PubMed Health⌕ Search

Biomedical subjects

R A Colvin

Publications and source records attributed to R A Colvin.

At least 37 records · Page 2Linked to original sources
PubMed1992

Unusual structure of the human immunodeficiency virus type 1 trans-activation response element.

The trans-activation response element (TAR) of human immunodeficiency virus type 1 is a structured RNA consisting of the first 60 nucleotides of all human immunodeficiency virus type 1 RNAs. Computer analyses and limited structural analyses indicated that TAR consists of a stem-bulge-loop structure. Mutational analyses showed that sequences in the bulge are required for Tat binding, whereas sequences in both the bulge and the loop are required for trans activation. In this study, we probed the structures of TAR and various mutants of TAR with chemical probes and RNases and used these methods to footprint a Tat peptide on TAR. Our data show that the structure of wild-type TAR is different from previously published models. The bulge, a Tat-binding site, consists of four nucleotides. The loop is structured, rather than simply single stranded, in a fashion reminiscent of the structures of the tetraloop 5'-UUCG-3' and the GNRA loop (C. Cheong, G. Varani, and I. Tinoco, Jr., Nature [London] 346:680-682, 1990; H.A. Heus and A. Pardi, Science 253:191-193, 1991). RNA footprint data indicate that three bases in the bulge are protected and suggest that a conformational change occurs upon Tat binding.

Base Sequence↗
PubMed1991

Na+/Ca2+ exchange activity is increased in Alzheimer's disease brain tissues.

These studies were performed to determine the changes that occur in Na+/Ca2+ exchange activity in Alzheimer's disease (AD) brain tissues. Cerebral plasma membrane vesicles were purified by sucrose density gradient centrifugation from frozen postmortem hippocampal/temporal cortex tissue slices derived from age matched brains of normal, AD and non-Alzheimer dementia (NAD) origin (autopsy confirmed). Membrane marker assays (Na/K ATPase, muscarinic receptor, cytochrome c oxidase) revealed no change in membrane purity across different preparations. Thin-section electron microscopy revealed predominantly intact unilamellar vesicles. Vesicles were preincubated for 15 min (37 degrees C) in buffer containing 132 mM NaCl, 5 mM KCl, 1.3 mM MgCl2, 10 mM glucose and 10 mM HEPES (pH 7.4). Ca2+ uptake was initiated by diluting vesicles 20-fold with buffer containing either 132 mM NaCl or 132 mM choline chloride and 45CaCl2 then terminated by addition of 200 microM LaCl3 and rapid filtration. Ca2+ content increased rapidly at first and then maintained a steady plateau for up to 5 min. When the Ca2+ ionophore A23187 (10 microM) with 100 microM EGTA was added after 4 min, Ca2+ content was reduced to 10% of its original value. Ruthenium red (10 microM) had no effect on Ca2+ content. Na(+)-dependent Ca2+ uptake (Ca2+ content measured in choline chloride minus that measured in NaCl) was increased in AD brains as evidenced by both an increase in the initial rise in Ca2+ content and in elevated values of peak plateau Ca2+ content.(ABSTRACT TRUNCATED AT 250 WORDS)

Adult↗
PubMed1991

Studies of the thermal inactivation of cardiac adenylyl cyclase: evidence for a conformational change in the reaction mechanism.

Membrane bound cardiac adenylyl cyclase was shown to undergo a spontaneous and irreversible thermal inactivation with a t1/2 of approximately 10 min. The loss of activity could not be explained by the action of endogenous proteases. Repeated freeze-thaw of membrane preparations resulted in a much increased rate of thermal inactivation (t1/2 = approx. 2 min). ATP, adenylimidodiphosphate, ADP, and PPi protected the enzyme from thermal inactivation with dissociation constants (Kd) of 193, 5.04, 84.4, and 6.3 microM, respectively. 5'-AMP and cyclic AMP were ineffective as protectors at concentrations as high as 3 mM. Activators of adenylyl cyclase such as Mn2+, forskolin, 5-guanylylimidodiphosphate, and NaF and 9 mM Mg2+ protected against thermal inactivation with Kd of 16.8 microM, 8.81 microM, 0.23 microM and 1.04 mM, respectively. Mg2+ alone was without effect. Thermal inactivation was first order under all conditions tested. Arrhenius plots of the rate constants for inactivation vs temperature were linear. The increased stability of ligand bound adenylyl cyclase was shown to be associated with an increased free energy of activation (delta G 0). These data provide evidence for the existence of two distinct conformations of cardiac adenylyl cyclase based on different susceptibilities to thermal inactivation. These enzyme conformations, termed E1 and E2, may be important reaction intermediates. The thermal stability of E1 was highly influenced by the enzyme's membrane lipid environment. The formation of E2 from E1 was enhanced by interaction with substrate, PPi, activators of adenylyl cyclase, and by interaction with dissociated stimulatory guanine nucleotide binding protein-alpha beta gamma heterotrimers.

Adenosine Triphosphate↗
PubMed1991

Calcium inhibition of cardiac adenylyl cyclase. Evidence for two distinct sites of inhibition.

Increasing the free calcium concentration from 10(-8) M to 10(-4) M inhibited cardiac sarcolemmal adenylyl cyclase activated by the addition of 5 X 10(-4) M forskolin or 1 X 10(-4) M GTP or Gpp(NH)p. The calcium inhibition curve in the presence of all three activators was shallow and best fit by a two site model of high affinity (less than 1.0 microM) and low affinity (greater than 0.1 mM). Gpp(NH)p appeared to decrease the sensitivity of adenylyl cyclase to inhibition by calcium at the high affinity site. Similar inhibition constants were obtained with each of the activators. Calmodulin content of native freeze-thaw vesicles was 76.2 +/- 14.2 ng/mg. Treatment of the vesicles with 1 mM EGTA to remove calmodulin significantly reduced calmodulin content to 19.7 +/- 1.35 ng/mg. This treatment had no significant effect on the calcium inhibition profile. Increasing free calcium to 3 X 10(-6) M was shown to have no effect on the EC50 estimated for either Gpp(NH)p or forskolin but did slightly increase the EC50 estimated for Mg2+ in the presence of maximal concentrations of either activator. Nevertheless, maximally stimulating concentrations of Mg2+ were unable to overcome calcium inhibition. Pretreatment of sarcolemmal membranes with pertussis toxin was shown to have no significant effect on calcium inhibition of adenylyl cyclase. The results suggest that the overall inhibitory action of calcium was most likely calmodulin independent and involved a direct interaction with the catalytic subunit at two distinct sites of high and low affinity. At the low affinity site calcium most likely competes with Mg2+ for an allosteric divalent cation binding site.(ABSTRACT TRUNCATED AT 250 WORDS)

Adenylate Cyclase Toxin↗
PubMed1990

Amiodarone, verapamil, and quinidine do not affect equilibrium binding of digoxin.

The binding of [3H]digoxin to purified canine cardiac sarcolemmal vesicles was characterized. Scatchard analysis of saturation isotherms yielded linear plots with a maximal binding capacity of 174 +/- 31.9 pmol/mg, a dissociation constant of 31.7 +/- 4.59 nM, and a Hill coefficient of 0.947 +/- 0.02 (mean +/- SEM), suggesting that [3H]digoxin bound to a single class of sites. In contrast to their marked effect on steady-state serum digoxin levels when administered in combination, quinidine, verapamil, and amiodarone were without effect on equilibrium binding of [3H]digoxin. Thus, increased steady-state serum concentrations of digoxin resulting from combination therapy with these particular drugs probably will have cardiac effects that may increase the risk of digitoxicity to the patient.

Amiodarone↗
PubMed1989

Interaction of amiodarone and desethylamiodarone with the cardiac muscarinic receptor in vitro.

We studied the interaction of amiodarone hydrochloride (Cordarone) and its major metabolite desethylamiodarone with the muscarinic receptor in purified canine cardiac sarcolemmal vesicles by measuring equilibrium binding of the muscarinic antagonist [3H]quinuclidinyl benzilate (QNB) and carbachol displacement of [3H]-QNB. At a [3H]-QNB concentration of 0.02 nM, equilibrium binding was inhibited by amiodarone and desethylamiodarone with an IC50 of 6.86 x 10(-6) M and 2.25 x 10(-6) M, respectively. The presence of increasing concentrations of [3H]-QNB in the incubation medium was able to reverse the inhibition seen with 1 x 10(-6) M amiodarone. Scatchard analysis of [3H]-QNB saturation isotherms (37 degrees C, pH 7.4) in the presence of 1 x 10(-6) M amiodarone showed an apparent increase in equilibrium dissociation constant (Kd) over control from 0.045 +/- 0.002 nM to 0.084 +/- 0.001 nM while maximal binding capacity (Bmax) was unaffected: 10.8 +/- 1.14 and 10.5 +/- 1.48 pmol/mg (means +/- S.E.M., n = 3), respectively. The inhibitory effect of amiodarone on equilibrium binding was highly dependent on the drug:membrane phospholipid mole ratio with effects beginning at a ratio of less than 0.1:1. Hill plot analysis was consistent with the interaction of [3H]-QNB at a single site in the presence or absence of amiodarone. Amiodarone (3 x 10(-6) M) decreased the pseudo-first order forward rate constant of [3H]-QNB (0.02 nM) with the muscarinic receptor (kobs = 4.05 +/- 0.61 x 10(-4)/s under control conditions and 2.36 +/- 0.15 x 10(-4)/s in the presence of amiodarone).(ABSTRACT TRUNCATED AT 250 WORDS)

Amiodarone↗
PubMed1989

Inhibition of 3H-quinuclidinyl benzylate binding to cardiac muscarinic receptor by long chain fatty acids can be attenuated by ligand occupation of the receptor.

Phospholipase A2 (PLA2) inhibits ligand binding to sarcolemmal muscarinic receptors in heart. To determine whether this effect of PLA2 is mediated by membrane accumulation of non-esterified fatty acids (FFA), the effect of selected fatty acids on the binding of 3H-quinuclidinyl benzylate (3H-QNB) to purified canine sarcolemmal membranes before and after PLA2 treatment was examined. Equilibrium 3H-QNB binding was inhibited by 5 min exposure of membrane vesicles to oleic, linoleic or arachidonic acid (IC50 = 6.3 +/- 0.9, 9.9 +/- 1.1, and 6.8 +/- 0.4 microM, respectively); the saturated fatty acids, stearic and palmitic acid (10 microM) had no effect. Scatchard analysis of equilibrium binding isotherms showed that the effect of the unsaturated fatty acids to inhibit 3H-QNB binding reflected a decrease of Bmax and a reduction of the affinity of the remaining receptors. The effect of unsaturated fatty acids was dependent on the mole ratio of fatty acid to membrane phospholipid present (FFA/PL ratio). Washing of fatty acid-treated membranes with bovine serum albumin (BSA) resulted in partial recovery of both maximal binding (Bmax) and affinity. The fatty acid-induced reduction of Bmax was also attenuated if binding was started by simultaneous addition of 3H-QNB and FFA. Similarity of the FFA induced effects on 3H-QNB binding to sarcolemmal muscarinic receptors to those induced by PLA2 suggest that membrane accumulation of unsaturated fatty acids underlies in part the effect of PLA2. Furthermore, modification of the receptor-ligand interaction by changes in the membrane lipid composition may be prevented by ligand occupation of the receptor.

Animals↗
PubMed1989

The Drosophila zeste locus is nonessential.

Diepoxybutane-induced mutations of the Drosophila zeste locus were generated in an effort to obtain a null allele. Of 33 mutations of this X-linked gene isolated, 16 were associated with multilocus deletions of zeste and adjacent complementation groups, while the remainder were defects restricted to zeste undetectable by Southern blot analysis. Two of these multilocus deletions (Df(1)zdeb3 and Df(1)zdeb92) were employed in the synthesis of females completely deleted for zeste. Such "zesteless" flies were produced, though at frequencies lower than Mendelian expectations. zeste-deleted females are fertile, and can give rise to zeste-deleted female progeny. These results demonstrate that the product of the zeste gene is not essential to viability or to female fertility, even if absent both as a maternal contribution and as a product of the zygotic genome. However, the possibility that zeste may influence relative viability cannot be excluded. In spite of previous in vitro indications that the zeste protein may activate transcription of the Ultrabithorax (Ubx) gene, zeste -deleted flies are Ubx+ in phenotype. This suggests that the zeste protein normally is either a very weak transcription factor, or that its function can be substituted by that of other regulatory proteins.

Animals↗
PubMed1987

Verapamil interaction with the muscarinic receptor: stereoselectivity at two sites.

Verapamil, in addition to blocking calcium channels, exhibits such "non-specific" effects on myocardium as inhibition of sodium and potassium conductances and modifications of muscarinic receptor-ligand interactions. To characterize further the effects of verapamil on the cardiac muscarinic receptor, we examined the abilities of the enantiomers of verapamil to modify the binding of the muscarinic antagonist [3H]quinuclidinyl benzilate ([3H]QNB) to purified canine sarcolemmal vesicles. Membranes were incubated with [3H]QNB and various concentrations of racemic, (+)-, or (-)- verapamil (25 or 37 degrees, pH 7.4), and reactions were terminated by rapid filtration. (-)-Verapamil (Ki of 5.3 +/- 0.2 microM) was twice as potent an inhibitor of equilibrium binding as (+)-verapamil (Ki of 11.4 +/- 0.6 microM), and this effect resulted from the ability of each enantiomer to slow [3H]QNB-receptor association. This degree of stereoselectivity, albeit at nanomolar concentrations, was similar to that observed for each enantiomer to compete for the specific phenylalkylamine site in this preparation. Verapamil also inhibited [3H]QNB-receptor dissociation, but this effect required high concentrations and demonstrated stereoselectivity opposite to that observed for association. These findings support the view that verapamil interacts with two distinct sites, possibly within membrane lipid, each with a different affinity and preference for (+)- and (-)-verapamil, to modify the muscarinic receptor.

Amines↗
PubMed1987

Effect of exogenous phospholipase A2 treatment on cardiac muscarinic receptors of highly purified canine sarcolemmal vesicles.

Effects of phospholipase A2 (PLA2)-catalyzed hydrolysis of sarcolemmal phospholipids on ventricular muscarinic receptors were examined by measuring specific binding of 3H-quinuclidinyl benzilate (3H-QNB) to purified canine sarcolemmal vesicles. Scatchard analysis of 3H-QNB saturation isotherms (25 degrees C, pH 7.4) yielded a dissociation constant (Kd) of 58 +/- 10 pM and maximal binding capacity (Bmax) of 5.7 +/- 1.3 pmol/mg. Pretreatment of the sarcolemmal membranes with PLA2 (1 U/ml) for 5 and 30 mins reduced Bmax to 38% and 7% of control, and increased Kd to 109 +/- 21 and 129 +/- 12 pM, respectively. Washing of PLA2-treated sarcolemmal vesicles with defatted albumin resulted in a partial recovery of Bmax, presumably by removing hydrolysis products. PLA2 also reduced equilibrium binding of 3H-QNB to 43% of control when reactions were started by simultaneous addition of 3H-QNB and 1 U/ml PLA2; however, under these conditions the inhibitory effect of PLA2 could be overcome by increasing 3H-QNB from 30 to 600 pM. PLA2 added at equilibrium (59 mins after reaction start) had no effect on 3H-QNB binding. Lipid hydrolysis by PLA2 was unaffected by the presence of bound 3H-QNB. The ability of ligand occupation and removal of hydrolysis products to attenuate the effects of PLA2-treatment on muscarinic receptor sites may be explained if modification of the membrane lipid bilayer leads to transitions between different states of the receptor.

Animals↗
PubMed1987

Deinhibition of cardiac Na+-K+-ATPase after exposure to exogenous phospholipase A2.

After 2 h of exogenous phospholipase A2 (PLA2) exposure, membrane phospholipid decreased from 3.22 +/- 0.31 to 1.06 +/- 0.13 mumol/mg (33% of control). All classes of phospholipid, except sphingomyelin, were hydrolyzed, whereas total cholesterol content was unaffected. Increases in nonesterified fatty acids (NEFA) were reflected primarily in oleic (18:1), linoleic (18:2), and arachidonic (20:4). Na+-K+-adenosinetriphosphatase (ATPase) activity was inhibited to 29% of control by 2 h of PLA2 treatment, and this inhibition was reversed (albeit, not completely after 5 min of PLA2 treatment) by removal of the hydrolysis products with 0.1% bovine serum albumin (BSA). In contrast, the apparent binding capacity for [3H]ouabain was not affected by PLA2 treatment. Unmasking of latent [3H]ouabain binding by alamethicin was utilized to estimate changes in the proportion of sealed vesicles present before and after PLA2 treatment. PLA2 treatment resulted in a time-dependent loss of sealed vesicles that paralleled the time course of phospholipid hydrolysis and was not reversed by washing with BSA. These studies demonstrate that cardiac Na+-K+-ATPase activity is inhibited by accumulation of endogenously produced lysophospholipids and NEFA. In contrast, loss of vesicle integrity may result from both accumulation of endogenously produced hydrolysis products and membrane phospholipid depletion.

Animals↗
PubMed1985

Structure-function studies of canine cardiac sarcolemmal membranes. I. Estimation of receptor site densities.

A novel method for the estimation of receptor site densities in purified canine cardiac sarcolemmal vesicles is described. Canine sarcolemmal vesicles, purified by the method of Jones et al. (Jones, L.R., Maddock, S.W. and Besch, H.R. (1980) J. Biol. Chem. 255, 9971-9980) had high (Na+ + K+)-ATPase specific activity (127 +/- 1.9 mumol Pi/mg per h). Total phospholipid content, estimated by measurements of total phosphorus and total fatty acid contents, was 3.09 mumol/mg. Saturation isotherms for several receptor ligands gave the following values for Kd and Bmax: ouabain 32.6 +/- 2.7 nM, 365 +/- 59 pmol/mg; quinuclidinyl benzilate 0.055 +/- 0.010 nM, 5.8 +/- 0.7 pmol/mg; dihydroalprenolol 4.6 +/- 1.0 nM, 2.2 +/- 0.2 pmol/mg; and nitrendipine 0.21 +/- 0.04 nM, 0.93 +/- 1.04 pmol/mg. Membrane phospholipid surface area per ligand-binding sites was estimated from the Bmax values for each receptor ligand utilizing 3.09 mumol phospholipid/mg and 60 A2 as the average surface area occupied by each phospholipid molecule. The following receptor site densities per micrometer 2 phospholipid surface were obtained: ouabain, 400; quinuclidinyl benzilate, 6; dihydroalprenolol, 2; and nitrendipine, 1. As the surface area contributed by protein was estimated to be less than 20% of the lipid surface area, these values must be reduced by approx. 20% to estimate site densities per micrometer 2 membrane surface. These data demonstrate much lower beta-adrenergic and muscarinic receptor density compared to that of Na+ pump sites.

Animals↗
PubMed1985

Structure-function studies of canine cardiac sarcolemmal membranes. II. Structural organization of the sarcolemmal membrane as determined by electron microscopy and lamellar X-ray diffraction.

The morphological and ultrastructural properties of highly purified canine cardiac sarcolemmal vesicles, prepared by a modification (Colvin, R.A., Ashavaid, T.F. and Herbette, L.G. (1985) Biochim. Biophys. Acta 812, 601-608) of the method of Jones et al. (Jones L.R., Madlock, S.W. and Besch, H.R. (1980) J. Biol. Chem. 255, 9971-9980), were examined by several techniques. Thin-section electron microscopy showed predominantly intact unilamellar vesicles with little staining beyond the lipid bilayer boundaries. Freeze-fracture electron microscopy demonstrated that the majority of particles are approx. 90 A diameter and present at a density of 780 +/- 190 micrometers-2 (+/- S.D.). If it is assumed that some of these particles represent the (Na+ + K+)-ATPase, the finding that they are largely confined to the convex fracture face suggests a predominant right-side-out orientation of these sarcolemmal vesicles that is consistent with biochemical assays. The sarcolemmal membrane width measured by electron microscopy (unhydrated membrane width of 50-70 A) is consistent with the unit cell dimensions of 56-77 A determined by lamellar X-ray diffraction (hydrated membrane width). A unit cell dimension of 56-62 A was also found by X-ray diffraction for sarcolemmal lipids extracted from these preparations, indicating that the isolated sarcolemmal preparations do not contain a significant surface coat (glycocalyx). As both cardiac and skeletal sarcoplasmic reticulum membranes have a 80-100 A membrane width, these findings demonstrate that the purified sarcolemmal membrane is structurally distinct from both cardiac and skeletal sarcoplasmic reticulum. In contrast to the protein-rich skeletal sarcoplasmic reticulum membrane, which contains a single essential protein responsible for the regulation of cytosolic Ca2+ concentration, the sarcolemma is a lipid-rich membrane that contains a variety of proteins associated with many regulatory functions served by this membrane in cardiac muscle.

Animals↗
PubMed1985

Effects of fatty acids on Na/Ca exchange in cardiac sarcolemmal membranes.

Three structurally distinct amphiphiles palmitic acid, oleic acid, and palmityl carnitine were studied to determine their effects on sodium dependent calcium uptake by purified cardiac sarcolemmal vesicles (PSL). Sodium dependent calcium uptake by PSL when studied over a 20 min reaction period was composed of an initial rapid uptake (20.9 +/- 0.93 nmol/mg X 30 s, mean +/- S.E. n = 20) a plateau in calcium content (42.4 +/- 3.2 nmol/mg, mean +/- S.E. n = 20) and a slow spontaneous release characterized by a first order rate constant of 0.68 +/- 0.08/h (mean +/- S.E. n = 18). Both palmityl carnitine and palmitic acid inhibited, whereas oleic acid stimulated initial calcium uptake. All three amphiphiles shortened the time to peak calcium content, inhibited peak calcium content and increased the rate constant for calcium release. All these effects were observed at fatty acid: membrane phospholipid mole ratios of 0.67 : 1 to 1.67 : 1 for oleic acid and palmityl carnitine and 0.02 : 1 to 0.42 : 1 for palmitic acid. These effects do not reflect disruption of membrane vesicle structure and may be explained, at least in part, by amphiphile induced increases in sarcolemmal membrane ion permeability. Although amphiphile accumulation has been implicated in the pathogenesis of cellular abnormalities in the ischemic myocardium, this study has shown that large amounts of amphiphile relative to membrane lipid are required to alter sarcolemmal membrane function in vitro.

Animals↗
PubMed1983

Binding of the calcium channel blocker nitrendipine to its receptor in purified sarcolemma from canine cardiac ventricle.

High affinity binding of the 1,4-dihydropyridine calcium channel blocker [3H]nitrendipine was found in cardiac sarcolemma but not cardiac sarcoplasmic reticulum or mitochondria. Sarcolemmal binding of [3H]nitrendipine was saturable and reversible, with a maximum (Bmax) of approximately 1 pmol/mg protein and a Kd of approximately 0.14 nM. Displacement of sarcolemma-bound [3H]nitrendipine by other nifedipine analogs was stereospecific. The Kd for nitrendipine binding was approximately three orders of magnitude lower than the IC50 for the negative inotropic effect of this drug on isolated cat myocardium.

Animals↗
PubMed1982

Effects of divalent cations, trypsin, and phospholipases on the passive permeability to sodium of inside-out vesicles from human red cells.

Inside-out vesicles (IOV) were prepared from human red blood cells. Steady-state uptake of 23Na was observed to generally follow an exponential time course with a rate constant of 1.57 +/- 0.09 h-1 (SE). One week of cold storage (0-4 degrees C) increased the rate constant to 2.50 +/- 0.12 h-1 (SE). Mg2+, Ca2+, or Sr2+ decreased the rate of 22Na uptake with no observable differences between the three divalent cations when tested at concentrations of 50 microM. Mg2+ was shown to decrease the rate of 22Na uptake at concentrations as low as 5 microM with maximal effect at 50 to 100 microM. The decrease in rate of 22Na uptake induced by Mg2+ could be enhanced by exposure of IOV to Mg2+ for longer periods of time. Trypsin treatment of OIV increased the rate of uptake of 22Na and was dependent on the concentration of trypsin added between 5 to 25 micrograms/ml (treated for 5 min at 25 degrees C). The ability of Mg2+ (50 microM) to decrease the rate of 22Na uptake was still observed after maximal trypsin treatment. Phospholipase A2 or phospholipase C treatment of IOV increased the rate of 22Na uptake and was dependent on the amount of phospholipase A2 (0.1 to 1.0 units/ml) or phospholipase C (0.25 to 2.5 units/ml) added (treated for 5 min at 25 degrees C). After phospholipase A2 treatment, the observed decrease in the rate of 22Na uptake induced by Mg2+ (50 microM) was generally greater than controls. After phospholipase C treatment, the observed decrease in rate of 22Na uptake induced by Mg2+ (50 microM) was less or absent when compared with controls. Phospholipase C treatment was less effective in preventing the Mg2+ effect the longer IOV were exposed to Mg2+. The results suggest that Mg2+ binds to phospholipid headgroups to reduce Na permeability perhaps by inducing a change in bilayer structure or phospholipid association.

Cations, Divalent↗