An official website of the United States government
NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.
Vink R, Nechifor M, editors. Magnesium in the Central Nervous System [Internet]. Adelaide (AU): University of Adelaide Press; 2011.
Mounir N. Ghabriel and Robert Vink.
Author Information and AffiliationsThe finding that magnesium levels are reduced in acute and chronic brain diseases has led to a recent surge in interest in the role of magnesium in the normal and injured nervous system, although the mechanisms of magnesium decline in pathological conditions and its availability in the neural tissue after administration are not fully understood. The brain has two main barrier systems: (1) the blood-brain barrier (BBB) formed by brain capillary endothelial cells which separate the blood from the extracellular fluid in the neuropil; and (2) the blood-CSF barrier (BCSFB) formed by choroidal epithelial cells which separate the blood from the CSF. Genetic studies in families with hereditary hypomagnesemia have identified mutations in two genes encoding claudin-16 (paracellin-1) and claudin-19, both localized at tight junctions between nephron epithelial cells and providing passive paracellular conductance for magnesium in the kidney. Endothelial cells of the BBB also express claudins, although whether members of the claudin family expressed at the BBB and BCSFB have similar conductance for magnesium akin to the role of claudin-16 and -19 in the nephron remains to be confirmed. Recently, the transient receptor potential melastatin (TRPM) members TRPM6 and TRPM7 have been identified as cation channels for magnesium transport. Although it is not known if choroidal epithelial cells express TRPM6 and TRPM7, these molecules are expressed by brain endothelial cells and may play a role in magnesium transport. While it is evident that magnesium enters the CNS through the BBB and is actively transported by choroidal epithelial cells into the CSF, the mechanisms of its entry into the brain will require further investigation.
Magnesium is a critical cation and an essential nutrient for normal body functions; hence mechanisms exist in the body for its homeostasis through a highly integrated feedback system involving intestinal absorption, renal excretion, bone metabolism and the parathyroid gland (Dai et al., 2001). Magnesium is involved in a myriad of biochemical processes including acting as a cofactor in the activation of many intracellular enzymes (Aikawa, 1976; Ebel and Gunther, 1980; Gunther, 2008) and is important for protein synthesis (Terasaki and Rubin 1985) and cell membrane stabilization (Bara and Guiet-Bara 1984). Magnesium deficiency reduces protein synthesis, serum antibody activity and the immune response (McCoy and Kenney, 1975), and induces CNS epileptiform activity (Morris, 1992) and hypomagnesemic tetany (Fontenot et al., 1989). There is an inverse correlation between dietary magnesium and the level of C-reactive protein (a marker of inflammation) and the level of E-selectin (a marker of endothelial cell dysfunction) (Song et al., 2007). Recent technical developments have allowed better assessment of the levels of magnesium in the body, although the blood, which is used in routine testing, contains only 0.3% of the total body magnesium (Elin, 1987). Approximately one half of the magnesium contained in the body is stored in the bone, while the rest exists in the soft tissues, mainly in the intracellular compartment, with less that 1% being present in the extracellular compartment (Elin, 1988). Indeed, magnesium is the second most common cation in the intracellular compartment after potassium. More than one half of the plasma content of magnesium is bound to plasma proteins, and the remaining free ionized magnesium (Mg2+), the metabolically active fraction, is held within a narrow range (0.53- 0.67mM) in normal healthy controls (Altura and Altura, 1991). The recommended daily allowance ranges from 320 to 420mg/day for women and men, respectively (Bergman et al., 2009).
There has been a recent surge in interest in the dynamics and role of magnesium in the normal and injured brain due to accumulating evidence of a reduction in the level of total and free Mg2+ in the brain in acute and chronic neurological diseases (Vink et al., 1987; 2009). In intensive care patients with traumatic brain injury (TBI) reduction of serum ionized Mg2+ correlates with the severity of TBI as determined by the Glasgow Coma Scale Score (Kahraman et al., 2003). In experimental TBI in the rat there is a sustained decline in intracellular Mg2+, detected by phosphorus magnetic resonance spectroscopy, that correlates with motor deficit (Cernak et al., 2004; Heath and Vink, 1996), while magnesium administration significantly improves motor outcome (Heath and Vink 1999; Turner et al., 2004). Magnesium administration reduces TBI- induced brain edema (Feldman et al., 1996) and restores blood-brain barrier (BBB) effectiveness to Evans blue tracer when compared to injured non-treated rats (Esen et al., 2003; Imer et al., 2009). Magnesium also reduces brain edema after cold-induced brain injury (Turkoglu et al., 2008). However, the mechanisms of magnesium decline in pathological conditions and its availability in the neural tissue after administration are still unclear (Vink et al., 2009).
Therapeutically, magnesium can be administered orally, intravenously or intramuscularly (Elin, 1988). It is most commonly used used as magnesium chloride, sulphate, gluconate, acetate (Fine et al., 1991) or lactate (Fine et al., 1991; Simoes Fernandes et al., 1985). Magnesium therapy has been utilized in numerous experimental and clinical settings including migraine (Peikert et al., 1996), asthma (Cheuk et al., 2005), depression (Eby and Eby, 2006), anxiety (Kara et al., 2002), diabetes (Resnick et al., 1993; Wester and Dyckner, 1987), hypertension (Wester and Dyckner, 1987), atrial fibrillation (Fanning et al., 1991), sleep disorders, insomnia and chronic fatigue (Takahashi et al., 1992), dementia (Glick, 1990), osteoporosis (Sojka and Weaver, 1995), fibromyalgia (Porter et al., 2010), pain (Soave et al., 2009), eclampsia (Euser and Cipolla, 2009), constipation (Guerrera et al., 2009), cerebral palsy (Rouse, 2009), lacunar stroke (Muir et al., 2004), TBI (McIntosh et al., 1988) and aneurysmal subarachnoid hemorrhage (van den Bergh, 2009; Yahia et al., 2005).
The current review will briefly outline the status of magnesium in the body, then will summarize current knowledge of its distribution in the CNS and discuss its transport across barrier membranes in the brain compared to its transport in other organs such as the kidney.
Much of our knowledge about magnesium absorption and excretion was gained in the latter part of the twentieth century and particularly in the nineties. Studies of normal subjects showed that magnesium bioavailability from high magnesium containing food sources such as almonds is equal to that obtained from soluble magnesium acetate, but enteric coating of magnesium chloride impairs magnesium bioavailability (Fine et al., 1991). Magnesium absorption occurs mainly in the small intestine (Brannan et al., 1976; Graham et al., 1960; Schroeder et al., 1969). The kinetics and rate of magnesium absorption are not dependant on calcium intake (Brannan et al., 1976). Under basal conditions the small intestine absorbs 30–50% of the magnesium intake, although this percentage diminishes with senescence, chronic renal disease and increasing intake (Musso 2009). The fractional magnesium absorption appears to fall progressively so that absorption as a function of intake is curvilinear (Fine et al., 1991). The fecal magnesium appears to be primarily derived from material that is not absorbed by the body rather than magnesium secreted by the intestine (Aikawa, 1976). Approximately 80% of the absorbed magnesium passes in the glomerular filtrate (Dai et al., 2001; Quamme and de Rouffignac, 2000).
More that 95% of the magnesium filtered by the glomerulus is reclaimed mainly in the thick ascending loop of Henle (60-70%) and to a lesser extent (10-15%) in the proximal convoluted tubules (Brunette et al., 1974; Dai et al., 2001; Quamme 1997; Quamme and de Rouffignac, 2000). A further 10% of magnesium in the filtrate is claimed in the distal convoluted tubule (de Rouffignac and Quamme, 1994) and this segment contributes to magnesium conservation (Quamme 1997). Magnesium reabsorption within the thick ascending loop appears to be passive and occurs via the paracellular route (de Rouffignac and Quamme, 1994), being driven by the trans- epithelial voltage (Quamme, 1997; Quamme and de Rouffignac, 2000), where positive luminal charge favors movement of magnesium from the luminal to the abluminal side of nephron epithelium (Quamme, 1997). On the other hand magnesium transport in the distal convoluted tubule is active and transcellular (Quamme, 1 997). While the mechanisms of magnesium transport were unclear and speculative (Dai et al., 2001; Quamme and Dirks, 1980) more recent reports point to specific paracellular (Efrati et al., 2005; Simon et al., 1999) and transcellular (Hoenderop and Bindels, 2005) routes across barrier membranes.
Two main fluid compartments are associated with the brain, the extracellular fluid (ECF) that bathes neurons and glial cells, and the cerebrospinal fluid (CSF) located in the subarachnoid space and ventricles of the brain (Figure 1).
Exchange between the blood and these fluid compartments can potentially occur at four sites, which have barrier mechanisms: (1) brain capillary endothelial cells, which separate the ECF from the blood and form the BBB; (2) the choroidal epithelial cells at the ventricular surface of the choroid plexuses, which interface between the blood and the CSF and form the blood-CSF barrier (BCSFB); (3) the arachnoid mater, which surrounds the CSF contained within the subarachnoid space; (4) the circumventricular organs of the brain, which have fenestrated capillaries. The four sites form potential entry routes for water, molecules, electrolytes, toxins, pathogens and drugs (Segal, 2000).
The arachnoid mater forms an external barrier, the morphological substrates of which are the flat tightly packed mesothelial cells (the arachnoid barrier cell layer) with their numerous tight junctions. They separate the CSF from the subdural space and the dura (Fig.1), which has fenestrated capillaries (Nabeshima et al., 1975; Vandenabeele et al., 1996). The arachnoid is avascular (Alcolado et al., 1988) and, relative to brain capillaries, has a much less surface area, thus its contribution to transport from the subdural space to the CSF and brain is negligible. Also the circumventricular organs of the brain, although they have fenestrated capillaries (McKinley et al., 2003; Sunn et al., 2003), they are provided with an internal mechanism that contributes to a barrier between these sites and the surrounding brain. They have rapid venous return systems, which compensates for the leaky capillaries, effectively preventing the spread of marker molecules to the surrounding brain tissue (Hashimoto, 1988; Segal, 2000). Therefore, the two main routes that are likely to have greater involvement in the regulation of the ECF and CSF environments are the BBB and BCSFB (Abbott et al., 2010).
The concept of a BBB has been proposed over a century ago (Bradbury 1979) and is still being developed (Wolburg et al., 2009). The BBB is formed by endothelial cells of brain capillaries, the characteristics of which are influenced by the surrounding microenvironment, including astro- cytes and pericytes (Wolburg et al., 2009). Apart from capillaries in the circumventricular organs which have fenestrated endothelial cells, those in the rest of the brain are lined with specialized endothelial cells which show few endocytotic vesicles, have no fenestrations and are rich in mitochondria (Brightman and Reese, 1969; Fenstermacher et al., 1988; Reese and Karnovsky, 1967; Rubin and Staddon, 1999; Sedlakova et al., 1999). Brain endothelial cells have high electrical resistance (Butt et al., 1990) and are joined by tight junctions (TJs) (Begley and Brightman, 2003; Rubin and Staddon, 1999). These TJs have long been recognized as the sites of exclusion of protein tracers as detected by electron microscopy (Brightman and Reese 1969; Reese and Karnovsky 1967; Sedlakova et al., 1999). Freeze-fracture electron microscopy revealed that these tight junctions have an elaborate and complex arrangement of network of inter- connecting strands of intramembranous particles, that particularly cleave with the P-face leaving corresponding grooves on the E-face of replicas of the split membrane (Sedlakova et al., 1999).
The BBB therefore plays a pivotal role in the tight regulation and rigorous stabilization of the chemical composition of brain ECF against fluctuations in the plasma chemistry, thus promoting normal neuronal signaling. The privileged status of the brain being protected by the BBB has a downside, in restricting access by drugs and therapeutic agents into the brain when needed (Abbott and Romero, 1996). Although the barrier nature of the BBB has been emphasized, it is obviously a selective barrier, as it should allow passage of nutrients and electrolytes. Thus endothelial cells of brain vessels are endowed with a host of receptors, enzymes and carriers in a polarized distribution to promote selective transport (Abbott et al., 2010; Betz et al., 1980; Ohtsuki and Terasaki, 2007; Roberts et al., 2008). Brain endothelial cells express the glucose transporter Glut-1 protein (Dick et al., 1984; Harik et al., 1994; Ohtsuki and Terasaki, 2007; Pardridge et al., 1990; Pardridge, 1991), which is coupled to neuronal demand (Leybaert, 2005) and its expression increases in hypoxia (Harik et al., 1994). Rat brain endothelial cells express a barrier antigen (Sternberger and Sternberger, 1987), neutralization of which leads to reversible opening of the BBB to exogenous and endogenous proteins (Ghabriel et al., 2000; 2004). Lipophilic molecules can penetrate endothelial cell membranes by diffusion, although some of these may be effluxed via ABC transporters, such as P-glyco-protein (Begley 2004; Miller, 2010; Shen and Zhang, 2010). Macromolecules may cross the BBB via receptor- mediated transcytosis (Descamps et al., 1996), or adsorptive-mediated transcytosis (Villegas and Broadwell, 1993; Zlokovic et al., 1990). Water- soluble nutrients and metabolites can to a limited degree pass passively across the BBB, but greater bulks of water, hydrophilic molecules and electrolytes have to be transported via carrier systems and channels (Zhang et al., 2002). Amino acids, nucleosides, small peptides, organic anions and organic cations are transported via solute carriers (Abbott et al., 2010; Koepsell et al., 2003; Ohtsuki and Terasaki, 2007).
Unlike in other organs, such as the intestinal wall where the extracellular space is large and is occupied by loose connective tissue, that of the brain is very narrow as neuronal and glial membranes are closely apposed leaving a gap of 20-50 nm (Vanharreveld et al., 1965) with a small volume of ECF. A small change in the ECF volume may lead to large alterations in electrolyte concentrations, deleteriously affecting neuronal excitability (Amiry-Moghaddam and Ottersen, 2003). Hence water transport across the BBB is intricately linked to ions transport. Brain perivascular astrocytic foot processes, sub-pial astrocytes and the ependyma, which lines the ventricles, express the water channels AQP-4 (Amiry-Moghaddam and Ottersen, 2003; Li et al., 2009; Nielsen et al., 1997). More recent studies demonstrated that brain endothelial cells also express AQP-4 at their luminal and abluminal membranes, albeit at a much lower density than in astrocytic foot processes (Amiry-Moghaddam and Ottersen, 2003; Kobayashi et al., 2001). The link between water movements and electrolyte concentrations is further strengthened by the co- expression of AQP4 and Kir-4.1 potassium channels, and the discovery that water move- ment is associated with potassium fluxes in the same direction (Amiry-Moghaddam and Ottersen, 2003; Niermann et al., 2001). The BBB is relatively permeable to water, allowing a slow bulk flow (Abbott, 2004) but less so to ions (Go, 1997). Electrolytes homeostasis in the ECF is actively maintained by the BBB, as evidenced by a high Na+, K+-ATPase activity in brain endothelial cells (Bauer et al., 1990; Go, 1997; Seda et al., 1984). Potassium concentration in the ECF is kept at a lower level than that of plasma (Hansen, 1985).
The brain has four choroid plexuses that project into the two lateral ventricles and the third and fourth ventricles. Each choroid plexus contains a core of loosely arranged stroma derived from the leptomeninges. Their ventricular surface is extremely fringed (Fig.1). The core of the plexus is richly supplied with capillaries, the endothelial cells of which have fenestrations and leaky tight junctions (Wolburg and Paulus, 2010). Large molecules such as peroxidase penetrate the choroid plexus capillaries and enter the interstitial space in the core of the plexus (Brightman, 1968). As the plexus invaginates the ventricle it acquires a covering of ependyma, the cell layer that forms the lining of the ventricle (Fig.1). This ependymal covering is known as the choroidal epithelium, which, although continuous with the rest of the ependyma at the neck of the choroid plexus, is structurally and functionally different, as it becomes modified and specialized to form the BCSFB (Brightman 1968; Redzic and Segal 2004; Tripathi 1973). The choroidal epithelium therefore separates the CSF from the milieu of the plexus core with its leaky capillaries. Choroidal epithelial cells are joined by tight junctions, which occlude the paracellular route (Wolburg and Lippoldt, 2002; Wolburg and Paulus, 2010). Choroidal epithelial cells actively produce the CSF and regulate its electrolyte, protein and water content (Bradbury et al., 1963). Choroidal epithelial cells express the water channels AQP-1 (Agre et al., 1993; Nielsen et al., 1993). In addition to their main function of CSF secretion choroidal epithelial cells transport some substances from the blood to the CSF, for example, nucleotides and ascorbic acid, and actively removes other substances from the CSF (Go, 1997).
There is no barrier between the brain and CSF located in the subarachnoid space, as the cells of the pia mater on the brain surface have gap junctions and no tight junctions (Alcolado et al., 1988). Peroxidase perfused into the subarachnoid space penetrates between pia mater cells, through the basement membrane of astrocytes at the brain surface and between astrocytes (outer glia limitans) to enter the neuropil (Brightman and Reese, 1969). Also free exchange occurs between the CSF located in the ventricles and the brain ECF across the ependymal lining of the ventricles, as peroxidase injected into the ventricles enter the neuropil (Brightman and Reese, 1969), thus the CSF and the ECF of the brain equilibrate.
The kidney is the main contributor to magnesium homeostasis, thus it is not surprising that research efforts to understand magnesium transport were focused on epithelial cells of the nephron. During the last decade greater understanding of the transport of magnesium across cell membranes was gained through genetic studies in families with hereditary abnormal magnesium balance, positional cloning and knockout mice investigations (Hoenderop and Bindels, 2005). Such studies led to the identification of genes encoding proteins involved in the transport of magnesium. One human gene, the paracellin-1 (PCLN-1), encodes the protein paracellin-1. PCLN-1 mRNA was detected using RT-PCR in the thick ascending limb of Henle’s loop and the distal convoluted tubule of the kidney in the rabbit (Simon et al., 1999) and rodents (Weber et al., 2001). The PCLN-1 protein was also detected using immunocytochemistry in the same segments in rabbit kidney using an antibody to PCLN-1 (Simon et al., 1999). Using confocal microscopy and double immuno- fluorescence for PCLN-1 and occuldin, a tight junction ubiquitous protein (Furuse et al., 1993), demonstrated colocalisation of the two proteins to intercellular tight junction at both segments of rabbit nephrons (Simon et al., 1999). The paracellin-1 protein is a member of the claudin family (claudin 16) (Simon et al., 1999). Families with mutation in PCLN-1 gene show hypo- magnesemia and renal magnesium wasting (Simon et al., 1999). The paracellin-1 located in the tight junction between epithelial cells in the nephron acts as a passive channel for reabsorbing magnesium along the paracellular route being favored by the positive luminal charge. Because these families also show hypercalciurea and urinary calculi, it was suggested that PCLN-1 is also a conductance for calcium (Simon et al., 1999). Such discovery modified the previously prevalent view that tight junctions exist to prevent passage of molecules and are static; rather, they do provide selective barriers. Further studies on human recessive renal magnesium loss mapped a second locus on chromosome 1p34.2 and have identified mutations in CLDN19, which encodes the TJ protein claudin-19 expressed in the nephron tubules (Konrad et al., 2006).
Recently plasma membrane cation channels that belong to the transient receptor potential (TRP) superfamily have been described and were classified into 7 subfamilies. The closely related subfamilies TRPC, TRPV, TRPM, TRPN and TRPA are classified as group 1, while group 2 includes TRPML and TRPP (Montell 2005). The TRPM subfamily includes 8 members, of which TRPM6 and TRMP7 have been recognized as highly permeable Mg2+ transporters into cells, and have been particularly studied in epithelial cells of the nephron (Aikawa 1976; Dai et al., 2001). In familial autosomal-recessive hypomagnesemia with sec-ondary hypocalcemia, affected infants suffer from seizures and tetany due to abnormal handling of Mg2+ in intestinal absorption and renal reabsorption. Genetic analysis pointed to a gene, TRPM6, which was mutated in these patients (Schlingmann et al., 2002; Walder et al., 2002). Using immunocytochemistry, it was shown that TRPM6 protein is localized to the apical membrane of renal distal convoluted tubules epithelium, the site of active reclaiming of magnesium, and also at the brush border of apical membranes of intestinal epithelium, the main site of magnesium absorption (Voets et al., 2004). Patch-clamp analysis and measurement of intracellular magnesium indicated that TRPM6 forms all or at least part of the magnesium channel in absorbing epithelia (Voets et al., 2004). It has been suggested that mammary epithelial cells in culture, with high and low magnesium content, adapt to low magnesium availability by upregulating magnesium influx via TRPM6, and to high magnesium availability by increasing magnesium efflux primarily via Na+/Mg2+ exchange (Wolf et al., 2010). The TRPM7 also has a central role in Mg2+ homeostasis, because TRPM7-deficient cells become Mg2+ deficient, and cannot survive (Schmitz et al., 2003).
Magnesium entry into the brain correlates with maturation of the BBB. In fetal sheep and the guinea pig, there is a drop in the concentration of Mg2+ in the cerebral hemispheres, which corresponds approximately to the minimum in potassium content and the maximum in chloride and sodium contents (Bradbury et al., 1972). During rat postnatal development, brain content of magnesium shows regional variations. At day 5, magnesium is most marked in the pons and medulla and least marked in the cerebral cortex. Magnesium levels in all regions decline after day 5, in parallel with the decrease in water content and the increase in tissue weight, suggesting that the maturation of the BBB plays an important role in brain magnesium homeostasis (Chan et al., 1992).
Magnesium is able to cross the BBB (Sacco et al., 2007), and it is transported via the barrier with a net flux from the blood into the parenchyma. Active magnesium transport from blood to the extracellular fluid of the brain is evidenced by its higher concentration in the cortical extracellular fluid than its concentration in the plasma- dialysate or cisternal CSF (Bito, 1969). Magnesium administration attenuates cell death due to cytoskeletal alteration (Saatman et al., 2001) and reduces apoptosis and the expression of p53 following TBI (Lee et al., 2004). Thus its administration rectifies the decline in intracellular Mg2+ levels and cellular functions (Saatman et al., 2001). In experimental closed head trauma, treatment with magnesium restores the polarity of astrocytes, in terms of aquaporin-4 distribution, to a preinjury state (Ghabriel et al., 2006). A similar protective effect for magnesium was demonstrated in induced hypoglycemia (Kaya et al., 2001). This indicates replenishing of magnesium in the extracellular space of the brain. However, the question remains, how does magnesium cross endothelial cells of the BBB? Is magnesium transport across the BBB similar to its transport across nephron epithelial cells? If so, does Mg2+ cross the BBB via a paracellular route or a transcellular route or both?
Concerning the paracellular route, the proposed role for PCLN-1 (claudin-16) in the nephron (Simon et al., 1999), acting as a conductance for Mg2+, and its exclusive location in the kidney (Efrati et al., 2005; Weber et al., 2001), raise the possibility that other members of the claudin family may play a similar role for paracellular conductance across other barrier membranes. Members of the claudin family are involved in the formation of TJ strands in various tissues (Morita et al., 1999). There is tissue-specific expression of claudin members in tight junctions, with a specific colocalisation of occludin with some but not all claudins (Peppi and Ghabriel, 2004). Magnesium is important for cell membrane stabilization. In the CNS, the myelin sheath is an elaboration of the cell membrane of oligodendrocytes, which express claudin-11. In claudin-11-deficient mice, the TJs are absent in the myelin sheath of oligodendrocytes and mice show demyelination (Gow et al., 1999). This indicates that claudin-11 is an important component in the stability and formation of the myelin tight junctions. Also endothelial cells of brain microvessels express several claudins, including claudin-1, -3, -5 and -12 (Coisne et al., 2005; Liebner et al., 2000; Matter and Balda, 2003; Morita et al., 1999; Tsukita and Furuse, 1999; Wolburg et al., 2003). Breakdown of the BBB in experimental autoimmune encephalo- myelitis in the mouse and the leaky vessels in human glioblastoma multiforme are accom- panied by selective loss of claudin-3 in BBB TJs (Wolburg et al., 2003). Although direct link between Mg2+ permeability and claudins has not been established in the CNS similar to the link between claudin-16 and epithelial cells of the nephron, it is tempting to speculate that, in addition to their supporting role in the morphology the BBB TJs, claudins in TJs of brain endothelial cells may contribute to Mg2+ conductance at the BBB. It is also speculated here that claudin-11 in the CNS myelin, in addition to its morphological supporting role, may act as a conductance for Mg2+ across the consecutive myelin lamellae, providing a faster access route from the ECF to the periaxonal space.
Concerning the transport of Mg2+ across the BBB via a transcellular route, it may also be prudent to compare the BBB to other barrier membranes. As stated above, TRPM6 transporter protein has been localized to the apical membrane of renal distal convoluted tubules epithelium, the site of active reclaiming of magnesium (Voets et al., 2004), and TRPM7 in the kidney also has a central role in Mg2+ homeostasis (Schmitz et al., 2003). In mouse brain, using quantitative RT-PCR, TRPM3 and TRPM7 mRNA were detected at high levels, and 19 other isoforms were also present (Brown et al., 2008; Kunert-Keil et al., 2006). Also in human brain TRPM1, 2, 3, 6 and 7 have been detected using RT-PCR (Fonfria et al., 2006). Evidence of TRPM presence in brain endothelial cells was obtained from endothelial cell cultures. Immortalized mouse brain microvessel endo- thelial cells, freshly isolated cerebral microvessels and primary cultured rat brain endothelial cells express multiple TRPC and TRPV isoforms, and also TRPM2, M3, M4, and M7 mRNA (Brown et al., 2008). We would like to suggest therefore that the TRPM6 and TRPM7, the gatekeepers of magnesium (Schlingmann et al., 2007), are likely to play a main role in brain endothelial cell transport of magnesium via the transcellular route. Western blotting and immunofluorescence assays for TRPM6 and TRPM7 similar to those performed for TRPV4 in mouse cerebral microvascular endothelial cells (Ma et al., 2008) would support this suggestion. Also direct evidence may be obtained from measuring intracellular magnesium in endothelial cell culture similar to previous reports on mammary epithelial cells in culture (Voets et al., 2004)
What has become apparent from experimental research and human studies is that magnesium concentration in the CSF is actively maintained above that of the plasma, and changes in CSF magnesium lag behind changes in its concentration in the plasma and are less pronounced (Bradbury and Sarna, 1977; Morris, 1992). Thus electrolyte composition in the CSF of adult mammals is different from that of plasma and is more stable, suggesting the existence of mechanisms for electrolyte homeostasis in the CSF, necessary for normal brain function (Bradbury et al., 1972; Somjen, 2002). The mechanisms determining the concentrations of ions in CSF appear to develop at different and largely independent rates (Bradbury et al., 1972). In the dog (Oppelt et al., 1963) and human (Nischwitz et al., 2008) the concentration of magnesium is higher in the CSF compared to plasma. Also in fetal sheep and guinea pigs magnesium concentration in the CSF is slightly higher than that of plasma (Amtorp and Sorensen, 1974; Bradbury et al., 1972). An investigation of the permeability of the human BCSFB to magnesium carried out in 29 individuals, using paired serum and CSF samples, showed a mean CSF/serum ratio of 1.3 (Nischwitz et al., 2008). This study also compared magnesium to other metals, and suggested that low molecular weight species such as magnesium and calcium cross the BCSFB through less specific ion channels compared to high molecular weight metals such as iron, copper and zinc, which can pass the barrier only via well controlled receptor mediated pathways (Nischwitz et al., 2008)
The choroidal epithelium plays an active role in maintaining the level of magnesium in the CSF by sensing changes in the CSF and altering the rate of active magnesium secretion (Oppelt et al., 1963; Reed and Yen 1978). Physiological studies of isolated choroid plexus of sheep (Allsop 1986) and cat (Reed and Yen 1978) showed that the choroid plexus is able to transfer magnesium against a concentration gradient. An in vivo study demonstrated a directional flow of magnesium into the CNS (Allsop and Pauli, 1985). In normal cows the magnesium concentration in the ventricular CSF was found to be higher than its concentration in the lumbar CSF (Allsop and Pauli, 1985). Under hypomagnesemic conditions, magnesium concentration in the ventricle decreased more rapidly than that of the lumbar CSF, while intravenous infusion of magnesium led to increased magnesium in the ventricular CSF before changes in the lumbar CSF (Allsop and Pauli, 1985). This study indicates that the access of magnesium from the blood to the brain is mainly via the choroid plexus to CSF, which equilibrates with the CNS parenchyma. Transport of magnesium in the choroid plexus is influenced by other selective electrolytes. For example, high levels of potassium in the perfusate of the choroid plexus leads to a reduction in the transfer rate of magnesium, but higher levels of calcium does not have a similar effect on the transfer of magnesium (Allsop, 1986).
A recent clinical trial on the protective role of magnesium in traumatic brain injury (Temkin et al., 2007) failed to replicate the improved outcome detected in experimental studies on rodents (Temkin et al., 2007; Vink and Cernak, 2000). In this clinical trial, plasma magnesium was maintained at higher levels compared to the placebo group and normal plasma level, to counteract the reported decline in Mg2+ in TBI patients (Kahraman et al., 2003). Since under normal conditions choroidal epithelial cells actively transport magnesium into the CSF, it is likely that such activity was disrupted in TBI cases in humans leading and failure of magnesium entry to the CNS. It is relevant here to state that in the rat, TBI induces severe morphological changes in choroidal epithelial cells detected by scanning and transmission electron microscopy that were still evident 4 weeks after trauma (Ghabriel et al., 2010).
Although one cannot dismiss the possibility that new channels may be involved in the transport of magnesium at the choroid plexus, it is natural to look at other systems for similar magnesium transport mechanisms. As stated above the strongest evidence for magnesium transport was obtained from genetic studies in familial hypomagnesemia, positional cloning and knockout mice investigations (Konrad et al., 2006; Schlingmann et al., 2005; Schlingmann et al., 2007; Simon et al., 1999; Walder et al., 2002) which identified TRPM6 and TRPM7 as candidates for transcellular transport, and Claudin16 (Efrati et al., 2005; Simon et al., 1999) for paracellular transport in the kidney. Recently TRPM7 has also been identified in smooth muscle cells in the vasculature (Callera et al., 2009).
Although TRPM6 and TRPM7 were detected in the brain using RT-PCR it is not clear if these channels are expressed in choroidal epithelial cells. The role of claudin-16 and claudin-19 in magnesium reabsorption from the kidney filtrate has been discussed in previous sections, and the role of claudins as selective conductance to ions through tight junction has been stated. The choroidal epithelial cells show selective expression of claudin-1, -2 and -5 at and near to their tight junctions. Also the endothelial cells within the choroid plexus show stronger expression of claudin-5 than claudin-1 and -2 (Lippoldt et al., 2000). However, claudin-16 and claudin-19 have not been reported in the choroid plexus. Whether other claudin members may contribute to magnesium conductance in the choroid plexus is speculative. Thus it appears that the channels for magnesium transport in the choroid plexus remain to be clarified.
Supported, in part, by funding from the Neurosurgical Research Foundation (Australia).
This book is copyright. Apart from any fair dealing for the purposes of private study, research, criticism or review as permitted under the Copyright Act, no part may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise without the prior written permission. Address all inquiries to the Director at the above address.
Your browsing activity is empty.
Activity recording is turned off.
See more...
1 and