Mg deficiency induces various metabolic disturbances including glucose metabolism in the liver. However, no comprehensive information is currently available on the metabolic pathways affected by Mg deficiency. The present study examined metabolite content in the liver of Mg-deficient rats using a metabolomic analysis. In this study, 4-week-old, male Sprague–Dawley rats were fed a control diet or a Mg-deficient diet for 8 weeks. The metabolomic analysis identified 105 metabolites in the liver, and significant differences were observed in the hepatic contents for thirty-three metabolites between the two groups. An analysis by MetaboAnalyst, a web-based metabolome data analysis tool, indicated that the Mg deficiency affected taurine/hypotaurine metabolism, methionine metabolism and glycine/serine/threonine metabolism; taurine, hypotaurine, glycine, serine and threonine contents were increased by Mg deficiency, whereas the amounts of 2-ketobutyric acid (a metabolite produced by the catabolism of cystathionine or threonine) and 5'-methylthioadenosine (a metabolite involved in spermidine synthesis) were decreased. The amount of glucose 6-phosphate, a hub metabolite of glycolysis/gluconeogenesis and the pentose phosphate pathway, was significantly decreased in Mg-deficient rats. Mg deficiency also decreased metabolite contents from the citric acid cycle, including citric acid, fumaric acid and malic acid. Aberrant metabolism may be related to the allosteric regulation of enzymes; the mRNA levels of enzymes were generally similar between the two groups. The present study suggests that the Mg deficiency-mediated modulation of hepatic metabolism is as yet uncharacterised.

Mg deficiency increases the concentration of Zn in the liver. We investigated the effect of Mg deficiency on the expression of Zn-regulating factors such as Zn transporters and metallothionein (MT) in the rat liver. Because Ca deficiency alleviates some of the effects of Mg deficiency, we also investigated the interactions associated with Ca and Mg deficiencies. Growing male rats were given a control diet, a Mg-deficient diet, a Ca-deficient diet and a Mg- and Ca-deficient diet for 3 weeks. Mg and Ca deficiencies additively increased the mRNA levels of MT-1 and MT-2, the MT protein concentration and the concentration of Zn in the liver. The hepatic mRNA level of Zip14 increased with Mg deficiency but not with Ca deficiency. The dietary treatments did not affect the mRNA levels of other Zn transporters such as Zip1, Zip5, ZnT1, ZnT5 and ZnT6 in the liver. Ca deficiency was found to decrease the amount of femoral Zn and increase serum Zn concentration. This did not occur in the case of Mg deficiency. These results suggest that Mg deficiency enhances hepatic Zn uptake by the up-regulation of Zip14 expression and increases hepatic Zn concentration, leading to the enhancement of MT expression. Ca deficiency causes a transfer of Zn from the bone to the liver, which increases hepatic Zn concentration and, in turn, up-regulates the expression of MT. Because Mg and Ca deficiencies increase hepatic Zn concentration and increase MT expression by different mechanisms, their effects are additive.

Mg deficiency accelerates Fe accumulation in the liver, which may induce various metabolic disturbances. In the present study, we examined the gene expression of Hepcidin, a peptide hormone produced in the liver to regulate intestinal Fe absorption negatively, in Mg-deficient rats. Although liver Fe concentration was significantly higher in rats fed an Mg-deficient diet for 4 weeks than in rats fed a control diet, Hepcidin expression in the liver was comparable between the dietary groups. Previous studies revealed that Fe overload up-regulated Hepcidin expression through transcriptional activation by Fe-induced bone morphogenetic protein (Bmp) 6, a growth/differentiation factor belonging to the transforming growth factor-β family, in the liver. Mg deficiency up-regulated the expression of Bmp6 but did not affect the expression of inhibition of DNA binding 1, a sensitive Bmp-responsive gene. In addition, the expression of Bmp receptors such as activin receptor-like kinase 2 (Alk2), activin receptor type IIA (Actr2a), activin receptor type IIB (Actr2b) and Bmp type II receptor (Bmpr2) was lower in the liver of Mg-deficient rats than in that of control rats. The present study indicates that accumulation of hepatic Fe by Mg deficiency is a stimulant inducing Bmp6 expression but not Hepcidin expression by blunting Bmp signalling possibly resulting from down-regulation of the receptor expression. Unresponsive Hepcidin expression may have a role in Mg deficiency-induced changes related to increased liver Fe.

The importance of Mg for the immune function is well recognized; however, there is no information available about the effect of Mg intake on the modulation of local immune response in the intestine. Thus, in the present study the hypothesis that short periods of Mg deprivation can affect intestinal mucosa and local immune response was tested. For this purpose, OF1 female mice were fed a semipurified diet (1000 mg Mg/kg diet). For 3 d before immunization and 1 d after, half of the animals were fed a Mg-deficient diet (30 mg Mg/kg diet), three immunizations per os were performed every 3 weeks with Escherichia coli producing the CS31A capsule-like protein (1010or 2×109 bacteria per animal). Mice were killed 10 d after the last immunization. The level of specific anti CS31A immunoglobulin (Ig) G and IgA in the serum and secretory IgA in the intestinal secretions and faeces were measured by ELISA. The results indicated that administration of a high dose of immunogen with a low-Mg diet led to lower specific IgA levels in the intestinal mucus and serum. Administration of a low dose of immunogen with a low-Mg diet led to lower IgA and IgG levels in the serum and secretory IgA coproantibodies. To assess alterations of intestinal mucosa caused by a low-Mg diet for a short period, histological and scanning electron microscopy analyses were performed on samples from mice (not submitted to the vaccination protocol) after 3 d on the Mg-deficient diet. These analyses showed several alterations, suggesting perturbations in the growth of the intestinal mucosa. These changes were accompanied by modifications in the expression of several genes involved in cell growth and stress response. From this present work, it may be concluded that short periods of Mg deprivation can affect the intestinal mucosa and local immune response of the intestine.

The serum mineral levels, glucose disappearance rate (kg), total area under the glucose (ΔG) and insulin (ΔI) curves, and static insulin secretion were compared among rats fed a Mg-deficient diet for 6 (DF-6) or 11 (DF-11) weeks, and rats fed a control diet for the same periods (CO-6 and CO-11 groups). No change in glucose homeostasis was observed among DF-6, CO-6 and CO-11 rats. DF-11 rats showed an elevated kg and a reduced ΔG and ΔI. For evaluating the effect of supplementation, rats fed a control or Mg-deficient diet for 6 weeks were then fed a Mg- supplemented diet for 5 weeks (SCO and SDF groups respectively). The serum Mg levels in SDF rats were similar to those in CO-11 and SCO rats, but higher than in the DF-11 group. SDF rats showed similar kg, ΔG and ΔI compared with the CO-11 and SCO groups. However, a significantly lower kg and higher ΔG and ΔI were observed in SDF compared with DF-11 rats. Basal and 8.3 mmol glucose/l-stimulated insulin secretion by islets from DF-11 rats were higher than by islets from CO-11 rats. These results indicate that moderate Mg depletion for a long period may increase the secretion and sensitivity to insulin, while Mg supplementation in formerly Mg-deficient rats may prevent the increase in sensitivity and secretion of insulin.

Experimental Mg deficiency leads to alterations in the immune response. Reduction of thymus weight and histological changes were previously observed in Mg-deficient rats after several weeks on a deficient diet, suggesting that functions of this immune organ may be affected by Mg deficiency. More recently, changes in the immune system during early Mg deficiency were shown. Thus, in the present study we examined modifications in the thymus during the early stages of Mg deficiency in weanling rats. From our results, it appears that Mg deficiency accelerates thymus involution. The assessment of apoptosis (enumeration of apoptotic cells on the basis of morphological criteria and intranucleosomal degradation of genomic DNA) showed greater values in thymuses from Mg-deficient rats as compared with controls. This was observed very early, since a significant difference was shown on the second day of deficiency, before reduced weight of thymus, which was recorded in the later period. These results indicate the relationship of accelerated thymus involution with an active process of cell death. Mg deficiency led to histological changes in the thymus. In the early stage of deficiency (second day) the presence of inflammatory cells was shown, suggesting that the inflammatory process was already occurring in the tissue studied. Later (eighth day) an increased proportion of epithelial reticular cells in the cortex was shown, indicating a remodelling process occurring in this period. Enhanced susceptibility to peroxidation also occurred very early during Mg deficiency. It may be hypothesized that disturbances in Mg status of short duration could have cellular effects with various deleterious consequences.

Nutritional deficiencies of magnesium or zinc lead to a progressive and often marked growth retardation. We have evaluated the effect of Mg and Zn deficiency on growth, serum insulin-like growth factor-1 (s-IGF-1), growth hormone (s-GH) and insulin (s-insulin) in young rats. In 3-week-old rats maintained on Mg-deficient fodder for 12 d the weight gain was reduced by about 34%, compared with pair-fed controls. This was accompanied by a 44% reduction in s-IGF-1, while s-insulin showed no decrease. After 3 weeks on Mg-deficient fodder, growth had ceased while serum Mg (s-Mg) and s-IGF-1 were reduced by 76 and 60% respectively. Following repletion with Mg, s-Mg was completely normalized in 1 week, and s-IGF-1 reached control level after 2 weeks. Growth rate increased, but the rats had failed to catch up fully in weight after 3.5 weeks. Absolute and relative pair-feeding were compared during a Mg repletion experiment. Absolute pair-fed animals were given the same absolute amount of fodder as the Mg-deficient rats had consumed the day before. Relative pair-fed animals were given the same amount of fodder, on a body-weight basis, consumed in the Mg-deficient group the day before. In a repletion experiment the two methods did not differ significantly from each other with respect to body-weight, muscle weight, tibia length and s-IGF-1, although there was a tendency towards higher levels in the relative pair-fed group. The peak in s-GH after growth hormone-releasing factor 40 (GRF 40) was 336 (se 63) μg/l in 5-week-old rats that had been Mg depleted for 14 d, whereas age-matched control animals showed a peak of 363 (se 54) μg/l (not significant).

In 3-week-old rats maintained on Zn-deficient fodder for 14 d weight gain was reduced by 83% compared with pair-fed controls. Serum Zn (s-Zn) and s-IGF-1 were reduced by 80 and 69% respectively, while s-insulin was reduced by 66%. The Zn-deficient animals showed a more pronounced growth inhibition than that seen during Mg deficiency and after 17 d on Zn-deficient fodder s-IGF-1 was reduced by 83%. Following repletion with Zn, s-Zn was normalized and s-IGF-1 had increased by 194% (P <0.05) after 3 d. s-IGF-1, however, was not normalized until after 2.5 weeks of repletion. Growth rate increased but the catch-up in weight was not complete during 6 weeks. The maximum increase in s-GH after GRF 40 was 774 (se 61) μg/l in control animals ν. 657 (se 90) μg/l in 6-week-old rats that had been Zn-depleted for 12 d (not significant). In conclusion, both Mg and Zn deficiency lead to growth inhibition that is accompanied by reduced circulating s-IGF-1, but unchanged s-GH response. Zn deficiency, but not Mg deficiency, caused a reduction in s-insulin. The reduction in s-IGF-1 could not be attributed to reduced energy intake, but seems to be a specific effect of nutritional deficiency of Mg or Zn. It is suggested that the growth retardation seen during these deficiency states may be mediated through reduced s-IGF-1 production.

The effects of magnesium or zinc deficiency on growth, tissue contents of Mg or Zn and protein synthesis have been compared in 4–13-week-old rats. When maintained on Mg-deficient fodder (1.6 mmol/kg) or Zn-deficient fodder (27 μmol/kg) rats showed a reduced weight gain, whereas repletion caused increased growth rates. Pair-feeding experiments showed that this could not be attributed to reduced energy intake only. In rats maintained on Mg-deficient fodder for 14 d [3H]leucine incorporation into skeletal muscle and the heart was reduced by 24–38% compared with pair-fed controls (P <0.001–0.002). The incorporation of [3H]phenylalanine was reduced by 19–31%. Tissue Mg contents, however, were only reduced by 6–7% (not significant). The pair-fed rats showed no reduction in the [3H]leucine incorporation compared with ad lib. fed animals. In rats maintained on Zn-deficient fodder for 15 d [3H]leucine incorporation into skeletal and heart muscle was reduced by 57–64% compared with pair-fed controls. The pair-fed rats showed no reduction in the [3H]leucine incorporation compared with ad lib. fed animals. In the Zn-deficient animals the content of Zn was not reduced in the skeletal muscles, whereas there was a small (15%) but significant loss of Zn in the heart. In another experiment, Zn depletion for 17 d caused a reduction in [3H]leucine incorporation of 35–41 %. After 5 d of Zn repletion this defect was restored, and the [3H]leucine incorporation was above control level in the skeletal muscles. It is concluded that the intact organism is very sensitive to dietary Mg or Zn deficiency, and that the reduced growth and protein synthesis cannot easily be attributed to the reduction of tissue Mg or Zn content per se. This points to the existence of other control mechanisms mediating down-regulation of growth and protein synthesis in response to reduced dietary supplies and the ensuing drop in the plasma concentrations of Mg and Zn.