Characterization of the membrane proteins of rat liver lysosomes. Composition, enzyme activities and turnover.

Lysosomes prepared from the livers of untreated rats and from the livers of rats injected with either Triton WR-1339 or dextran yielded membranes that were similar in both polypeptide composition and activities of ATPase and acid 5'-nucleotidase. The administration of Triton WR-1339 (and dextran) resulted in an increase in ATPase activity of liver homogenates that was associated with a parallel increase in the ATPase activity of the lysosomal membrane. On the other hand, plasma membranes appear to be different from lysosomal membranes with respect to polypeptide composition and enzyme activities. The ATPase activity of lysosomal membranes is not affected by ouabain and suramin, inhibitors of the plasma-membrane ATPase. The plasma-membrane alkaline 5'-nucleotidase has little activity at acid pH. Pulse-labelling of lysosomal membranes with [3H]fucose and with [3H]- and [14C]-leucine occurred rapidly, faster than labelling of plasma membranes. The labelling kinetics indicate that lysosomal membranes may be assembled independently of plasma membranes. These data suggest that, in liver, little bulk transport of plasma membrane to lysosomes takes place, and lysosomal-membrane proteins may not be derived from those of plasma membranes.     

Increased uptake of fatty acids by the isolated rat liver after raising the fatty acid binding protein concentration with clofibrate

Following the administration of clofibrate to rats, the concentration of Z protein or fatty acid binding protein in liver cytosol increases by 98 %. Ligandin concentration remains unchanged. Isolated perfused livers of clofibrate-treated rats take up free fatty acids from the perfusate at a significantly higher rate (+ 76 %) than controls. Lipid synthesis from radioactive fatty acids is not modified by clofibrate administration. The yield of plasma membranes obtained from liver homogenates as well as their lipid composition are similar in control and clofibrate treated livers. These results seem to exclude the possibility that the enhancement of FFA uptake could result from an indirect effect of the drug on FFA metabolism and/or plasma membrane surface and thus support the view that Z protein plays a role in intracellular fatty acid transport in the liver.     

Synthesis of CDP-diacylglycerol by rat liver rough microsomes as visualized by electron microscopic autoradiography: relationship to GTP-stimulated membrane fusion

Using detergent-free conditions of incubation for the analysis of liponucleotide synthesis, we compared GTP-dependent formation of CDP-diacylglycerol (CDP-DG) and membrane fusion in RNA-depleted rough microsomes from rat liver. After incubation of stripped rough microsomes (SRM) in the presence of GTP and [5-3H]-CTP, radioactivity was recovered in lipid extracts and identified by thin-layer chromatography as a single spot which co-migrated with CDP-DG. The nucleotide requirement for CDP-DG synthesis and that for membrane fusion were observed to be identical. We next carried out an electron microscopic autoradiographic analysis on incubated membranes to determine the site of incorporation of [5-3H]-CTP. Silver grains were observed directly over the unilamellar membranes of natural vesicles. In confirmation of the biochemical data, quantitation of silver grain density indicated more grains over membranes incubated in the presence of GTP than over those incubated in the absence of this nucleotide. For membranes incubated in the presence of GTP, the grain density was similar over fused and unfused membranes in the same preparation. When SRM were incubated with the enzyme co-factors required for synthesis of phosphatidylinositol, a GTP-independent membrane fusion was observed by both transmission and freeze-fracture electron microscopy. Together with the biochemical and autoradiographic data, this suggests that phospholipid metabolism may be activated by GTP and lead to the fusion of RER membrane.     

Membrane Tethering and Nucleotide-dependent Conformational Changes Drive Mitochondrial Genome Maintenance (Mgm1) Protein-mediated Membrane Fusion

Cellular membrane remodeling events such as mitochondrial dynamics, vesicle budding, and cell division rely on the large GTPases of the dynamin superfamily. Dynamins have long been characterized as fission molecules; however, how they mediate membrane fusion is largely unknown. Here we have characterized by cryo-electron microscopy and in vitro liposome fusion assays how the mitochondrial dynamin Mgm1 may mediate membrane fusion. Using cryo-EM, we first demonstrate that the Mgm1 complex is able to tether opposing membranes to a gap of ∼15 nm, the size of mitochondrial cristae folds. We further show that the Mgm1 oligomer undergoes a dramatic GTP-dependent conformational change suggesting that s-Mgm1 interactions could overcome repelling forces at fusion sites and that ultrastructural changes could promote the fusion of opposing membranes. Together our findings provide mechanistic details of the two known in vivo functions of Mgm1, membrane fusion and cristae maintenance, and more generally shed light onto how dynamins may function as fusion proteins.     

Changes in the activities of the enzymes of hepatic fatty acid oxidation during development of the rat.

1. Changes in the activities of several enzymes involved in mitochondrial fatty acid oxidation were measured in livers of developing rats between late foetal life and maturity. The enzymes studied are medium- and long-chain ATP-dependent acyl-CoA synthetases of the outer mitochondrial membrane and matrix, GTP-dependent acyl-CoA synthetase, carnitine acyltransferase, enoyl-CoA hydratase, 3-hydroxyacyl-CoA dehydrogenase, general 3-oxoacyl-CoA thiolase and acetoacetyl-CoA thiolase.     

Diethylnitrosamine-induced increase in gamma-glutamyltranspeptidase in rat liver: its association with thyroid hormone deficiency and its reversal by tri-iodothyronine.

1. The nodular phase of hepatic premalignancy was induced in male Fischer 344 rats by the administration of diethylnitrosamine, 200 mg/kg i.p., followed by promotion utilizing the Solt-Farber promoting regime. 2. Relative to the situation in normal non-treated control rats: the activity of gamma-glutamyltranspeptidase was found to be increased 9.42-fold in homogenate and 7.33-fold in plasma membrane fractions prepared from the livers of saline-injected control rats; and 81.37-fold in homogenates and 91.92-fold in plasma membranes prepared from the livers of diethylnitrosamine-injected rats; plasma levels of total T3 and total T4 were found to be decreased 42.06 and 47.45% in saline-injected control rats and 88.7 and 83.2% in diethylnitrosamine-injected rats, respectively. 3. An early pre-nodular phase of hepatic premalignancy was produced in young immature and mature adult male Fischer 344 rats by the administration of diethylnitrosamine, 75 mg/kg, without subsequent application of the promotion regime. 4. Relative to the situation in control rats: the activity of gamma-glutamyltranspeptidase was found to be increased in liver homogenates prepared from diethylnitrosamine-treated rats, 1.62-fold in young immature rats 1.20-fold in mature adult rats; plasma levels of total T3 were found to be reduced in diethylnitrosamine-treated rats, 28% in young immature rats 9% in mature adult rats. 5. Treatment of diethylnitrosamine-injected young immature male Fischer 344 rats at the prenodular phase of hepatic premalignancy with tri-iodothyronine at 0.005 micrograms/kg s.c. daily for 7 days reversed the diethylnitrosamine-induced increase in liver homogenate gamma-glutamyltranspeptidase activity and the decrease in plasma total T3, restoring these parameters to normal levels.(ABSTRACT TRUNCATED AT 250 WORDS)     

Association of a photoreceptor-specific tetraspanin protein,ROM-1, with triton X-100-resistant membrane rafts from rod outer segment disk membranes

This study reports the isolation and characterization of a Triton X-100-resistant membrane fraction from homogenates of rod outer segment (ROS) disk membranes purified free of the surrounding plasma membrane. A portion of the ROS disk membrane was found to be resistant to Triton X-100 extraction at 4 degrees C. This detergent-resistant fraction was isolated as a low buoyant density band on sucrose density gradients and exhibited an increase in light scattering detected at 600 nm. Biochemical analysis of the Triton X-100-resistant fraction showed it to be enriched in cholesterol and sphingomyelin relative to phospholipid and in phospholipid relative to protein compared with the soluble fraction. The Triton X-100-resistant membranes described herein did not arise simply from partial solubilization of the ROS disk membranes because detergent-treated low buoyant density fractions isolated from homogenates with octyl glucopyranoside had cholesterol and sphingomyelin content indistinguishable from that of solubilized ROS disk homogenates. Analysis of proteins associated with the Triton X-100-resistant fraction showed it to be enriched in the rim-specific protein ROM-1 and caveolin; surprisingly, the fusion protein peripherin/rds (where rds is retinal degeneration slow), also localized to the disk rim, was entirely absent from the membrane raft domain. The lipid profiles of the Triton X-100-resistant membranes were virtually identical in preparations homogenized in either the light or dark. Slightly more ROM-1 was recovered from samples prepared in the light (23%) than from samples prepared in the dark (13%), but peripherin/rds could not be detected in either preparation. When the Triton X-100-resistant membranes were treated with methyl-beta-cyclodextran to deplete membrane cholesterol, the resultant membranes contained slightly lower levels of ROM-1, specifically in the dimeric form. Cholesterol depletion also resulted in the collapse of the large caveolin complex to monomeric caveolae. The results presented herein characterize a pool of ROM-1, a photoreceptor tetraspanin protein, that may play a regulatory role in peripherin/rds-dependent fusion.     

The mechanisms of up-regulation of the hepatic insulin receptor in hypoinsulinemic diabetes mellitus

The mechanism underlying the increased insulin binding found in hepatic plasma membranes from streptozotocin-diabetic rats was evaluated by measuring insulin binding to intact and Triton X-100-soluble extracts of plasma membranes prepared from the livers of control rats and rats administered streptozotocin (85 mg/kg). In addition, to assess whether the cellular content of hepatic insulin receptors is also increased in diabetic animals, we measured insulin-binding activity in intact and soluble extracts of total hepatic cellular membrane preparations (100,000 X g cellular pellets). The data indicate that while insulin binding is increased (52 +/- 3%) in intact hepatic plasma membranes from diabetic rats compared to control rats, there is no comparable increase in insulin binding in intact total cellular membranes or in Triton X-100-soluble extracts of plasma membranes or total cellular membranes. We therefore conclude that the enhanced insulin binding found in the livers of diabetic rats is the result of a local redistribution of plasma membrane insulin receptors from cryptic to exposed sites. Finally, the data suggest the presence of a negative modulator of insulin-binding affinity in intact plasma and total cellular membranes.     

The cesium-induced delay in myoblast membrane fusion is accompanied by changes in cellular subfraction lipid composition.

We have recently demonstrated that the delay in myoblast membrane fusion induced by cesium is accompanied by changes in isolated membrane lipids (Santini, M.T., Indovina, P.L., Cantafora, A. and Blotta, I. (1990) Biochim. Biophys. Acta 1023, 298-304). In the present study, we have investigated changes in the lipid profile of total cell homogenates and microsomal membrane fractions during myoblast membrane fusion as well as the effects that addition of cesium may have on these lipid variations in order to try to understand the production and translocation of lipids during this myogenic process. The data presented here indicate that the lipid composition of cell homogenates and microsomes varies in a different manner from isolated plasma membranes during myogenic fusion. In addition, cesium affects, in a different manner, the normally-occurring lipid production and distribution which takes place in each subcellular fraction.     

GTP hydrolysis is required for vesicle fusion during nuclear envelope assembly in vitro

Nuclear envelope assembly was studied in vitro using extracts from Xenopus eggs. Nuclear-specific vesicles bound to demembranated sperm chromatin but did not fuse in the absence of cytosol. Addition of cytosol stimulated vesicle fusion, pore complex assembly, and eventual nuclear envelope growth. Vesicle binding and fusion were assayed by light and electron microscopy. Addition of ATP and GTP to bound vesicles caused limited vesicle fusion, but enclosure of the chromatin was not observed. This result suggested that nondialyzable soluble components were required for nuclear vesicle fusion. GTP gamma S and guanylyl imidodiphosphate significantly inhibited vesicle fusion but had no effect on vesicle binding to chromatin. Preincubation of membranes with 1 mM GTP gamma S or GTP did not impair vesicle binding or fusion when assayed with fresh cytosol. However, preincubation of membranes with GTP gamma S plus cytosol caused irreversible inhibition of fusion. The soluble factor mediating the inhibition by GTP gamma S, which we named GTP-dependent soluble factor (GSF), was titratable and was depleted from cytosol by incubation with excess membranes plus GTP gamma S, suggesting a stoichiometric interaction between GSF and a membrane component in the presence of GTP gamma S. In preliminary experiments, cytosol depleted of GSF remained active for fusion of chromatin-bound vesicles, suggesting that GSF may not be required for the fusion reaction itself. We propose that GTP hydrolysis is required at a step before the fusion of nuclear vesicles.     

Cytoplasmic membrane transplants in early Xenopus embryos indicate cell-cycle-specific effects.

Fragments of rough endoplasmic reticulum or Golgi complex isolated from normal adult rat liver homogenates were injected into one cell of cleaving two-cell Xenopus laevis embryos and the effects on development were monitored during early cleavage by morphological analysis. Scanning electron microscopy revealed the formation of large cells on the injected side of the embryos. Such large cells were not present in controls and thus were considered to have been formed as a consequence of delayed cleavage. Delay of cleavage was obtained with as little as 1 ng of membrane protein giving a ratio of membrane protein to embryo protein of 1:10(5). Cytological observations of microinjected embryos confirmed the occurrence of delayed cytokinesis and suggested that nuclear division became asynchronous. Since rough microsomes from proliferating tissues (i.e., livers with primary tumors and livers undergoing regeneration) showed little or no effect on cytokinesis after microinjection into early embryos, we conclude that cytoplasmic membranes may exhibit cell-cycle-specific properties important for normal development.     

Accumulation of polyunsaturated free fatty acids coincident with the fusion of rough endoplasmic reticulum membranes.

The accumulation of polyunsaturated free fatty acids (PUFAs) was observed coincident with GTP-dependent fusion of liver rough microsomes. Whereas 0.5 mM NADPH led to a parallel reduction (greater than 50%) in membrane fusion and PUFA accumulation, indomethacin (50 microM) either had little effect or slightly augmented both processes. CTP was observed to stimulate accumulation of PUFAs and diacylglycerol (DAG). Therefore PUFAs may be relevant for GTP-dependent membrane fusion and together with DAG may play a role in fusion stimulated in the presence of CTP.     

5-Methyltryptamine stimulates phospholipase C-mediated breakdown of exogenous phosphoinositides by blowfly salivary gland membranes

5-Methyltryptamine, through a GTP-dependent mechanism, stimulated breakdown of endogenous [3H]inositol-labeled phosphoinositides in membranes prepared from blowfly salivary gland homogenates through a phospholipase C exhibiting a pH optimum of approximately 7.0. Unlabeled membranes, prepared from salivary gland homogenates, hydrolyzed exogenous [3H]phosphatidylinositol 4,5-bisphosphate substrate with generation of labeled inositol phosphates. Inositol trisphosphate formation was increased approximately 200% by 10 microM guanosine 5'-(O-thio)-trisphosphate (GTP gamma S) within 30 s. 5-Methyltryptamine, in the presence of 10 microM GTP gamma S, increased the rate of inositol trisphosphate formation by approximately 500% within 30 s. Half-maximal activation of hormone-stimulated breakdown of exogenous substrate required approximately 0.05 microM GTP gamma S. [3H]Phosphatidylinositol was also hydrolyzed during incubation with membranes, resulting in the generation of inositol, glycerol phosphoinositol, and inositol monophosphate. Formation of inositol monophosphate was stimulated approximately 30% by 10 microM GTP gamma S and 10 microM 5-methyltryptamine. Neither inositol nor glycerol phosphoinositol formation was affected by hormone. These results indicate that in a cell-free system from blowfly salivary glands, 5-methyltryptamine, through a GTP-dependent mechanism, directly activates a phospholipase C which mediates phosphoinositide hydrolysis.     

Quantitation of epinephrine- and glucagon-sensitive adenylate cyclases of rat liver implications of alterations of enzymatic activities during preparation of particulate fractions and membranes

Stimulated and basal adenylate cyclase activities from livers of young and old rats were lower in particulates than in homogenates. Particulates were compared to homogenates by reconstituting the suspensions to the volume of the homogenates from which they were derived; enzyme activities in paired homogenates and particulates therefore reflected the same amounts of membrane-bound enzyme. The magnitude of the losses of hormone-sensitive activities in particulates was dependent on the age and sex of the animals and the concentrations of hormone. Particulates from 3-month-old animals showed glucagon-( (1 · 10^?5 M) and epinephrine-sensitive (1 · 10^?4 M) activities which were 67 and 78% of homogenate activities, respectively; particulates from 24-month-old animals had activities relative to homogenates of 55% for glucagon and as low as 32% for epinephrine. The glucagon dose vs. response curve in particulates and membranes showed maximal activity at 1 · 10^?7 M glucagon while in homogenates activity increased linearly with increasing glucagon concentrations up to 1 · 10^?5 M. Losses of basal and anion-stimulated activities were similar at both ages. Fluoride and azide stimulations relative to basal activities were greater in particulates than in homogenates, while relative epinephrine activity was lower in particulates, suggesting qualitative alteration of adenylate cyclase during preparation of particulates. These studies show that adenylate cyclase activity in rat liver is presently best quantitated in homogenates and suggest caution in comparisons of enzyme activities based on particulates or membranes prepared from animals of differing physiologic states.     

Membrane recruitment of effector proteins by Arf and Rab GTPases

In their GTP-bound form, Arf and Rab family GTPases associate with distinct organelle membranes, to which they recruit specific sets of effector proteins that regulate vesicular transport. The Arf GTPases are involved in the formation of coated carrier vesicles by recruiting coat proteins. On the other hand, the Rab GTPases are involved in the tethering, docking and fusion of transport vesicles with target organelles, acting in concert with the tethering and fusion machineries. Recent structural studies of the Arf1-GGA and Rab5-Rabaptin-5 complexes, as well as other effector structures in complex with the Arf and Rab GTPases, have shed light on the mechanisms underlying the GTP-dependent membrane recruitment of these effector proteins.     

Possible Involvement of Pertussis Toxin Substrates (Gi, Go) in Desipramine-Induced Refractoriness of Adenylate Cyclase in Cerebral Cortices of Rats

To evaluate the efficiency of coupling between beta-receptor and adenylate cyclase catalyst via a GTP-binding protein, Gs, in the brain membrane two parameters were employed: a beta-agonist-induced increase in the membrane GTP-dependent adenylate cyclase activity and a beta-agonist-induced shortening of the lag time preceding the onset of the steady-state activation by guanyl-5'-yl-beta-gamma-imidodiphosphate [Gpp(NH)p] of the membrane cyclase. Both parameters showed lower values in membranes from desipramine-treated rats compared with untreated rats. Thus, coupling of beta-adrenergic receptors to adenylate cyclase in the brain membrane was impaired by the desipramine treatment. Rats once injected intraventricularly with islet-activating protein (IAP), pertussis toxin, were subjected to desipramine treatment, for the purpose of studying effects of another kind of the GTP-binding protein (Gi), which loses its function as a signal transducer on being ADP-ribosylated selectively by the toxin. IAP treatment did not impair the beta-receptor coupling by itself, since neither of the above two parameters for the coupling were reduced by IAP treatment. Moreover, the first parameter was normalized, though the second one was not, by superimposition of the IAP treatment upon the desipramine-treated rats. It seems likely, therefore, that Gi interacts with a Gs-adenylate cyclase coupling in an inhibitory fashion in brain membranes. The desensitization might be overcome when the inhibitory interaction of Gi on the subsequent process is attenuated by IAP treatment.     

Freeze-fracture analysis of the effects of intermediates of the phosphatidylinositol cycle on fusion of rough endoplasmic reticulum membranes.

While searching for the identity of the effector of the putative GTP-binding protein involved in fusion of rough endoplasmic reticulum (RER) cell-free incubation conditions were found permitting fusion in a GTP-independent manner. Membrane fusion was obtained using medium required to study synthesis of phosphatidylinositol (PI). We now report on the effects of various co-factors and intermediates of the PI cycle on the interaction of rough microsomes. By freeze-fracture, fusion of rough microsomes was defined as the appearance of fracture-planes of membrane larger than those of unincubated membrane. Cytosine triphosphate (CTP, 3 mM) in the presence of 2 mM MnCl2 was most effective in stimulating fusion. Guanosine triphosphate (GTP) at the same concentration, could substitute for CTP to stimulate fusion, ATP, ITP, UTP and guanosine 5'-[gamma-thio]triphosphate (GTP gamma S) could not. When combined together in the same medium CTP potentiated the effect of GTP. Arachidonic acid (20 micrograms/ml) also stimulated fusion in the presence of MnCl2. This led to the appearance of large fracture-planes of membrane with a heterogeneous distribution of intramembranous particles. Other saturated fatty acids at the same concentration did not stimulate fusion. Phosphatidylinositol (PI, 50 micrograms) and 2 mM MnCl2 had a similar effect as arachidonic acid and MnCl2 in stimulating fusion. The PI effect was largely augmented in the presence of CTP. Our results are consistent with the concept that metabolism of phospholipids may modulate GTP-dependent fusion of RER membranes.     

Age-related changes in the translocation of phosphatidate phosphohydrolase from the cytosol to microsomal membranes in rat liver

The effects of oleate, spermine and chlorpromazine were assayed in the presence or absence of 0.15 M KCl on the translocation of phosphatidate phosphohydrolase activity from cytosol to endoplasmic reticulum membranes in liver homogenates obtained from rats aged 1, 30, 60, 180 and 360 days. Marked age-associated decreases in phosphatidate phosphohydrolase distribution onto the membranes were demonstrated under nearly all conditions. In liver homogenates taken from 1-day-old rats and incubated with 0.15 M KCl, most of the enzyme was active (associated with the membranes). Physiological salt concentration (0.15 M KCl) produced a 2-fold increase of oleate-induced translocation of phosphatidate phosphohydrolase activity in liver homogenates from 1-day-old rats; it had no effect on those from 60-day-old rats, and produced a notable decline in liver homogenates obtained from 180- and 360-day-old rats. The promoting effect of spermine on oleate-induced translocation of this enzyme activity was higher in younger rats when incubated in the absence of 0.15 M KCl. Chlorpromazine did not show its usual antagonizing effect on oleate-induced translocation of phosphatidate phosphohydrolase when added to homogenates taken from 1-day-old rats. The antagonizing effect was slightly apparent in liver homogenates from 30-day-old rats and was more pronounced in those from 60-day-old rats in which the values diminished to one-half and to one-third either in the presence or absence of 0.15 M KCl.     

Hydrolysis of retinyl palmitate by enzymes of rat pancreas and liver. Differentiation of bile salt-dependent and bile salt-independent, neutral retinyl ester hydrolases in rat liver

Previous studies have demonstrated that homogenates of the livers of rats contain a neutral retinyl ester hydrolase activity that requires millimolar concentrations of bile salts for maximal in vitro activity. The enzymatic properties of this neutral, bile salt-dependent retinyl ester hydrolase activity in liver homogenates are nearly identical to those observed in the present report for the in vitro hydrolysis of retinyl palmitate by purified rat pancreatic cholesteryl ester hydrolase (EC 3.1.1.13). Moreover, anti-rat pancreatic cholesteryl ester hydrolase IgG completely inhibits the bile salt-dependent retinyl ester hydrolase activity of rat liver homogenates whereas normal rabbit IgG does not. We also show that liver homogenates contain a neutral, bile salt-independent retinyl ester hydrolase activity that differs from the bile salt-dependent activity in that 1) its absolute activity does not vary markedly among individual rats, 2) it is not inhibited by antibodies to pancreatic cholesteryl ester hydrolase, and 3) it is localized in the microsomal fraction of liver homogenates. Subfractionation of microsomes demonstrates that the neutral, bile salt-independent retinyl ester hydrolase activity is associated with liver cell plasma membranes and thus may play a role in the hydrolysis of retinyl esters delivered to the liver by chylomicron remnants.     

Altered Mg2+ transport across liver plasma membrane from streptozotocin-treated rats

Type-I diabetes is associated with a decrease in magnesium content in various tissues, including liver. We have reported that hepatocytes from streptozotocin-injected rats have lost the ability to accumulate Mg2+ following hormonal stimulation. To assess whether the defect is inherent to the Mg2+ transport mechanism located in the hepatocyte cell membrane, plasma membrane vesicles were purified from diabetic livers. Diabetic plasma membranes do not retain intravesicular Mg2+ as tightly as vesicles purified from livers of age-matched non-diabetic rats. In addition, the amount of intravesicular Mg2+ these vesicles exchange for extravesicular Na+ or Ca2+ is 2-3-fold larger than in non-diabetic vesicles. The partition of Ca2+/Mg2+ and Na+/Mg2+ exchange mechanisms in the apical and basolateral domains of liver plasma membrane is maintained under diabetic conditions, although the Na+/Mg2+ exchanger in diabetic basolateral membranes has lost the ability to operate in reverse and favor an accumulation of extravesicular Mg2+ within the vesicles in exchange for entrapped Na+. These data indicate the occurrence of a major alteration in Mg2+ transport across the hepatocyte membrane, which can explain, at least in part, the decrease in liver magnesium content observed in diabetic animals and patients.