Table of Contents
The free calcium inside the cell is essential for its functions. It plays a role in muscle contraction, gland secretion and neurotransmitter release. It is involved in the cell growth and differentiation into different types. It is also indispensable for neuronal excitability. Increase in the calcium inside the cell beyond a certain limit may result in cell damage. For this reason, calcium requires a mechanism to maintain a dynamic level that changes in time and space according the needs of the cell.
Neuronal cells have a resting calcium level. The remaining fraction of calcium is bound to membranes and cytoplasmic components or stored in organelles such as mitochondria, dense-cored vesicles and small synaptic vesicles. Outside the cell, calcium is bound to plasma membranes with its phospholipids, proteins, gangliosides and glycolipids. It exchange rapidly at these sites. Calcium takes part in gene expression and transmitter storage or release.
The most important calcium stores inside the cell are endoplasmic reticulum and microsomes where it is rapidly exchanged with the cytoplasm of the cell.
Endoplasmic Reticulum (ER) and microsomes are probably the most important Ca2+ pools in terms of rapid exchange with the cytosol. For example, smooth ER in cerebellar cortex has a high Ca2+ content, which can increase five-fold after prolonged depolarisation. ER organelles are heterogeneous in composition (and also in function). In the ER, calciosomes have been shown to exist in a number of cells. They are exclusively in a class of smooth vacuoles and are apparently distinct from, although often adjacent to, other ER organelles. They are rich in Ca2+-binding proteins such as calsequestrin and contain a Ca2+-ATPase which is immunologically distinct. Because of their specific molecular components, calciosomes appear to be well suited to act as rapidly exchangeable, membrane-segregated Ca2+ pools.
It should be noted that the information described above is applicable to cells which are not being stimulated by external signals. Information is now accummulating which indicates that when neurones in culture are subjected to electrical field stimulation (e.g 1 Hz for 15 sec) or to depolarising concentrations of K+, there is a 3-5 fold increase in Ca2+ but that this is unevenly distributed throughout the cell. The greatest increases apparently occur in the cell’s nucleus and at the terminal regions. The rise in nuclear [Ca2+]i may be related to changes in gene transcription whereas the changes at the nerve endings are probably associated with neurotransmitter release or possibly with synaptic plasticity.
Cells regulate [Ca2+]i mainly by controlling Ca2+ movement across the plasma membrane, and/or the membranes of ER, mitochondria or calciosomes. [Ca2+]i increases are normally due to entry from the extracellular fluid via channels as a result of a depolarisation or by release of Ca2+ from intracellular stores. The increased [Ca2+] initially interacts with its effectors and then multiple mechanisms operate to buffer it in the cytosol, sequester it in and/or mobilise it from intracellular stores, and extrude it across the plasma membrane. Eventually, all of this Ca2+ must be extruded to maintain Ca2+ homeostasis. Although Ca2+ entry by exchange and diffusion does not significantly contribute to influx under normal conditions, in states such as ischaemia, influx by these pathways is greatly enhanced and may contribute to irreversible cell damage.
Cytosolic Ca2+-buffering capacity has high and low affinity components. Various molecules, e.g. citrate, nucleotides and inositol phosphates, account for the low affinity binding of Ca2+ (<100 nM/mg protein) and hence are of dubious importance. The high affinity is due primarily to proteins, most of which share the typical EF hand structure of their binding sites e.g. calmodulin, parvalbumin, and vitamin D-dependent Ca2+-binding protein (CaBPr): the distribution and concentration of these proteins varies in different neurons. Calmodulin, which is highly concentrated in brain (30-50 nM), and widely distributed among neurons, may account for a Ca2+-buffering capacity of 120 nM. Although Ca2+-binding proteins have a relatively higher buffering capacity for Ca2+ in neurons than in smooth muscle and liver, they are only able to buffer Ca2+ that enters the neuron during the first few action potentials (about 5 pmol /mg protein/ms if Ca2+ is evenly distributed in the cytosol). Neurons firing at high frequency, and especially in those in which a relatively large fraction of the inward current during the rising phase of a action potential is carried by Ca2+, these cytosolic Ca2+ buffers may saturate.
Ca2+ sequestration and mobilisation. Cytosolic proteins buffer transient rises in [Ca2+]. Intracellular pools are then required to sequester it until it can be extruded. Because of the low affinity of Ca2+ for the mitochondrial Ca2+ transporter, there is only appreciable accumulation when [Ca2+]i is increased by neuronal stimulation. when [Ca2+]i rises as a result of neuronal activity, mitochondria respond by a net (ATP-dependent) uptake of Ca2+ and in the matrix it increases up to approximately. When [Ca2+]i exceeds 5~M (i.e. under pathological conditions ), mitochondria sequester substantial amounts.
Synaptic vesicles sequester Ca2+ by an ATP-driven mechanism, but they have a low affinity and therefore probably contribute little to [Ca2+]i regulation.
ER and microsomes are the major non-mitochondrial Ca2+ stores and probably play the most important role in sequestering Ca2+ following neuronal activity. They accumulate it in an ATP-dependent fashion at a rate of 1-2 pmol/ mg protein/ms which is sufficient to remove Ca2+ from the cytosol following its entry. In neurons, calciosomes have not been investigated in detail: however, calciosome-like proteins are expressed, in cultured neurons at least, and appear to be localised in cell bodies and neurites. Depolarisation, hormones and ionomycin can induce increases in [Ca2+]i even in the absence of [Ca2+]0 . These increases are believed to result from Ca2+ mobilisation from the ER and mitochondria. Release of Ca2+ from the ER is quantal and more will be released as the ER fills. Indeed, the ER may need to be loaded in this way to contribute significantly to increased [Ca2+] during neuronal activity. Evidence suggests that during excitation, release of Ca2+ from the ER, in particular those rapidly exchangeable pools,is modulated by IP3: decreasing [Na+]0 triggers IP3 production and Ca2+ mobilisation. However, whether Ca2+ release from calciosomes is triggered by IP3 remains to be established. Mitochondria are not involved in IP3-mediated Ca2+ mobilization, but release it by a Na/Ca2+ exchange mechanism.
Calcium extrusion The concentration gradient, [Ca2+]0 » [Ca2+]i (10~-fold), and the electrical driving force (50-100 mV negative membrane potential) promote the net gain of Ca2+ at rest and during neuronal activity. To maintain Ca2+ homeostasis, neurons have two parallel, independent mechanisms in their plasma membranes for extruding Ca2+: Ca2+/Mg2+ ATPase and a Na+/Ca2+ exchanger.
Two general classes of Ca2+ pumps (Ca2+/Mg2+ ATPase) have been identified: those in the ER, that participate in intracellular Ca2+ sequestration, and those in the plasma membrane, that extrude it from cells. The ER systems are calmodulin-insensitive, Ca2+-dependent ATPases with coupling ratios of Ca2+:ATP 2:1 and MW of about 105 KDa. The plasma membrane systems are calmodulin-modulated, Ca2+-dependent ATPases with coupling ratios of Ca2+:ATP 1:1 and MW of about 140 kDa: they have a high affinity for Ca2+ but a low transport capacity and are proposed to be primarily responsible for Ca2+ efflux in the resting state.
The Na+/Ca2+ exchanger :
An electrogenic Na+/Ca2+ exchanger extrudes 1 Ca2+ in exchange for 3 Na+ by utilising the energy stored within the Na+ gradient across the plasma membrane, which is in turn restored by the Na+/K+ ATFase. The direction of Na+/Ca2+ exchange can be either inward or outward depending on the Na+ gradient. Initially, Ca2+ flux through this antiporter would be inward after depolarisation due to the increase in intracellular Na+. However, after repolarisation promoted by Na+/K+ ATPase, there would primarily be an outward Ca- current. The rate of transport mediated by the Na+ / Ca2+ exchanger is controlled by factors such as the difference between the membrane potential and the reversal potential for the exchanger. It is also influenced by kinetic factors based on the fractional occupancy of the carriers by transported ions as well as by activating (non-transported) internal Ca2+. The system has a low affinity for Ca2+ but a large capacity. Thus, when [Ca2+)i is low, e.g. in resting neurons (-100 nM), the turnover of the exchanger is very low because the internal Ca2+ sites that participate in Ca2+ extrusion as well as those that activate exchanger-mediated Ca2+ entry are largely unoccupied. This means that the exchanger operates primarily to restore [Ca2+]i during depolarisation when [Ca2+]i is raised.
Changes in [Ca2+]i following agonist addition to neurons have shown that the responses may be oscillatory. Mechanisms involved in the generation of Ca2+ oscillations appear to be of two types: oscillations that occur secondary to spontaneous action potentials as in some secretory and retina cells and oscillations induced by Ca2+-mobilizing substances, e.g.IP3. The latter effect can be initiated in the absence of extracellular Ca2+ and thus, intracellular Ca2+ release is the primary source of Ca2+ although extracellular Ca2+ is required for maintenance of Ca2+ oscillations. A minimum oscillating system requires a feedback loop, and for sustained oscillations, some delay step is necessary to generate the periodicity. A possible mechanism for negative feedback is provided by the inhibitory effect of Ca2+ on IP3 binding to its receptor, i.e. the time required for reaccumulation of Ca2+ into the IP3-sensitive pool could provide the delay which would allow the generation of the characteristic periodicity.
An increase in [Ca2+] triggers exocytotic release of neurotransmitters. Several ion channels that help to shape the frequency and duration of electrical responses in neurons are regulated directly by Ca2+ or indirectly by Ca2+-dependent protein kinases or phosphatases . Synapse formation is linked to [Ca2+]. Actively elongating neurites and motile growth cones have a narrow range of [Ca2+] of 100-300 nM: a decrease below or an increase above this range produced by neurotransmitters or electrical activity is associated with the arrest of neurite elongation and cessation of growth cone movement . Associative long-term potentiation (LTP), an activity-dependent, persistent increase in synaptic strength, is initiated by a local increase in [Ca2+]i . Finally, transient changes in [Ca2+]i can trigger the transcription of “immediate-early genes” such as c-fos, and c-jun in neuronal nuclei which encode DNA-binding proteins that alter the express-ion of other genes.
In neurons, there are two major classes of Ca2+ targets in the plasma membrane and three in the cytosol. In the plasma membrane, the first target group are channel proteins ie certain K+ channels, cation-selective channels, and Cl- channels are directly regulated by Ca2+ : these are distinct from the well known receptor operated (ROCC) calcium channels (N, L,F and T). The second target group in the plasma membrane are two important families of membrane phospholipases : these are phospholipase C (PLC), which hydrolyses phosphatidylinositol phosphates (PIPs), and phospholipase A2 (PLA2), which cleaves fatty acids from the glycerolipid backbone . The products of phospholipase C are well established second messengers eg. diacylglycerol( DAG) and inositol 1,4,5, trisphosphate (1P3): DAG activates the various isoforms of protein kinase C and IP3 mobilises calcium from intracellular stores. PLA2 activation leads to the formation of the detergent lysolecithin and to various fatty acids including arachidonate. Arachidonic acid production is the rate limiting step in eiconosoid production and is possibly involved in the generation of synaptic plasticity. Thus, prolonged activation of PLA2 is likely to result in uncontrolled membrane deacylaction and generation of excess second messengers.
The three major cytosolic targets for calcium are protein kinase C (PKC), calpain (a Ca2+-dependent protease) and calmodulin (CaM). PKC exists as a family of 83 I:Da serine/threonine protein kinases eg PKC1, PKC11 and PKCI11, that are activated synergistically by Ca2+ and diacylglycerol (DAC) (34). They are also activated by phorbol esters. In their activated state, cytosolic PKC isoforms (of which their are at least seven) move to the plasma membrane where they phosphorylate and regulate membrane proteins. In neurons, PKC regulates electrical excitability by phosphorylating certain ion channels. In addition, they can regulate synaptic efficacy and appear to have a role in the initiation and/or main tenance of LTP. PKC has been implicated in the intracellular changes which result from toxic exposure to methyl mercury. For example, in cell culture, the abnormal protein phosphorylation of neuronal but not glial proteins which follows treatment with methyl mercury is similar to that seen following addition of phorbol esters and is blocked by the adition of staurosporine.
Calpains are a group of neutral cystei ne proteases that are activated directly by Ca2+ . They have a widespread distribution, are present in both membrane and cytosolic forms, and are implicated in the regulation of membrane proteins (eg. receptors) and the cytoskeleton (eg. tubulin and fodrin are substrates) in a number of cells, including neurons. Obviously, they are likely to be involved in the catabolism of a variety of proteins including other enzymes and thus the potential for causing cell damage is considerable: in fact there are some reports of increased calpain activity in age-related pathologies. In the hippocampus, excitotoxicity due to administration of NMDA results in loss of CAl neurons (similar to that seen following global ischaemia): these pathological changes are associated with calpain mediated proteolysis of the major cytoskeletal protein spectrin (such changes are clearly seen using Western blots and image analysis).
C~pains have an inactive form, procalpain which is converted to the active enzyme by Ca dependent proteolysis. This step is irreversible. The active form has increased sensitivity to Ca and under these conditions, Ca promotes dissociation of the natural inhibitor, calpastatin.
In its Ca2+-bound form, CaM binds with different affinities to specific proteins and alters their functions. Among these proteins are a family of Ca2+/CaM-dependent protein kinases. CaM kinase II is the predominant Ca2+-dependent protein kinase in neurons of the mammalian cortex and hippocampus. It has a relatively broad substrate specificity and can phosphorylate several neuronal proteins. It is presumed to be a target for the postsynaptic Ca- current produced by activation of N-methyl-D aspartate (NMDA) receptors. Furthermore, its rapid on/off regulation by autophosphorylation suggests that it could play a role in the initiation of LTP. In the presynaptic terminal, CaM kinase II may mediate synaptic transmission and increase glutamate and NA release from synaptosomes. One of its important presynaptic substrates is synapsin I, a protein that associates with synaptic vesicles and binds to the cytoskeleton. Phosphorylation by CaM kinase reduces the affinity of synapsin I for vesicles: this causes dissociation of vesicles from the cytoskeleton, making them available for fusion and leading to increased neurotransmiUer release.
A second neural protein activated by Ca2+/CaM is protein phosphatase-2B (calcineurin). Calcineurin is abundant in brain, but has a rather narrow substrate specificity. It was recently implicated in the Ca2+-dependent inactivation of L-type Ca2+ channels in neurons. Phosphorylation of L-type Ca2+ channels by cAMP-dependent protein kinase enhances their activation by depolarisation. Conversely, dephosphorylation by calcineurin d esensitises the channels. Calcineurin also dephosphorylates and inactivates DARRP-43, a protein inhibitor of the broad-specificity brain protein phosphatase, phosphatase-I. Thus, activation of calcineurin may initiate a cascade of dephosphorylation.
Two additional targets of Ca2+/CaM are isozymes of adenylate cyclase and cyclic nucleotide phosphodiesterase. Thus a rise in [Ca2+]i may promote either production or degradation of cAMP depending on the nature of the local cyclases and phosphodiesterases. In addition, Ca2+/CaM participates in the feedback control of [Ca2+]i by activating membrane Ca2+/Mg2+ ATPase that functions as a Ca2+ pump.
There are reports that several growth factors can protect CNS neurons against a variety of insults. For example, fibroblast growth factor (FCF) protects hippocampal neurons against glutamate neurotoxicity and FGF, nerve growth factor (NCF) and insulin-like growth factors (ICFs), protect human cortical and rat hippocampal and septal neurons against hypoglycemic damage. In each case, the growth factor prevents the sustained elevation in intracellular calcium levels that normally mediate the cell damage. The processes which are involved in these protective effects are unclear but, for example, may involve enhanced calcium buffering or alterations in the expression of the NMDA-type glutamatergic receptor.
In the sections above, the putative sites of calcium’s cytotoxic effects have been described. However, there are some important unresolved issues. For example although excitotoxicity proceeds via an NMDA type glutamatergic receptor mediated increase in calcium influx, it is clear that high external K+ concentrations or cyanide exposure also increase [Ca2+]i but do not necessarily result in cell death. Thus, it is likely that excitotoxic damage is due to changes in [Ca2+]i in spatially or temporally restricted domains.
The importance of temporal factors is apparent from studies of hippocampal neurones exposed to toxic levels of glutamate eg there is a period in which [Ca2+]i levels are apparently normal even although irreversible injury may have occurred.
The significance of spatial aspects of calcium distribution are strongly suggested by optical measurements which have shown that increases in [Ca2+]1 are often confined to specific parts of a neuron. Therefore, responses to a rise in [Ca2+]i will depend on the spatial organisation of Ca2+ target proteins, their relative affinities for Ca2+ or for Ca2+/CaM, and the arrangement of more distal proteins in the response pathway. Factors that will influence a local response include clustering of Ca2+ target proteins within the membrane and their association with the cytoskeleton. In this context, it should be noted that the affinities for Ca2+/CaM vary widely among its target proteins. Calcineurin and Ca2+/Mg2+ ATPase have relatively high affinities for Ca2+/CaM (Kds: 5 nM), whereas CaM kinase II has a considerably lower affinity (Kd: 50 nM) (30). Thus, calcineurin will bind a larger proportion of available Ca2+/CaM than the CaM kinase II when the two enzymes are present at the same concentration. However, in general, information on the subcellular distribution and concentration of most neuronal Ca2+ target proteins is still inadequate to permit quantitative predictions of local cellular Ca2+ responses.
These various findings described above indicate that in most cases the mechanisms involved in calcium mediated cell damage are still ill defined in terms of local intracellular events. Finally, it should be borne in mind that elevation of [Ca2+]i could be as a terminal consequence of cell damage eg as a result of metabolic compromise and thus the increase may not have any causal significance.
The concentration of extracellular calcium is in the millimolar range whereas free intracellular calcium is nanomolar (measurement of free calcium in cells has become possible largely because of the development of calcium-binding fluorescent dyes). The massive external:internal concentration gradient is maintained by a large number of factors and this can probably be seen as a reflection of the importance of calcium homeostasis in cells. Homeostatic failure can obviously arise as a result of a large variety of toxic insults eg. it could result from metabolic changes or could be due to a specific effect on (for example) glutamatergic (NMDA) receptors. Therefore, in the former case, increases in intracellular free calcium are likely to be a result rather than a cause of neuronal damage.
There is data which shows that increases in ~Ca2+]i resulting from toxic insult are not uniform throughout the cell but that calcium’s cytoxicity has both a spatial and temporal component.
There is now some evidence that beta-amyloid protein potentiates the toxicity of glutamate which is produced by increases in neuronal free calcium. This may be important in the aetiology of Alzheimer’s disease. There is also some evidence that growth factors such as fibroblast growth factor (FCF) protect cells from glutamate induced damage.
The consequence of increased levels of free calcium in cells is that a large number of enzymes become activated (eg. proteases) and hence cell damage can occur. Elucidation of the specific effects of many of these enzymes in the damage process remains to be accomplished.