Aimin Zhang¹, Bella T. Altura^2,3 and Burton M. Altura^1,2,3

Departments of ¹Physiology and ²Medicine, and ³The Center for Cardiovascular and Muscle Research, State University of New York, Health Science Center at Brooklyn, NY 11203

Received 5/15/97; Accepted 5/21/97; On-line 6/1/97

5. DISCUSSION

Based on digitized video-image analysis techniques, utilizing a fluorescent probe (mag-fura-2) and a superfusion system, we were able to: 1. measure both the mean values of [Mg²⁺]_i and their spatial distributions within single endothelial cells, and 2. continually define [Mg²⁺]_i responses to alterations of [Mg²⁺]_o in the same cells. The present study shows that the mean basal level of [Mg²⁺]_i of human aortic endothelial cells was in the sub-millimolar range. This value is in keeping with the levels of [Mg²⁺]_i observed in other cells and tissues studied by the same indicator (16,19,20,22-26). Overall, most recent studies, including those using ³¹P-nuclear magnetic resonance (NMR) spectroscopy on heart and brain, suggest that [Mg²⁺]_i in most tissue types is on the order of submillimolar (14,15,19,27-30).

It is evident from our new findings that [Mg²⁺]_i in human aortic endothelial cells is rapidly responsive to [Mg²⁺]_o, unlike that observed for several other cell types (17,18). Increasing [Mg²⁺]_o to 4.8 mM Mg²⁺ caused an increase of [Mg²⁺]_i within 2-10 min, from a basal value of 0.51 mM to a mean value of 0.80mM, i.e., about a 1.6-fold increase (Table 1). Recently, possible interference of (Ca²⁺)_i with measurements of [Mg²⁺]_i, using mag-fura-2, has been suggested (31). However, an overestimation of [Mg²⁺]_i caused by interference with excess (Ca²⁺)_i seems unlikely in the present stusdies, because: 1. the basal value of (Ca²⁺)_i in these human aortic endothelial cells was found to be 74±22 nM and is not affected by elevation of [Mg²⁺]_o (4); 2) the dissociation constant of the mag-fura-2­Ca²⁺ complex is about 65 µM; and 3) the fluorescence intensity of the latter is not altered by changes in Ca²⁺ until it exceeds 5 µM (22).

Little is known about the cellular regulation of [Mg²⁺]_i homeostasis in endothelial cells. Although we do not show here the kinetic changes of [Mg²⁺]_i in response to elevation of [Mg²⁺]_o, our data clearly indicate that rises in [Mg²⁺]_i in human aortic endothelial cells occur within a short time (2-10 min), much faster than previously recognized (17,18,32). Somewhat similar findings have been noted recently in several other types of cells and tissues (16,20,28-30,33-35). In contrast to a slow process of transport for many other cell types (31), the plasma membranes of vascular endothelial cells may be highly permeable to [Mg²⁺]_o. The rise in [Mg²⁺]_i could be a direct result of increasing Mg²⁺ influx via changes of the transmembrane Mg²⁺ gradient. However, at both 1.2 and 4.8 mM [Mg²⁺]_o, the levels of [Mg²⁺]_i were regulated at levels well below the equilibrium potential given by the Nernst equation. This suggests that Mg²⁺ may, at least partially, be actively-transported out. Several carrier-mediated magnesium extrusion systems have been proposed, such as Mg²⁺/Ca²⁺ exchange (36), or Na⁺/Mg²⁺ exchange (37). In addition, intracellular Mg²⁺ buffering may also be involved. For instance, alteration of [Mg²⁺]_o could influence proton (H⁺) and Ca²⁺ transport as well as the status of cellular bioenergetic metabolism (28,30,35,38). All these intracellular events could modify Mg²⁺ binding at intracellular sites. Undoubtedly, these direct or indirect mechanisms, which remain to be unmasked, may play a role in the regulation of Mg²⁺ homeostasis.

Similar to that observed for subcellular (Ca²⁺)_i distributions, [Mg²⁺]_i also appears to be segregated in various eukaryotic cell types, including liver, skeletal and vascular muscle (15,16,28,39-41). However, to our knowledge, the present studies are the first to show a heterogeneous distribution of the concentration of [Mg²⁺]_i in endothelial cells; the latter being seen, irrespective of [Mg²⁺]_o. The brightness spots, observed herein, most likely represent Mg²⁺ release from extensive Mg²⁺ binding and uptake elements of internal storage sites which limit Mg²⁺ diffusion. However, the role of such heterogeneous distribution of [Mg²⁺]_i in the regulation of cellular functions is not clear, and further studies will be needed to define, precisely, the intracellular compartments for [Mg²⁺]_i in human endothelial cells, particularly with respect to Mg²⁺ permeability and transport characteristics.

Recently, agonist-induced changes of [Mg²⁺]_i have been reported in the intact brain (29,42), vascular smooth muscle cells (43-45), fibroblasts (46), renal epithelial cells (26), and pancreatic acinar and beta-cells (25,34). The elevations in [Mg²⁺]_i by increases of [Mg²⁺]_o in endothelial cells, observed herein, could be of significant physiological relevance, because [Mg²⁺]_i at this submillimolar range is known to fit the Michaelis-Menton constant (K_m) values for Mg²⁺ activation or inhibition of many enzymes (32), regulatory functions of G-proteins (17,18,47) and cation channel activity (5,6).

We propose that modulation of endothelial cell functions by [Mg²⁺]_o may be mediated by numerous Mg²⁺-sensitive regulatory systems through changes in plasma ionized Mg²⁺. Through the use of a new ion selective electrode for [Mg²⁺]_o, it has been shown that despite no alteration in total plasma Mg, the ionized [Mg²⁺]_o significantly varies in a number of pathophysiological states in humans, including ischemic heart disease, angina, hypertension, diabetes mellitus, migraine headache, renal disorders, and stroke (14,15,19,48).