Intracellular remodelling of Ca2+ stores in pulmonary hypertension

This editorial refers to ‘Stretch-induced Ca2+ signalling in vascular smooth muscle cells depends on Ca2+ store segregation’ by G. Gilbert et al., pp. 313–323, this issue. The low resistance of the pulmonary circulation, a function of its unique structure and regulation, normally allows it to accept the entire cardiac output in the pulmonary vascular bed while maintaining an internal pressure that is low enough to prevent pulmonary oedema. In pulmonary hypertension (PH), however, pulmonary vascular resistance can increase enormously due to structural remodelling and abnormalities in the regulation of pulmonary artery (PA) diameter. Both changes are driven, at least in part, by increases in the intracellular Ca2+ concentration ([Ca2+]i) in PA smooth muscle cells (PASMCs). This is associated with the up-regulation of multiple pathways mediating the influx of Ca2+ into these cells.1 [Ca2+]i, however, is not determined by Ca2+ influx alone; there is recent evidence2 that the expression of SERCA2a, which mediates Ca2+ uptake into the sarcoplasmic reticulum (SR), is decreased in PASMCs from humans with PH. An elegant study by Gilbert et al.3 in the current issue of Cardiovascular Research now demonstrates that the abnormalities in intracellular Ca2+ handling in PH go well beyond SERCA; they find that two animal models of PH exhibit a re-organization of intracellular Ca2+ stores of PASMCs leading to a potentiation of the rise in [Ca2+] evoked by Ca2+ influx. In order to appreciate the significance of their findings, it is useful to first briefly review some of what is known about the SR in these cells. Although apparently an interconnected and continuous network of tubules and sacs, the SR has long been known to behave as if it comprises two or more distinct compartments which handle Ca2+ differently (e.g. Yamazawa et al.4). In 1997, Golovina and Blaustein5 used confocal Ca2+ imaging to show that this functional heterogeneity of the SR reflects a spatial compartmentalization in mesenteric artery SMCs. They observed that although most of the SR was depleted of Ca2+ by the SERCA inhibitors cyclopiazonic acid (CPA) and thapsigargin, and also by 5-HT, a small fraction of the SR was in contrast loaded with Ca2+ by these drugs. Moreover, while the larger compartment was insensitive to the RyR activators caffeine and ryanodine, the smaller compartment was depleted of Ca2+ by these drugs. The same year, initial evidence emerged for the presence of two SR compartments acting as distinct Ca2+ stores in PASMCs as well.6 One of these stores, sensitive to ryanodine, was released by hypoxia, whereas the other could take up Ca2+ released from the ryanodine-sensitive store but was emptied by the α-receptor agonist phenylephrine. Subsequent work by a number of groups has established a much fuller picture of the physical and functional compartmentalization of the SR in PASMC. Yang et al.7 observed the presence in transiently cultured PASMCs of subplasmalemmal and perinuclear SR ‘pockets’ which they found to express mainly RyR1/2 and RyR1/3, respectively; interestingly, a fraction of the Ca2+ ‘sparks’ due to release from the centrally located SR was very prolonged, consistent with a role for these events in regulating gene transcription. A series of papers by Evans and colleagues8–10 have also supplied detailed functional and morphological evidence for two SR compartments in freshly isolated cells, PASMCs. One of these is located close to the plasmalemma, is highly enriched with SERCA2b and RyR1, is depleted by CPA, and when released by isoprenaline caused BKCa channel-dependent vasodilatation. The other store, which initiates Ca2+ signalling induced by endothelin-1 (ET-1), is enriched in SERCA2a and RyR3, and is CPA-resistant. It is perinuclear and is in close proximity to lysosomes, allowing it to undergo Ca2+-induced Ca2+ release triggered by NAADP-dependent lysosomal Ca2+ release. They propose that release of Ca2+ by this store triggers Ca2+ waves by activating RyR2 on SR widely distributed through the cell. The study by Gilbert et al. focuses on how the compartmentalization of the SR and other Ca2+ stores in PASMC affects Ca2+ signals initiated by stretch-activated channels, which this group has shown previously are up-regulated in chronic hypoxia (CH)11 and examine how this interaction is altered in two models of PH. They describe in PASMCs from normoxic animals a segregation of RyR and SERCA isoforms between the subplasmalemmal and perinuclear SR, which closely resembles that described by Evans' laboratory8 (Figure 1). They also confirm that RyR1 and RyR3 are involved in activating BKCa and mediating the response to ET-1, respectively. SR spatial and functional compartmentalization in normal PASMCs and changes observed in chronic hypoxic and MCT models of PH. The diagram is based on Gilbert et al.3 and also on work by others.2,7,8–10,12 See text for more details. Remarkably, this arrangement changes when PH is induced with either CH or monocrotaline (MCT). In both models, the mitochondria and lysosomes, which are normally clustered near the plasma membrane and the nucleus, respectively,10,12 assume a more diffuse distribution. In CH-PASMCs, a new subplasmalemmal SR element containing RyR3 and SERCA2a appears alongside the RyR1/SERCA2b compartment. The effect on MCT-PASMCs is more dramatic: the segregation of SERCA and RyR isoforms is completely lost. This finding is reminiscent of the loss of functional SR segregation previously seen when PASMCs are cultured.13 This may reflect the transition of MCT-PASMCs to a proliferative phenotype, but importantly emphasizes the inadvisability of using cultured SMCs to study changes in Ca2+ signalling caused by disease. The authors also show that the stretch-induced Ca2+ signal is enhanced to a similar extent in both PH models. This is associated with the development of a functional association between the SR and caveoli and, particularly in MCT-PASMCs, a much greater amplification by Ca2+ release. Hadri et al.2 notably used SERCA2a gene transfer into PA to cause regression of PH and associated remodelling in MCT-treated rats. However, the current study by Gilbert et al.3 implies that PH is associated with a much more profound dysregulation of the expression, compartmentalization and specific interactions14 between the organelles and intracellular proteins involved in controlling [Ca2+]. Their observations imply that the structural abnormalities of the pulmonary vasculature, which are a hallmark of PH, reflect an equally profound remodelling at the subcellular level and raise important questions: do similar changes occur in any type of human PH, and if so is there a single causative factor which can be identified and targeted for therapeutic purposes?