nitric oxide (NO) plays an important role in intracellular signaling, and stole the limelight in 1992 by winning science's equivalent of an “Oscar”: Science magazine's “Molecule of the Year.” Subsequently, the importance and versatility of its known roles has expanded. Furthermore, this actor now seems equally at home playing in the “local theater” of organelles such as mitochondria because it is in the grand stage of the cytosol. For instance, Dedkova and Blatter (8) have previously demonstrated that NO modulates cytosolic Ca2+ levels by inhibiting capacitative Ca2+ entry from outside the cell, as well as accelerating its reuptake into the endoplasmic reticulum. In their present work (Ref. 9; see p. 836 in this issue), they turn the spotlight on mitochondria and find that NO can modulate Ca2+ levels in mitochondria by inhibiting its uptake via the mitochondrial uniporter, with the possible additional involvement of the mitochondrial permeability transition pore (mPTP).
The first identification of a physiological role for NO came about with the discovery that it was playing the part of the “endothelium-derived relaxing factor” (reviewed in Refs. 5 and 22). NO synthase (NOS) produces NO enzymatically by oxidizing the terminal guanidine nitrogen of l-arginine to form NO and citrulline. There are three mammalian isoforms, namely, the endothelial NOS (eNOS), endotoxin- and cytokine-inducible (iNOS), and neuronal (nNOS) isoforms. Although the highest expression levels of these enzymes are found in the endothelial, blood cell, and neuronal lineages, respectively, other cell types, such as cardiomyocytes, can also produce NO. This NO appears to exert a physiological influence on myocardial contractility, ventricular relaxation, and mitochondrial respiration (reviewed in Ref. 16), and apparently emanates from eNOS, although other isoforms can be expressed as well (22). In fact, heart failure can lead to expression of nNOS in the myocardium and this may have detrimental effects on Ca2+ cycling (16).
The intracellular localization of NOS enzymes is believed to control aspects of their regulation, and the possible existence of a mitochondrially localized NOS (mtNOS) has sparked much interest (1, 14–16). Yet, whether mtNOS is of physiological, or even pathophysiological, relevance has been questioned partly because of the relatively low rates of mitochondrial NO production, and also because the diffusibility and relatively long half-life of several seconds makes it more difficult to conceive of a purpose for highly localized NO production. However, it is possible that the diffusion distance of NO is significantly limited by the crowded intracellular milieu and by the molecule's high innate reactivity. Therefore, concentration of small quantities of NOS enzymes in certain subcellular compartments may result in levels of NO production that affect localized cellular processes (11). NOS activity has been identified in purified mitochondria, as well as by electron microscopy of immunolabeled cells and by electron paramagnetic resonance (1, 14, 15, 18). The isoform of NOS in mitochondria was found to be stimulated by Ca2+, and therefore cannot be nNOS, which is Ca2+ insensitive. Subsequent experiments identified it as eNOS and showed it was located in the inner membrane and matrix (18, 20). In the present study (9), Ca2+ uptake was increased to 128 ± 3% by l-NIO, an inhibitor of eNOS, suggesting that eNOS is in the mitochondria of primary vascular endothelial cells. Interestingly, other studies (20) have shown that mtNOS activity is increased after hypoxia, suggesting that mtNOS may be pathophysiologically relevant.
One form of pathophysiology in which an important role for NO (but not necessarily mtNOS) has been repeatedly observed is that of myocardial ischemia and reperfusion (comprehensively reviewed in Ref. 5). In this case, the source of NO may be the endothelial cells acting in a paracrine manner, or may be intrinsic to the cardiomyocytes. NO has also been implicated in hypertrophy, although the exact role it plays is controversial and may depend on the intensity and location of its production (25). In fact, this is an important and recurring theme: the dose of NO is important, and higher or lower concentrations may even have opposite effects (16).
Significantly, an important component in the pathophysiology of both hypertrophy and ischemia-reperfusion is Ca2+. Cytosolic and mitochondrial Ca2+ levels increase progressively during ischemia, until the intracellular Ca2+ handling capacity is exceeded, leading to cell death (21). Opening of the mPTP (a pore involved in necrosis and possibly apoptosis; see Ref. 7) is believed to be integral to this outcome and is sensitized to opening by high Ca2+ levels. Alterations in Ca2+ homeostasis also contribute to the development of hypertrophy. Obviously, Ca2+ levels in the cytosol and within organelles must be carefully controlled by the cell, and an entire “molecular toolkit” of Ca2+ buffers, pumps, exchangers, etc., exists solely for this purpose (4). However, this raises the question of what operates the toolkit. The current study by Dedkova and Blatter adds weight to the suggestion that NO may be one molecule with a subtle and important action in modulating this toolkit. By repeatedly measuring the mitochondrial accumulation of Ca2+ in permeabilized primary endothelial cells, they were able to determine that prior application of NO reduced mitochondrial Ca2+ uptake rates, and this appeared to be secondary to a decrease in mitochondrial membrane potential. Importantly, they tested a range of NO concentrations and found the effect was much more pronounced at “pathophysiological” (>1 μM) concentrations.
While one might expect NO to modify the rate of mitochondrial Ca2+ uptake, Dedkova and Blatter measured a reduction in the maximum level of Ca2+ accumulated by the mitochondria. This is presumably because the “set point” of mitochondrial Ca2+ concentration is determined by Ca2+ cycling: i.e., a combination of uptake and extrusion rates (7). It has been observed that mitochondrial Ca2+ import has a very steep voltage sensitivity, which suggests that even modest mitochondrial depolarizations could have profound effects on mitochondrial Ca2+ uptake (12), and, hence, levels. On the other hand, Ca2+ uniporter activity plateaus at ∼110 mV in isolated liver mitochondria (7), so unless this value is different for mitochondria in situ, a small drop in mitochondrial membrane potential (ΔΨm) may not be sufficient to affect Ca2+ uptake. However, the effect of NO on ΔΨm is sufficient to induce physiological effects in at least some situations, e.g., treatment with a NO donor reduces mitochondrial potential in isolated cardiomyocytes, and limits Ca2+ overload during ischemia (18, 24). Exposure to NO has previously been shown to dose-dependently protect isolated rat hearts from subsequent ischemia and reperfusion, and application of NO directly depolarizes isolated mitochondria (2).
There are various downstream targets of NO. First, NO directly inhibits several mitochondrial dehydrogenases, and this is likely to be the mechanism by which it reduces ΔΨm. NO may directly nitrosylate some proteins, and also increases cGMP, which activates protein kinase G and leads to the phosphorylation of downstream substrates. NO can react stoichiometrically with superoxide to form peroxynitrite, which has been shown to activate PKC and open KATP channels, although the involvement of peroxynitrate was excluded in the current study. Finally, NO appears to have a direct inhibitory effect on the mPTP.
Dedkova and Blatter (9) also have detected a delaying effect of NO on mPTP opening. The relationship that this has with regulation of mitochondrial Ca2+ by inhibition of the uniporter is not clear. mPTP opening has been proposed to be an “escape valve” releasing excess mitochondrial Ca2+, although this rather contradicts the concept of regulation of mitochondrial Ca2+ by Ca2+ cycling envisioned by Crompton (7). Perhaps instead, as Dedkova and Blatter discuss in their present article (9), mPTP opening serves as a rapid mechanism of mitochondrial depolarization, thereby reducing the driving force for Ca2+ uptake. This mechanism would presumably be invoked when a more rapid reduction in ΔΨm was required than that which could be achieved by inhibition of mitochondrial dehydrogenases. Thus inhibition of mPTP opening by NO would be secondary to a reduction in mitochondrial [Ca2+]. Depletion of Ca2+ has previously been shown to protect hepatocytes against anoxia by inhibiting mPTP opening (23).
Yet, while NO can depolarize mitochondria and delay mPTP opening, other investigators have found that it can sustain mitochondrial repolarization after reperfusion and can prevent Ca2+-induced mPTP opening in isolated mitochondria (19). The diversity of effects observed is reminiscent of the concentration-dependent effects of NO on Ca2+ uptake. Indeed, while low concentrations of NO inhibit mPTP opening, concentrations >2 μM can accelerate mPTP opening in isolated mitochondria (6). It may be relevant that more than one “mode” of mPTP opening has been reported: a transient, “flickering” mode, as well as a more stable opening (7, 17). The different assays used to assess mPTP opening cannot always distinguish these modes. For example, application of cyclosporin A (or other mPTP inhibitor) would be expected to inhibit both mechanisms, and it is not known how well the modes can be distinguished by examining redistribution of fluorescent calcein, as used in the current study.
An aspect of NOS regulation that was not examined in the present study by Dedkova and Blatter is its phosphorylation. Although its pattern of phosphorylation is complicated, Ser1177 phosphorylation by Akt has been identified as being critical in its regulation, and this regulatory mechanism appears to be independent of Ca2+ levels (10, 13). Phosphorylation of NOS appears to be fundamental to cardioprotection because various cardioprotective agents that act via Akt cease to function in eNOS-knockout mice (3, 26). Interestingly, some of the residual cardioprotection in eNOS−/− mice treated with diazoxide was eliminated by inhibition of iNOS (26), so the various NOS enzymes may have redundancy in function. It will be important to establish the relationship between NOS as a target of cardioprotective kinase pathways and mitochondrial calcium levels.
eNOS (and nNOS) were already known to be Ca2+ dependent, and with evidence accumulating (including the present article) for the effect of NO on Ca2+ regulation, NO seems to be well placed as a potential component of a feedback mechanism regulating Ca2+ levels in the cell, cytosol, and mitochondria, if not in other organelles as well. In view of these findings, NO seems certain to retain its star status for some time to come.
- Copyright © 2005 the American Physiological Society