The smooth muscle cell membrane and peripheral sarcoplasmic reticulum: their interactions are organized and may be crucial

BK or MaxiK channels are nearly ubiquitous in smooth muscle cells. These channels are regulated both by voltage and by the local Ca2+ concentration: depolarization and elevated local [Ca2+]i enhance their opening probabilities. When studied in expressed systems, it is found that [Ca2+]i levels greater than those normally found in resting smooth muscle (approximately 100–150 nmol) are required to activate them unless the transmembrane voltage is much more depolarized than −50 mV [1–6]. For this reason, it has usually been assumed and/or concluded that their role in smooth muscle physiology is to limit the effects of depolarizing stimuli which increase Ca2+ entry or stimuli which release Ca2+ from the sarcoplasmic reticulum. However, there is also evidence that the Ca2+ sensitivity of these channels varies depending on the composition of the channels, clustering and the β1 subunit present [7–10]. Recently, it has become clear that intracellular [Ca2+]i is not homogenous in the cytosol. A variety of evidence shows that the level of [Ca2+]i between the peripheral sarcoplasmic reticulum and the plasma membrane may be elevated by the uptake and subsequent vectorial extrusion of sarcoplasmic reticulum Ca2+ towards the plasma membrane [11–13]. The operation of this system, called the buffer barrier system by Van Breemen and colleagues since the uptake of Ca2+ entering by the peripheral sarcoplasmic reticulum buffers its entrance into the cytosol, results in there being a high enough [Ca2+]i near the plasmalemma to allow the Na+–Ca2+ exchange system to operate in a mode to extrude Ca2+ from the cytosol in those muscles in which it plays such a role; some muscles appear not to use it for extrusion [14]. In some smooth muscles, the buffer barrier function may operate to provide a [Ca2+]i sufficient to activate BKCa2+ channels directly or indirectly and limit tone development. For example, in the lower oesophageal sphincter, which like many arteries normally has sufficient [Ca2+]i to elevate tone, there is membrane-bound neuronal nitric oxide synthase (nNOS) [15,16] which, activated by the local [Ca2+]i concentration normally present, limits tone development by initiating iberiotoxin-, tetraethylammonium- and L-NOARG-sensitive outward currents. Since the dependence of the outward currents on [Ca2+] pipette was found to have an EC50 of approximately 100 nmol Ca2+, it seems unlikely that the general cytosolic [Ca2+] was activating nNOS or BKCa2+ channels directly. In fact, the outward currents were found to depend on activity of L-Ca2+ channels, even when the [Ca2+] pipette was approximately 6 nmol. A further study [17] suggested that there was recycling of Ca2+ between an external pool, associated with the plasmalemma but not freely diffusible into the extra-cellular fluid, and the superficial sarcoplasmic reticulum. This extracellular Ca2+ pool supplied sufficient Ca2+ to the sarcoplasmic reticulum using L-Ca2+ channels in the absence of Ca2+ in the medium to allow carbachol to initiate sustained contractions which were blocked not only by nifedipine, but also by high [EGTA] in the external medium. Similar findings have been made in airway smooth muscle [18,19]. Subsequent studies in the lower oesophageal sphincter and airway muscle suggest that the nNOS, L-Ca2+ channels, the plasmalemma Ca2+ pump and Ca2+ binding proteins were all colocalized with caveolin-1 in a special subdomain of the membrane, the caveolae [20,21]. Some BKCa2+ channels appeared colocalized in this subdomain as well. Na+-Ca2+ channels appear to be also located there in vascular muscle [11,12], Such a location would provide an efficient organization to control [Ca2+]i near the membrane independently of the levels in the cytosol and not affecting contractile proteins directly. Moreover, sarcoplasmic reticulum Ca2+ may be released spontaneously in bursts causing Ca2+-sparks, as observed by Ca2+-imaging. The initiation of this Ca2+ release is still unclear and may differ in different systems. In the original description of these sparks in arterial muscle by Nelson et al. [22,23], the sparks were associated with the opening of BKCa2+ channels and were inhibited by thapsigargin, an inhibitor of the SERCA pump, and by ryanodine, which is presumed to act by affecting Ca2+-release channels on sarcoplasmic reticulum. In the paper by Asano and Nomura, published in this issue of the journal, it was reported that three endothelium-denuded arteries from the spontaneously hypersensitive rat (SHR) (femoral, mesenteric and carotid) were contracted by charybdotoxin (ChTX) to a greater extent than the same arteries from Wistar–Kyoto (WKY) rats. These contractions were increased by depolarization with 10 mmol K+ and were abolished by nifedipine. Treatment with thapsigargin decreased the ChTX contractions, as did a combination of ryanodine and caffeine. Only after addition of 10 mmol K+ did the same arteries from WKY rats contract to ChTX appreciably. An interesting aspect of this report is the finding that Ca2+ to support ChTX contractions comes from L-Ca2+ channels. This may be a result of a lower membrane potential in SHR arteries compared to WKY rat arteries, leading to activation of these channels, Ca2+ entry and activation of BK channels, as Asano and Nomura suggest, but which was not established directly. The fact that elevation of [K+]e increased sensitivity to ChTX in arteries of SHR and WKY rats is consistent with that possiblity. However, recent studies show [6–10] that the β1 subunit of the BK channel controls its sensitivity to Ca2+ and one possibility is that there is a different subunit in arteries of SHR, with a lower threshold for [Ca2+], or that there is different modulation of the channel. The most likely interpretation of the results is that, in these arteries, as in the muscles described above, Ca2+ entry through L-Ca2+ channels limits tone by activating K+-Ca2+ channels, probably BKCa2+ but possibly IKCa2+ (intermediate conductance channels). ChTX inhibits both BK and IK channels [24]. A better choice of BKCa2+ blocker would have been iberiotoxin, which acts highly selectively to block BKCa2+ channels. However, the data also suggest that the Ca2+ entering through L-Ca2+ channels was supplemented by Ca2+ released from the sarcoplasmic reticulum to activate K+-Ca2+ channels. Thapsigargin, a SERCA pump inhibitor, first elevated tone and, when this declined to baseline presumably because sarcoplasmic reticulum Ca2+ had been released and extruded, ChTX was less able to increase tone. The authors consider that the Ca2+ released from sarcoplasmic reticulum is the result of Ca2+-induced Ca2+ release (CICR) based on their findings with ryanodine and caffeine. In skeletal and cardiac muscle [25] and in some smooth muscles, but not all [26–33], a ryanodine receptor (RyR3) exists which responds to lower concentrations of ryanodine by opening to release sarcoplasmic reticulum Ca2+, and to higher concentrations of ryanodine by closing. Caffeine acts to enhance the opening of this channel [25,26], but this is not the only action of caffeine. It also raises cAMP (e.g. [30]). Moreover, knockout of RyR3 does not noticeably impair function except in skeletal muscle and the central nervous system [25]. Although the actions of ryanodine to release Ca2+ from sarcoplasmic reticulum have been well established in whole-cell patch-clamp studies of isolated smooth muscle cells and some Ca2+ imaging studies [26–33], the actions of ryanodine on intact smooth muscle preparations, and whether it acts to enhance or inhibit release of sarcoplasmic reticulum Ca2+ (and at what concentration) or has some other actions [34], are not as well understood as in skeletal muscle. In skeletal muscle, ryanodine receptors and L-Ca2+ channels are colocalized on foot proteins [25] and play a major role in excitation–contraction coupling. Caffeine was used by the authors of this study on the presumption that it potentiates the action of ryanodine to release Ca2+ from the sarcoplasmic reticulum as it does in skeletal muscle. This might have been expected to enhance the contraction from prior ryanodine, but instead it relaxed it in all arteries except the carotid, where it produced an initial phasic contraction before inhibiting. The authors did not establish the mechanism of caffeine relaxation, which may have been the result of inhibition of phosphodiesterases and elevation of cAMP or cGMP. protein kinase A activity has been shown to affect BKCa2+ channel activity [35]. Only after washing several times was the ryanodine response restored. It appears to be uncertain from these data whether CICR was involved or whether spontaneous release of Ca2+ from sarcoplasmic reticulum into a buffer barrier region together with increased Ca2+ entry and/or increased sensitivity of the BK channels to local [Ca2+], contributed to the opening of K+Ca2+ channels. A recent publication [36] reports that, in both arterial muscle and neonatal cardiomyocytes, depletion of caveolae led to a reduction in the frequency, amplitude and spatial size of Ca2+ sparks. However, temporal characteristics of Ca2+ sparks were not significantly affected. The possibility that the decreases in Ca2+ spark frequency and size are caused by changes in dihydropyridine sensitive L-type channels, sarcoplasmic reticulum Ca2+ load or changes in membrane potential was ruled out. These results suggest a novel signalling model that explains the formation of Ca2+ sparks in a caveolae microdomain. The authors concluded that a transient elevation in [Ca2+]i at the inner mouth of a single caveolemmal L-Ca2+ channel, colocalized in caveolae, induces simultaneous activation and thus opens several RyRs to generate a local Ca2+ release event, a Ca2+ spark. It is possible that a precise juxtaposition of peripheral sarcoplasmic reticulum membrane to caveolae containing L-Ca2+ channels is required for Ca2+ sparks to be induced. It has been suggested several times that colocalization of sarcoplasmic reticulum, BK channels and L-Ca2+ channels is required for the spark mechanism, requiring activation of multiple RyR channels by CICR to work [26,29,37]. The mechanism promoting this alignment is not known. However, it may be a general mechanism for organizing signalling in smooth and skeletal muscle [15–17]. Neither that study nor the study by Asano and Nomura examined whether nNOS was also colocalized with caveolae, as it is in some smooth muscles [15–17], and/or whether its activation by entering Ca2+ contributed to the activation of K+Ca2+ channels directly or indirectly [38,39]. NO is known to affect loading of Ca2+ into sarcoplasmic reticulum as occurs in other smooth muscles [40]. A study of the effects of NOS inhibition on responses to BK channel block in arteries of SHR versus WKY rats would be an interesting follow-up. In addition, studies of the subunit composition of BK channels in SHR compared to WKY rat arteries would be of interest, as well as high resolution studies of the [Ca2+]cav near caveolae in both types of arteries. Caveolin multimers in the inner plasmalemma may be the organizing mechanism of a signalling unit in and near the membrane necessary for sparks using peptide domains to which the relevant proteins have high affinity. Therefore studies of the colocalization of caveolins and Ca2+-dependent or -modulated proteins in these arteries and the binding of these proteins to caveolin's cytoplasmic arms would be very helpful. The role of caveolae may also be revealed by studying cardiovascular function in caveolin 1 or 3 knockout mice. To determine if CICR is actually required in the release of sarcoplasmic reticulum Ca2+, studies in animals with knockout of the RyR3 receptor would be useful. Obviously, studies of whole animals, as well as at the cellular or molecular level, will be required. Further study of the control of the mechanisms underlying the buffer barrier and Ca2+ sparks is needed to provide more pathophysiological insight into mechanisms of hypertension and to reveal new therapeutic targets.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.