phasic changes associated with beat-to-beat changes. Based on the outwardly rectifying current–voltage relationship, and reversal potential around the Cl⁻ reversal potential (E_Cl = −30 mV), we speculate that inward chloride current at voltages below −30 mV may contribute to diastolic depolarization and SAN automaticity, while outward current at voltages above −30 mV may contribute to SAN AP shortening (116) (Figure 3). The integrated effects on automaticity and the response to stretch, however, are likely to be complex.

Similarly, Decher et al. found in guinea-pig atrial myocytes that I_Cl,SWELL induced by osmotic swelling leads to a shortening of AP duration that was inhibited by DCPIB (a relatively selective I_Cl,SWELL inhibitor) (117). Furthermore, Seol et al. found the I_Cl,SWELL can be activated by axial stretch in cardiomyocytes isolated from the pulmonary vein (30, 59); and Egorov et al. determined that I_Cl,SWELL activation in response to mechanical stretch can depolarize resting membrane potential, generate arrhythmic substrates, and confirm that it can also shorten APs (59). In isolated rabbit SAN tissue, Arai et al. also showed that application of various non-specific anion channel blockers that can block VRACs, such as DIDs, caused a reduction in the stretch-induced increase in firing rate at a high level of distension (118). On the other hand, Cooper et al. (56) reported that the application of 9-AC at 1 mM concentration had no effect on the stretch-induced increase in heart rate when a significant stretch-stimulus was applied, suggesting that I_Cl,SWELL may not underlie the SAN response to mechanical stretch. However, application of such high concentrations of 9-AC is highly non-specific and therefore complicates the interpretation of this result. Furthermore, the use of different stretching techniques between the two studies may account for the differences observed. These studies, albeit contradictory, indicate the potential role of I_Cl,SWELL in modulating SAN function on a slow, non-beat-to-beat basis, which may be present during periods of chronic stretch, and demonstrate the need for additional experiments. Since SWELL1 (LRRC8a) and LRRC8 subunit proteins are now known to encode I_Cl,SWELL in numerous other cell types (103–108), future studies examining cardiac specific and SAN targeted Swell1 KO mice, transient knockdown in isolated cells, and/or more specific small molecules such as DCPIB will provide important new insights into the contribution of I_Cl,SWELL and VRAC in cardiac automaticity and the response to SAN stretch.

Other candidates for the molecular identity for I_Cl,SWELL in SAN are the ClC ion channels. While both ClC-2 and ClC-3 have been studied in cardiac myocytes (26, 62), and ClC-3 has a high mRNA count in murine SAN (Figure 2), only ClC-2 has been directly studied with respect to regulating SAN automaticity. Huang et al. (26) showed that inwardly rectifying chloride current induced by osmotic swelling in isolated guinea-pig SAN pacemaker myocytes could be blocked though intracellular dialysis of anti-ClC-2 antibody, which did not affect other pacemaker currents, including I_f, I_Ca,L, and I_Ks and the volume-regulated outwardly rectifying Cl⁻ current (I_Cl,vol). Anti-ClC-2 antibody reversed the changes in SAN APs induced by osmotic swelling. The authors also showed that ClC-2 KO (ClCN2^−/−) mice demonstrate a decreased chronotropic response to acute exercise stress when compared with their age-matched ClCN2^+/+ and ClCN2^+/− littermates. It was then concluded that targeted inactivation of ClC-2 does not alter intrinsic heart rate but prevented the positive chronotropic effect of acute exercise stress through sympathetic regulation of ClC-2 channels. While ClC-2 channels may contribute in part to cardiac I_Cl,SWELL, there have been no studies examining the signaling mechanisms underlying ClC-2 mechanoresponsivity.

ClC-3, on the other hand, has been proposed to be mechanoresponsive in osteoblasts (119) and has been shown to be expressed in cardiac myocytes, to underlie I_Cl,SWELL, and to be involved in numerous pathophysiological processes, including ischemic preconditioning, myocardial hypertrophy, and heart failure (120). However, there have yet to be any studies directly examining ClC-3 in SAN cells, and neither global nor cardiac specific ClC-3 KO mice were noted to show differences in heart rates (62).

Other possible contributors to SAN mechanosensitivity are calcium-activated chloride channels (CaCCs) such as anoctamin1 (ANO1) and chloride channel accessory 2 (ClCA2). Ye et al. (121) determined that ANO1 is expressed in mouse ventricular myocytes and facilitates accelerated AP repolarization. Given that ANO1 is implicated to be mechanoresponsive (121) and expressed in murine SAN (Figure 2), it is plausible that ANO1 may contribute similarly to shorten pacemaker potentials, as Sung et al. speculated (122). In addition, Mao et al. found that ClCA2 is highly expressed in SAN tissue and, when mutated, induces mild conduction block and ectopic pacemaker activity (71). While no study has examined ClCA2 mechanosensitivity, given its calcium-activated properties, it is likely to be affect by pressure-induced calcium transients (123). Given these findings, it is feasible that calcium-activated chloride channels could play a partial role in the response of SAN beating rate to stretch.

Mechanical modulation of Ca²⁺ spark activity was linked to stretch-induced activation of ROS signaling that is also graded in a stretch-dependent manner (142). The stretch-induced NOX2-dependent ROS response sensitizes RyR to Ca²⁺ possibly through direct oxidation but may also do so indirectly via oxidation of calmodulin, displacing it from the RyR and promoting activation (143) or via RyR phosphorylation by oxidized CaMKII (144). These pathways of mechano-transduction are termed X-ROS signaling and require an intact microtubule network and functions independently of stretch-activated channels (SACs) and transsarcolemmal Ca²⁺ influx (33, 137). Furthermore, X-ROS signaling is confined to dyads (the cytosolic space between the sarcolemmal and SR membranes) (145) and has been proposed to be an important regulator of beat-to-beat adaptation to hemodynamic load in working cardiomyocytes (142). However, these pathways have not been confirmed specifically in SAN myocytes, but they are known to contain the necessary components (146).

In addition to regulation of Ca²⁺ signaling, the X-ROS pathway has also been found to be involved in the modulation of mechanoresponsive ion channels as well (147). Patch-clamp studies on stretched ventricular myocytes revealed NOX-dependent modulation of SACs (148), and this modulation may be facilitated by co-localization of NOX2 and SAC in caveolae (149, 150). Although SACs are not involved in X-ROS signaling per se, these channels may contribute to stretch-induced modulation of AP as discussed above for mechano-electrical signal transduction. Additionally, I_Cl,SWELL activated by osmotic swelling has been found to be controlled by an angiotensin II-dependent ROS cascade that is previously implicated by integrin stretch (113). This is consistent with persistent activation of I_Cl,SWELL and ROS present in several models of cardiac disease. Furthermore, Gradogna et al. demonstrated that LRRC8 channel subunits and their currents are differentially modulated by oxidation depending on LRRC8 channel subunit composition (151). Given that inflammation and oxidation are present in the setting of hypertension, it is possible that SAN mechanosensitivity could differ from other cardiac regions due to varying SWELL1 subunit expression (151).

Nitric oxide synthase (NOS) also plays a discrete role facilitating cardiac stretch as Petroff et al. demonstrated that stretch increases nitric oxide (NO) production with concurrent increases in Ca²⁺ spark frequency and transient amplitudes (33). Pharmacological inhibition or genetic deletion of both neuronal NOS (nNOS) and endothelial NOS (eNOS) demonstrates that subtypes contribute to the increase of systolic Ca²⁺, but only nNOS participated in the afterload induced Ca²⁺ sparks (152). Due to the short lifetime of NO, its effective signaling range is limited and dependent on the diffusion distance, amount produced, and the buffer capacity of the cell (33). Therefore, one possible explanation for the distinct effects of nNOS vs. eNOS-derived NO on Ca²⁺ sparks is their different subcellular localizations. While eNOS is localized at the caveolae (153, 154), nNOS is preferentially localized at the SR membrane in the vicinity of RyR, and nNOS increases RyR Ca²⁺ leak, directly by S-nitrosylation or indirectly via CaMKII (155). In addition, nNOS facilitates SERCA Ca²⁺ reuptake (155), which may compensate for the increased SR Ca²⁺ leak and reduced basal I_Ca,L (156). In regard to the SAN, in a study by Vila-Petroff et al. using exogenous NO donors, high levels of NO induced a large increase in cGMP and a negative inotropic effect, while low levels of NO increased cAMP and caused positive inotropy via cGMP-independent activation of adenylyl cyclase (157). Furthermore, it has been shown that inhibition of NOS has a negative chronotropic effect on SAN activity and that NOS activation has an opposite effect, indicating that SAN function is somewhat dependent on NOS activity (158, 159). However, unlike X-ROS signaling, NO mechanosensitivity operates on a slower time scale of minutes rather than seconds (33), suggesting that it may play a more significant role in conditions where chronic stretch is a factor (i.e., hypertension and chamber filling pressures).

Another important factor in myocyte stretch signaling is atrial natriuretic peptide (ANP) (160). Similar to X-ROS, ANP is a mechanosensitive signaling factor that is activated by a caveolae and angiotensin II-dependent pathway (161). ANP has been found to enhance reflex bradycardias (162); therefore, it is likely that ANP has a compensatory mechanosensitive role on the SAN, acting to restore it to normal function in response to elevated stretch. ANP has been found to shift midpoint activation of pacemaking I_f current toward less negative potentials (163) and thus accelerate SAN rhythm. ANP is also able to increase intracellular cGMP and cAMP levels (163), which play a crucial role in SAN automaticity via phosphorylation of the calcium clock proteins (50, 164). Indeed, ANP has been identified as a critical regulator of SAN automaticity (38, 165).

IP₃Rs are another type of SR Ca²⁺ release channels, which are activated by IP₃ through the hydrolysis of phosphatidylinositol-(4, 5)-bisphosphate by phospholipase C and thus may also contribute to the LCR generation via hypersensitization of RyRs. They are highly abundant in atrial and SAN myocytes (166–168), and recent studies demonstrated that this signaling pathway might be confined within specific microdomains, including lipid rafts and dorsal ruffles (169). Stimulation of IP₃Rs was found to accelerate spontaneous beating rate of the mouse SAN likely through regulation of Ca²⁺ spark activity and RyR function (170). In rabbit ventricular myocytes, upregulation of IP₃R-induced Ca²⁺ releases was detected and linked to enhanced spontaneous SR Ca²⁺ releases (170). It has been shown that mechanical stretch can directly activate phospholipase C with production of IP₃ (171), which may subsequently modulate SAN automaticity (Figure 4).

While there are numerous mechanochemical signaling factors that may affect SAN automaticity, they are united as facilitators of cardiac mechano-transduction through their association with caveolae membrane structures (Figure 4) (123, 172–174). For example, NOS (173), NOX2-mediated ROS (150), and calcium dynamics (123) are all affected by the presence of caveolae, which are suspected to play an inhibitory role on these factors, which are disrupted by shear stress. Digging deeper, angiotensin II mediates activation of cAMP production (175) and X-ROS through caveolae membrane structures (176), further linking the discussed factors to these structures. The suspected regulation of these signaling factors of SAN automaticity by caveolae may explain the connection between caveolae loss and cardiac pathology (177) as chronic shear stress depletes caveolae (178), allowing these factors to activate constitutively and/or enter unusual feedback loops. For example, as shear stress increases, caveolae flatten and release NOS, which should reduce the initial mechanical stimuli and allow caveolae to reform. However, if the mechanical stimulus is prolonged, membrane caveolae density will decrease, eliminating a crucial regulator of nNOS activity. Without this negative regulation, these signaling factors can enter positive feedback loops, inducing the generation of excess ROS from sarcolemmal and mitochondrial sources that can ultimately lead to changes in myocyte electrophysiology as calcium kinetics are subsequently altered. For these reasons, it is highly plausible that SAN caveolae may regulate downstream signaling factors that are known to alter SAN automaticity and consequently heart rate.